Communications
(Scheme 2a). The 1-phenylethanol derivatives 1b–1d bearing
electron-donating substituents in ortho, meta, and para posi-
tions gave the desired products 3b–3d in good yields. Similar-
ly, the electron-deficient substrates 1e and 1 f furnished the
upgraded alcohols 3e and 3 f in good yields. Furthermore, the
alkylation of the naphthyl substrate 1g afforded the expected
product in 71% yield. Importantly, the substrate scope could
be extended to the use of the less active aliphatic secondary
alcohols 1h–1l; the corresponding alcohols 3h–3l were ob-
tained in good-to-moderate yields. Noteworthy, even the steri-
cally demanding alcohol 1k was tolerated in this Mn-catalyzed
reaction.
Table 1. Optimization of the reaction conditions.[a]
Entry Catalyst
Base
Solvent
Conv. Yield of 3a/3a’
([mol%]) ([mol%])
[%]
[%]
After successfully varying the secondary alcohols, we
became interested in studying the scope of the b-alkylation of
1
2
3
4
5
6
7
Mn-1 (3) Cs2CO3 (10)
Mn-2 (3) Cs2CO3 (10)
Mn-3 (3) Cs2CO3 (10)
Mn-4 (3) Cs2CO3 (10)
Mn-1 (3) Cs2CO3 (10)
Mn-1 (3) Cs2CO3 (10)
Mn-1 (3) Cs2CO3 (10)
TAA
TAA
TAA
TAA
1,4-dioxane
2-Me-THF
toluene
57
49/8
42/14
69/30
55/34
20/3
30/4
50/22
50/8
65/8
76/5
81/5
57/17
92/8
60
>99
90
62
60
73
77
90
93
88
85
>99
>99
>99
1-phenylethanol (1a) with different primary alcohols
2
(Scheme 2b). The alkylation of 1a with the electron-rich benzyl
alcohols 2b–2 f as well as the electron-poor substrates 2g–2i
resulted in the secondary alcohols 4b–4i in good yields. Also,
the alcohol 4j containing a naphthyl group was obtained in
77% yield. Additionally, various aliphatic primary alcohols
could also be used as a coupling partner to afford the alcohols
4k–4m.
Encouraged by these promising results, we decided to inves-
tigate the more challenging coupling reaction between satu-
rated aliphatic primary and secondary alcohols (Scheme 2c).
Indeed, our catalytic system also proved to be suitable to
couple the branched cyclohexanemethanol and unbranched 1-
octanol with 1-cyclohexylethanol and the products 5a and 5b
were isolated in good yields. Furthermore, 1-cyclopenylethanol
and 1-cyclopropylethanol were alkylated with different linear
and non-linear alcohols to produce the alcohols 5c–5e in
good yields.
8
Mn-1 (3) KHMDS (10) toluene
9
Mn-1 (3) KH (10)
toluene
toluene
toluene
toluene
toluene
toluene
toluene
10
11
12
13
14
15[b]
Mn-1 (3) KOH (10)
Mn-1 (3) KOtBu (10)
Mn-1 (3) KOtBu (5)
Mn-1 (3) KOtBu (25)
Mn-1 (1) KOtBu (25)
Mn-1 (1) KOtBu (25)
92/8
83/13
[a] Reaction conditions: 1a (0.5 mmol), 2a (0.55 mmol), [Mn] and base in
0.5 mL of solvent at 1358C in a glass tube under argon for 20 h. Conver-
sions and yields were determined by the H1 NMR analysis of the crude re-
action mixture using mesitylene as an internal standard. TAA=tert-amyl
alcohol. [b] A drop of mercury was added.
3 mol% of our PNN complex Mn-1 led to promising results,
providing the desired product 3a in 49% yield along with 8%
of the corresponding ketone 3a’ (Table 1, entry 1). The applica-
tion of the di-tert-butyl complex Mn-2 resulted in similar dehy-
drogenation activity and slightly lower hydrogenation activity
(Table 1, entry 2).[16f] In the presence of PNP complexes, such as
Mn-3 and Mn-4, excellent conversion was observed. However,
the inefficient hydrogenation of the unsaturated intermediate
provided considerable amount of the undesired ketone 3a’
(Table 1, entries 3 and 4). Thus, we decided to further optimize
the model reaction using Mn-1 in combination with different
bases and solvents. 1,4-Dioxane and 2-Me-THF proved to be
unsuitable for this reaction. However, performing the reaction
in toluene resulted in better results (Table 1, entries 5–7). Addi-
tionally, we tested various bases such as KHMDS, KH, KOH, and
KOtBu (Table 1, entries 8–13). From these experiments the reac-
tion proceeded best when 25 mol% of KOtBu were applied
and the desired product was obtained in 92% yield (Table 1,
entry 13). Interestingly, we could reduce the catalyst loading to
1 mol% while still obtaining excellent yield (Table 1, entry 14).
Finally, full conversion was still observed upon adding Hg to
the reaction mixture, proving the homogenous nature of the
Mn catalyst (Table 1, entry 15).
To gain insight into the reaction mechanism, we performed
deuterium-labeling experiments (Scheme 3). When 1-phenyle-
thanol-a-d1 [D1]1a was reacted with benzyl alcohol-a,a-d2
[D2]2a, a very strong kinetic isotope effect was observed and
3a’’ was obtained in 31% yield with 86% deuteration in the a-
position and 60% deuterium incorporation at C3. No deutera-
tion occurred at the OH moiety and only 10% deuterium incor-
poration occurred at the C2 position (Scheme 3a). Similarly,
the reaction between [D1]1a and 2a gave the alkylated prod-
uct with deuterium incorporations of 32% at C1, 15% at C3,
and >3% at C2 (Scheme 3b). The presence of only 40% deu-
terium incorporation at C3 in the reaction between 1a and
[D2]2a indicates the reversibility of the primary alcohol dehy-
drogenation process and supports the hydrogen autotransfer
pathway (Scheme 3c).
Interestingly, the high deuterium content at C1 and C3 and
the low deuterium incorporation at C2 are not in alignment
with both the classical dihydride mechanism and proposed
amidate-assisted pathway.[16e] The presented deuterium experi-
ments support a monohydride mechanism and highlight the
involvement of both the metal and the non-innocent ligand in
the transfer-hydrogenation pathways.[18–20]
The metal monohydride can be formed by the b-hydride
elimination of the Mn alkoxide (inner sphere pathway). Alterna-
tively, the alcohol can be (de)hydrogenated through the outer
Next, the substrate scope of the b-alkylation of different sec-
ondary alcohols 1 with benzyl alcohol 2a was conducted
&
ChemSusChem 2018, 11, 1 – 5
2
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!