Tunable Metal-Catalyzed Heterocyclization Reactions of Allenic Amino Alcohols: An Experimental and Theoretical Study

Article abstract of DOI:10.1021/acs.joc.6b00934

Controlled preparation of 2,5-dihydro-1H-pyrroles, 3,6-dihydro-2H-pyrans, and pyrroles has been achieved through switchable chemo- and regioselectivity in the metal-catalyzed heterocyclization reactions of allenic amino alcohols. The gold-catalyzed cycloi

Full text of DOI:10.1021/acs.joc.6b00934

Subscriber access provided by United Arab Emirates University | Libraries Deanship
Article
Tunable Metal-Catalyzed Heterocyclization Reactions of Allenic
Amino Alcohols: An Experimental and Theoretical Study
Benito Alcaide, Pedro Almendros, Cristina Aragoncillo, Gonzalo
Gómez-Campillos, M. Teresa Quirós, and Elena Soriano
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.6b00934 • Publication Date (Web): 27 Jul 2016
Downloaded from http://pubs.acs.org on August 4, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Tunable Metal-Catalyzed Heterocyclization
Reactions of Allenic Amino Alcohols: An
Experimental and Theoretical Study
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
,†
,‡
†
Benito Alcaide,* Pedro Almendros,* Cristina Aragoncillo, Gonzalo
†
†
‡
GómezꢀCampillos, M. Teresa Quirós, and Elena Soriano
†
Grupo de Lactamas y Heterociclos Bioactivos, Departamento de Química Orgánica I,
Unidad Asociada al CSIC, Facultad de Química, Universidad Complutense de Madrid,
8040ꢀMadrid, Spain
2
‡
Instituto de Química Orgánica General, IQOGꢀCSIC, Juan de la Cierva 3, 28006ꢀ
Madrid, Spain
Eꢀmail: alcaideb@quim.ucm.es; Palmendros@iqog.csic.es
ABSTRACT
Controlled preparation of 2,5ꢀdihydroꢀ1Hꢀpyrroles, 3,6ꢀdihydroꢀ2Hꢀpyrans, and
pyrroles has been achieved through switchable chemoꢀ and regioselectivity in the metalꢀ
catalyzed heterocyclization reactions of allenic amino alcohols. The goldꢀcatalyzed
cycloisomerization reaction of αꢀaminoꢀβꢀhydroxyallenes was effective as 5ꢀendo
cyclization by addition of the amino functionality to the distal allene carbon to yield
enantiopure 2,5ꢀdihydroꢀ1Hꢀpyrroles; whereas their palladiumꢀcatalyzed cyclizative
coupling reactions furnished 3,6ꢀdihydroꢀ2Hꢀpyrans through
a
chemoꢀ and
regioselective 6ꢀendo cycloetherification. In opposition, the goldꢀcatalyzed
heterocyclization reaction of βꢀaminoꢀγꢀhydroxyallenes exclusively generated pyrrole
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
derivatives. These results could be explained through a chemoꢀ and regioselective 5ꢀexo
aminocyclization to the central allene carbon, followed by aromatization. Chemoꢀ and
regioselectivity depend on both the linker elongation as well as the type of catalyst. This
behaviour can be justified by means of Density Functional Theory calculations.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table of Contents (TOC)
HO
(
)n
•
ZHN
palladium salt/
or gold salt
cross-coupling reagent
NHZ
or
or
HO
HO
N
N
Z
O
Z
R
NHZ
R
or HO
or
N
Z
O
INTRODUCTION
Cycloisomerization reactions have been developed as the most efficient, atomꢀ
economical synthetic strategy to build up heterocycles. In particular, catalytic
intramolecular addition of a pendant Y–H group to an allene moiety avoids byꢀproduct
1
formation of classical substitution reactions. How to control chemoselectivity to get the
desired adducts over unwanted isomers is the principal issue in these cyclization
reactions.
Despite that this performance has previously been reported for allenols and
allenamines, related transformations for hydroxy allenamines are undeveloped;
probably because of additional chemoselectivity problems. An interesting case of
2
ACS Paragon Plus Environment

Page 3 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
selectivity arises in allenic amino alcohols, because two potentially reactive moieties,
namely, alcohol and amine, are present in the same substrate. In 2009, the Krause group
2
reported the goldꢀcatalyzed
5ꢀendo cycloisomerization of Nꢀ(butaꢀ2,3ꢀ
3
dienyl)hydroxylamines to give Nꢀhydroxypyrrolines. Apart from the above mentioned
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
aminocyclization reaction, there exist too few reports on the heterocyclization reactions
4
of allenic amino alcohols. Even so, the switchable synthesis of different heterocycles
from the same starting allenic amino alcohol is very attractive, but very difficult because
each moiety could react to afford diversely sized oxaꢀ or azacycles. Besides, it may also
5
benefit the understanding of the reaction mechanisms. We envisioned that a metalꢀ
catalyzed heterocyclization reaction of αꢀaminoꢀβꢀhydroxyallenes would produce
different 3ꢀ, 4ꢀ, 5ꢀ or 6ꢀmembered heterocycles. Depending on both the
chemoselectivity (aminocyclization versus oxycyclization) as well as the
regioselectivity (endo-trig versus endo-dig versus exo-dig versus exo-trig attack) either
of the eight possible cycloisomerization adducts could be the reaction products (Scheme
1). In consideration of the importance of both aminocyclization and cycloetherification
reactions, we decided to examine the influence of different metal activators on the
heterocyclization reaction.
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 1. Possible Cycloisomerizations that Can Take Place in αꢀAminoꢀβꢀ
hydroxyallenes
NHZ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
NHZ
HO
N
O
Z
NHZ
HO
HO
•
O
ZN
ZHN
NHZ
HO
ZN
O
HO
N
Z
RESULTS AND DISCUSSION
To explore the effects of various complexes on the metalꢀcatalyzed heterocyclization
reaction of allenic amino alcohols, αꢀaminoꢀβꢀhydroxyallene 1a and βꢀaminoꢀγꢀ
hydroxyallenes 2a,b were synthesized in optically pure form. Starting allene 3 was
made from the corresponding alkyne by treatment with paraformaldehyde in the
presence of diisopropylamine and copper(I) bromide (Crabbé reaction) using our
6
previously described methodology. Starting allenes 4a,b were obtained from Garner’s
aldehyde via the coupling with propargyl bromides promoted by indium, followed by
7
protection of the hydroxyl functionality in the form of methyl ether. Precursors for the
heterocycle formation, enantiopure allenic amino alcohols 1a and 2a,b were smoothly
4
ACS Paragon Plus Environment

Page 5 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
obtained in excellent yields by BiCl ꢀcatalyzed aminal cleavage (Scheme 2).
3
Unfortunately, we faced significant difficulties at the purification stage in the case of
highly polar substrates 2a,b. Chromatography, even with neutralized silica gel or
alumina, resulted in decomposition products. Consequently, we devised a oneꢀpot
sequence from oxazolidineꢀderived allenes 4a,b avoiding the purification of βꢀaminoꢀγꢀ
hydroxyallenes 2a,b. Thus, it was our hope that the resulting crude products displayed
sufficient purity to be engaged in the metalꢀcatalyzed cyclization step.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Preparation of Enantiopure Allenic Amino Alcohols 1 and 2
O
HO
•
BiCl
3
(20 mol%)
•
ref. 3
N
HN
Boc
(–)-1a (quantitative yield)
CH CN: H O (10:1)
3
2
Boc
RT, 24 h
(
–)-3
CHO
O
N
Boc
HO
R
MeO
R
MeO
R
•
Br
Me
SO , TBAI
2 4
,
In
R
•
NaOH (aq. 50%)
•
BiCl (20 mol%)
3
O
N
O
N
HO
NH
Boc
Boc
Boc
THF/NH
4
Cl (aq. sat.)
CH
2
Cl
2
, RT
CH
3
CN: H
2
O (10:1)
0
ºC to RT
RT, 24 h
(
(
–)-4a-OH R = Me (44%)
–)-4b-OH R = Ph (54%)
(–)-4a R = Me (67%)
(–)-4b R = Ph (61%)
(–)-2a (quantitative yield)
(–)-2b (quantitative yield)
Next, we targeted on the use of metal catalysis to the selective preparation of
heterocycles starting from allenic amino alcohols. Initially, it was hoped that a catalytic
activation strategy could be developed through the use of noble metal salts. In the
investigation of the reaction conditions, αꢀaminoꢀβꢀhydroxyallene 1a was chosen as the
model substrate. Firstly, the effect of the catalyst on the model reaction was examined.
Fortunately, a variety of gold and platinum catalysts were applied with success to αꢀ
aminoꢀβꢀhydroxyallene 1a. These reactions were completely selective for the formation
of 2,5ꢀdihydroꢀ1Hꢀpyrrole 5a (Table 1). When the reaction was catalyzed by
[
PtCl
2
(CH
2
=CH
2
)]
2
or PtCl
2
/AgOTf, comparable results of azacycle 5a were obtained
P)AuNTf ] provided lower
(Table 1). However, the addition of Gagosz’ catalyst [(Ph
3
2
yield of the cycloisomerization product (Table 1). To our delight, the use of a gold(III)ꢀ
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
based catalyst, such as AuCl , enhanced the hydroamination reaction. Thus, 2,5ꢀ
3
dihydroꢀ1Hꢀpyrrole 5a could be isolated after purification by column chromatography
3
in 77% yield after AuCl treatment. A screening of solvents (toluene, polyethylene
glycol 400, 1,2ꢀdichloroethane) revealed that the reaction is best performed in
dichloromethane. The above outcome could be rationalized invoking a 5ꢀendo
cycloisomerization by chemoꢀ and regiospecific addition of the amino group to the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
terminal allene carbon.
Table 1 Selective aminocyclization reaction of αꢀaminoꢀβꢀhydroxyallene 1a under
modified metalꢀcatalyzed conditions
X mol% metal precatalyst
HO
HO
•
Y mol% additive
N
HN
Boc
–)-1a
o
solvent, t, 20 C
Boc
(
(–)-5a
d
entry
metallic salt (X mol%) additive (Y
mol%)
solvent/t (h)
yield
1
2
PtCl
2
(5)
AgOTf (5)
CH
2
Cl
2
/23
59
52
a
[PtCl (CH =CH )] (5) TDMPP
CH Cl /23
2
2
2
2
2
2
(
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
10)
b
3
4
5
6
7
[(Ph
3
P)AuNTf
2
] (5)
DCE /3
20
68
49
75
77
AuCl
AuCl
AuCl
3
3
3
(5)
(5)
(5)
toluene/5
c
PEGꢀ400 /4
DCE /3
AuCl (5)
CH Cl /3
3
2
2
a
b
c
TDMPP = tris(2,6ꢀdimethoxyphenyl)phosphine. DCE = 1,2ꢀDichloroethane. PEG = Polyethylene
d
glycol 400. Yield of pure, isolated product with correct analytical and spectral data.
A need of modern organic synthesis is to investigate how different types of
molecules can be created from the same starting material merely through catalyst choice.
6
ACS Paragon Plus Environment

Page 7 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Thus, having found a solution for the selective allenic aminocyclization, our next aim
was to study the catalystꢀcontrolled chemoselective heterocyclization of allenic amino
alcohols in order to obtain oxacycles. Our initial investigations on Pdꢀcatalyzed
chemodifferentiation reactions focused on the heterocyclization of αꢀaminoꢀβꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
hydroxyallene 1a with allyl bromide. Worthy of note, under PdCl
2
catalysis the
reaction took place, affording 3,6ꢀdihydroꢀ2Hꢀpyranꢀ3ꢀamine derivative 6a in fair yield
Scheme 3). Disappointingly, along with cyclic ether 6a, dihydropyrrole 7a was also
(
obained as minor product. Fortunately, isomeric heterocycles 6a and 7a could be easily
separable by flash chromatography. The formation of dihydropyran 6a can be somehow
explained through a chemoselective 6ꢀendoꢀtrig allenic oxycyclization followed by
crossꢀcoupling.
Scheme 3. PalladiumꢀCatalyzed Preparation of Dihydropyran 6a and
Dihydropyrrole 7a
Br
HO
NHBoc
5
mol% PdCl2
•
HO
+
HN
Boc
DMF, RT, 2 h
N
O
Boc
(
–)-1a
(
–)-6a (45%)
(+)-7a (21%)
We thought it would be of interest to apply to αꢀaminoꢀβꢀhydroxyallene 1a our
previously developed palladiumꢀcatalyzed heterocyclization/crossꢀcoupling sequence
10
between αꢀallenols and αꢀallenic esters. Gratifyingly, it was observed that these
reaction conditions using αꢀallenic esters 8 were compatible with substrate 1a, resulting
in the formation of the desired oxacycles 9 in fair yields with complete regioꢀ and
1
chemoselectivity (Scheme 4). The H NMR spectra of the crude reaction mixtures show
the presence of oligomeric material but did not reveal the formation of any
cyclization/crossꢀcoupling adducts different to products 9.
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 4. PalladiumꢀCatalyzed Preparation of Dihydropyrans 9. PMP = 4ꢀ
MeOC H
6
4
HO
OAc
NHBoc
5
mol% PdCl2
R
•
•
+
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HN
Boc
–)-1a
R
DMF, RT, 26 h
O
8
a R = Ph
(–)-9a R = Ph (44%)
(
8b R = PMP
(–)-9b R = PMP (47%)
Lanthanide amido complexes have shown remarkable success in catalyzing both
11
12
hydroamination and hydroalkoxylation reactions. The above precedents suggest that
lanthanideꢀbased catalysts may be effective for the cycloisomerization reaction of
allenic amino alcohols. In marked contrast, it was found that the NꢀBoc cleavage of αꢀ
aminoꢀβꢀhydroxyallene 1a occurred using lanthanum bis(trimethylsilyl)amide as
promoter (Scheme 5). Of note, the allene moiety in product 1b remained intact despite
its inherent reactivity. To the best of our knowledge, such controlled reactivity has not
been reported before.
Scheme 5. LanthanumꢀCatalyzed Preparation of αꢀAminoꢀβꢀhydroxyallene 1b
HO
•
5 mol% La[N(SiMe
3
)
2
]
3
HO
•
HN
Boc
–)-1a
Toluene, reflux, 2 h
H N
2
(
(+)-1b (quantitative yield)
To address the selectivity issue on the metalꢀcatalyzed heterocyclization reaction of
allenic amino alcohols, we further investigated additional substrate 1b. We were pleased
to observe that αꢀaminoꢀβꢀhydroxyallene 1b also worked well under the above
optimized catalytic conditions of Table 1 and Schemes 3 and 4. Thus, goldꢀcatalyzed
cycloisomerization and palladiumꢀcatalyzed cyclization/coupling reactions lead to the
controlled formation of heterocycles 5b–7b and 10a,b in slightly higher yields (Scheme
8
ACS Paragon Plus Environment

Page 9 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
6). Dendralenic products 10 are potential precursors for more complex structures
1
3
through the dieneꢀtransmissive Diels–Alder reaction.
Scheme 6. Chemoꢀ and Regiocontrol in the MetalꢀCatalyzed Heterocyclization
Reactions of αꢀAminoꢀβꢀhydroxyallene 1b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HO
5 mol% AuCl3
HO
•
N
H
H N
2
CH Cl , RT, 2 h
2 2
(
+)-1b
(+)-5b (81%)
Br
HO
5 mol% PdCl2
NH2
•
•
HO
+
H N
DMF, RT, 2 h
N
2
O
H
(
+)-1b
(–)-6b (53%)
(+)-7b (39%)
HO
H N
OAc
NH2
5 mol% PdCl2
DMF, RT, 26 h
R
•
+
R
2
O
(
+)-1b
8
a R = Ph
(–)-10a R = Ph (56%)
8b R = PMP
(–)-10b R = PMP (52%)
Next, we focused on finding a heterocyclization procedure for functionalized
allenes 2a,b that could be both regioꢀ and chemoselective. βꢀAminoꢀγꢀhydroxyallenes
2a,b have diverse reactive sites, at which two different transformations (Nꢀcyclization
versus Oꢀcyclization) can take place. Gold(I)ꢀbased complexes were the best catalysts
among several complexes screened (AuCl
of gold(I) catalysts were tested to reveal that [(Ph
choice. The best result was obtained from a reaction of 2a using [(Ph
3
, PtCl
2
, PdCl
2
, PtCl
2
(CH
] was the catalyst of
P)AuNTf ] (5
2 2 2
=CH )] ). A number
3
P)AuNTf
2
3
2
mol %) in dichloromethane, producing tertꢀbutyl 5ꢀ(hydroxymethyl)ꢀ2,3ꢀdimethylꢀ1Hꢀ
pyrroleꢀ1ꢀcarboxylate 11a in 53% isolated yield (Scheme 7). When the present reaction
was conducted with allenic amino alcohol 2b, pyrrole derivative 11b was also
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
exclusively produced (Scheme 7). Unfortunately, extensive decomposition was
observed on sensitive pyrroles 11 during purification by flash chromatography, which
may be responsible for the moderate isolated yields. The goldꢀcatalyzed
heterocyclization reaction of βꢀaminoꢀγꢀhydroxyallenes
2
shows the same
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chemoselectivity but different regioselectivity as that observed under gold catalysis for
αꢀaminoꢀβꢀhydroxyallenes 1, that is, the initial cyclization takes place by selective
nucleophilic addition of the amino group to the activated central allene carbon. By
contrast, the cyclization of βꢀaminoꢀγꢀhydroxyallenes 2a,b was not as rewarding when
they were treated with allyl bromide using a Pd–based catalytic system. Palladium salts,
which were effective for the oxycyclization reaction of allenic amino alcohols 1a,b,
gave rise to complex reaction mixtures. These results clearly show that the metalꢀ
catalyzed aminocyclization is preferred over cycloetherification for βꢀaminoꢀγꢀ
hydroxyallenes 2. Pyrrole formation could be explained through a 5ꢀexo cyclization by
chemoꢀ and regioselective attack of the amino group to the central allene carbon,
followed by elimination of methanol with concomitant aromatization.
Scheme 7. GoldꢀCatalyzed Preparation of Pyrroles 11
R
MeO
R
•
5
mol% [(Ph P)AuNTf ]
3 2
HO
N
CH Cl , RT, 2 h
2
2
HO
NH
Boc
Boc
(
–)-2a
1
1a R = Me (53%)
1b R = Ph (40%)
(–)-2b
1
A possible mechanism for the catalytic achievement of 2,5ꢀdihydroꢀ1Hꢀpyrroles 5
involving a goldꢀbased carbophilic πꢀacid may proceed through initial coordination of
the metal to the distal double bond of αꢀaminoꢀβꢀhydroxyallenes 1 leading to species 1ꢀ
Au. Next, chemoꢀ and regioselective 5ꢀendoꢀtrig allenic aminoauration forms
10
ACS Paragon Plus Environment

Page 11 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
zwitterionic dihydropyrroles 12. Loss of HCl in intermediates 12 generates neutral
alkenylmetal species 13. Protonolysis of the carbon–gold bond of 13 liberates 2,5ꢀ
dihydroꢀ1Hꢀpyrroles 5 with concomitant regeneration of the Au(III) catalytic species
(Scheme 8).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 8. Mechanistic Explanation for the GoldꢀCatalyzed Cycloisomerization
Reaction of αꢀAminoꢀβꢀhydroxyallenes 1
HO
N
HO
•
R
5
HN
R
AuCl3
1
protonolysis
HCl
AuCl2
HO
HN
•
HO
AuCl3
N
R
R
1
-Au
13
AuCl3
–HCl
5-endo-trig
aminoauration
HO
N
R
H
12
A mechanistic hypothesis for the achievement of 3,6ꢀdihydroꢀ2Hꢀpyrans 9 and 10
from αꢀaminoꢀβꢀhydroxyallenes 1 is outlined in Scheme 9. Chemoꢀ and regiocontrolled
palladium(II)ꢀcatalyzed oxypalladacyclization of the aminoꢀβꢀhydroxyallene molecule 1
produces a palladadihydropyran species 14, which then suffers a crossꢀcoupling reaction
with the αꢀallenic ester 8. Alkenyl palladium(II) species 14 reacted with the central
allene carbon of 8 in a regioselective fashion giving rise to intermediates 15.
Subsequently, butaꢀ1,3ꢀdienyl 3,6ꢀdihydroꢀ2Hꢀpyranꢀ3ꢀamines 9 and 10 are formed as
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
single Eꢀisomers through transꢀβꢀdeacetoxypalladation. The metal catalyst is reformed
at the same time. Notably, the heteropalladation of αꢀaminoꢀβꢀhydroxyallenes 1 to form
the corresponding 3,6ꢀdihydropyranyl palladium species 14 is totally chemoselective
toward the allenol functionality. We did not observe the formation of azacycles from the
aminocyclization of the allenamine functionality in 1.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 9. Mechanistic Explanation for the PalladiumꢀCatalyzed
Oxycyclization/CrossꢀCoupling Sequence between αꢀAminoꢀβꢀhydroxyallenes 1
and αꢀAllenic esters
HO
•
AcOH
II
HN
R
Pd Cl
2
1
HCl
ClPd(OAc)
ClPdII
NHR
O
NHR
R
14
Me
O
OAc
H
NHR
ClPdII
9
, 10
•
R
R
OAc
O
Me
8
Me
1
5
Scheme 10 outlines a mechanistic proposal for the achievement of pyrroles 11 from
I
allenic amino alcohols 2. Initial Au ꢀcoordination to the 1,2ꢀdiene moiety gave an
allenegold complex 2ꢀAu, which suffers a chemoꢀ and regioselective 5ꢀexo
aminocyclization reaction to give the intermediate auradihydropyrrolium 16. Loss of
proton generates neutral species 17, which followed by protonolysis of the carbon–gold
bond afforded dihydropyrroles 18 with concurrent regeneration of the gold catalyst.
Heterocycles 18 then suffer a methanol elimination to give the pyrrole products 11.
12
ACS Paragon Plus Environment

Page 13 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Scheme 10. Mechanistic Explanation for the GoldꢀCatalyzed Aminocyclization
reaction of βꢀAminoꢀγꢀhydroxyallenes 2
MeO
HO
–MeOH
R
MeO
R
•
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
1
8
NTf2
HO
NH
Boc
R
Boc
AuPPh3
HO
2
protonolysis
HNTf2
N
Boc
11
R
MeO
R
•
MeO
HO
N
Boc
AuPPh3
AuPPh3
HO
NH
Boc
17
2-Au
MeO
HO
R
–
HNTf2
5-exo-dig
aminoauration
N
AuPPh3
Boc
H
16
In order to get further insights into the chemoꢀ and regioselectivity, as well as the
reaction pathway, a density functional theory (DFT) study has been performed.
As a model reaction, we have selected precursor 1b. Table 2 summarizes the
activation barriers and relative energies for the possible heterocyclization steps for both
the amino and alcohol functionalities, from the allene–metal complexes 1bꢀAu
less stable 1bꢀAu . These complexes are formed by coordination of the distal and
proximal alkene bonds of the allene to the catalyst, respectively.
D
and the
P
ꢀ1
Table 2. Relative freeꢀenergy values [M06/def2ꢀQZVP/PCM, in kcal mol ] for the
possible Auꢀcatalyzed heterocyclization steps of 1b.
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
#
Complex
bꢀAuD
Mode
ꢀG
.7
ꢀG
3
ꢀ13.4
1
aminocyclizations
4ꢀexo-digꢀ
ꢀendo-trigꢀ
5ꢀexo-dig-
ꢀendo-trigꢀ
5
2.6
8.6
6.1
ꢀ17.7
ꢀ12.2
ꢀ2.3
ꢀ15.0
ꢀ9.1
oxycyclizations
aminocyclizations
oxycyclizations
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1bꢀAu
P
3ꢀexo-trig-
4ꢀendo-digꢀ
5ꢀendo-digꢀ
7.8
13.3
10.8
12.8
ꢀ10.5
6.8
4
exo-trig-
According to the calculated values, the lowestꢀenergy transition structure for the
Au(III)ꢀcatalyzed heterocyclization is found for TS15ꢀendoꢀtrig, which belongs to the 5ꢀ
1
,14
endo-trig cyclization.
Moreover, it drives to the most stable Au σꢀalkenyl
zwitterionic intermediate I5ꢀendoꢀtrig. Other plausible transition structures for the
intramolecular nucleophilic attack of the alcohol or amine on the triggered allene, show
#
ꢀ1
higher barriers than TS15ꢀendoꢀtrig (ꢀꢀG = 1.1–10.7 kcal mol ). The formation of the
heterocyclization intermediates is exergonic for all; only one exception is found, as the
strained attack of the βꢀallenol group onto the central allene position, 4ꢀexo-trig
cyclization, is slightly endergonic. In fact, the formation of the most stable Au σꢀalkenyl
–
1
intermediate I5ꢀendoꢀtrig is the most exergonic process (–17.7 kcal mol ), pointing to a
moderately irreversible character of this heterocyclization mode.
These results agree with the experimentally observed selective formation of 3ꢀ
pyrroline 5b and would propose a kinetic and thermodynamic control of the chemoꢀ and
regioselective intramolecular aminocyclization. However, according to previous
15
16
estimation of the reaction mechanism by us and others, the rateꢀdetermining step
would be the 1,3ꢀH migration to form the product which is assisted by the metalꢀ
complex and proceeds stepwise: release of HCl and Au–C bond excision by
1
7
protonolysis with HCl to end the cycle and restore the active catalyst.
14
ACS Paragon Plus Environment

Page 15 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
For the formation of 5b, the loss of HCl from I5ꢀendoꢀtrig produces the neutral
ꢀ1
complex I25NꢀAu in an endergonic process (ꢀG = +16.2 kcal mol ). To end,
protonolysis of the CꢀAu bond yields 5bꢀAu via the transition state TS3 ꢀAu with an
5
N
#
ꢀ1
activation energy of ꢀG = +4.9 kcal mol . Thus, according to the calculated higher
energy barriers, the 1,3ꢀH migration would be the bottleneck of the Auꢀcatalyzed
formation of 5b (Figure 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ1
Figure 1. Calculated freeꢀenergy profile [M06/def2ꢀQZVP/PCM, in kcal mol ] for
some plausible Auꢀcatalyzed reactions of 1b.
For the Pdꢀcatalyzed process, the calculations also suggest that, the 5ꢀendo-trig
heterocyclization by nucleophilic attack of the αꢀamino moiety to the terminal allene
ꢀ1
carbon of the most stable complex, 1bꢀPd
D
(by 0.8 kcal mol ) (via TS15ꢀendoꢀtrig) (Table
3
), is the favored ringꢀclosure step both from the kinetic and thermodynamic
#
perspectives, in view of the higher activation energy barriers (ꢀꢀG = 4.8–18.1 kcal
ꢀ1
ꢀ1
mol ) and the lower exergonicity (ꢀꢀG = 3.7–25.7 kcal mol ) computed for other
plausible heterocyclizations. Notably, the endo-trig cyclization by nucleophilic attack of
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the βꢀalcohol on the terminal allene carbon, through TS16ꢀendoꢀtrig, also shows a low
activation barrier, although 4.8 kcal/mol higher, which unfortunately contrasts with the
experimental observations.
ꢀ1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 3. Relative freeꢀenergy values [M06/def2ꢀQZVP/PCM, in kcal mol ] for the
possible Pdꢀcatalyzed heterocyclization steps of 1b.
#
Complex
bꢀPdD
Mode
ꢀG
ꢀG
9
.7
ꢀ9.2
1
aminocyclizations
4ꢀexo-digꢀ
5
ꢀendo-trigꢀ
2.3
15.8
7.1
11.7
19.9
19.0
20.4
ꢀ12.9
ꢀ8.1
4.1
ꢀ8.1
ꢀ7.7
ꢀ3.0
12.8
oxycyclizations
aminocyclizations
oxycyclizations
5ꢀexo-dig-
6ꢀendo-trigꢀ
3ꢀexo-trig-
4ꢀendo-digꢀ
5ꢀendo-digꢀ
1bꢀPd
P
4
exo-trig-
On other hand, the results depicted in Scheme 6 suggest that the selectivity of the
Pdꢀcatalyzed process depends on the nature of the coupling alkene. Thus, in the
presence of allyl bromide, the kinetically favored formation of the I5ꢀendoꢀtrig and I6ꢀendoꢀ
trig alkenyl intermediates would undergo a crossꢀcoupling reaction to form 7b and 6b,
respectively, while in the presence of αꢀallenic ester, the formation of the oxacycles 10
takes place with complete regioꢀ and chemoselectivity. To account for these
observations and apparent discrepancies, we have performed a deeper analysis.
The crossꢀcoupling with allyl bromide from the alkenyl intermediate I6ꢀendoꢀtrig takes
ꢀ1
place as follows (Figure 2). After loss of HCl through TS2
Pd
probable fast ligandꢀinterchange displacement mechanism to yield the palladium
6
(+6.0 kcal mol above 1bꢀ
D
), the Pdꢀcoordinated HCl is easily moved by the entering allyl bromide in a
ꢀ1
complex I36Br. This step is slightly endergonic (ꢀG = –0.9 kcal mol ) and has a low
ꢀ1
activation barrier (1.2 kcal mol ). This species evolves into complex I46Br via the nearly
16
ACS Paragon Plus Environment

Page 17 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
planar fourꢀmembered transition structure TS46Br, a saddle point related to the insertion
of the allylic halide C=C bond into the Pdꢀalkenyl bond to yield a σꢀC–Pd intermediate
I46Br and the formation of a new carbon–carbon bond. This intermediate is formed in a
ꢀ1
ꢀ1
highly exergonic step (–30.1 kcal mol ) with a low activation barrier (2.8 kcal mol ). A
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
final elimination step yields the complexed dihydropyran 6bꢀPd in a exergonic step (by
ꢀ1
ꢀ1
–8.8 kcal mol ) through TS5
(–18.0 kcal mol ). Overall, the reaction is highly
Br
6
exergonic and the cyclization and coupling steps constitute the bottleneck of the
transformation.
Earlier computational results on palladiumꢀcatalyzed reactions have revealed that,
in presence of an unsaturated reactant, the crossꢀcoupling from the alkenyl intermediate
12d,18
is favored over a proton shift process to yield protonolysis products.
Expectedly,
the competitive protonolysis step from I2 involves a less favored process both from
6
kinetic and thermodynamic viewpoints, since our calculations reveal a higher activation
ꢀ1
D
barrier (transition structure 11.9 kcal mol above 1bꢀPd ) and less exergonicity (–14.4
kcal/mol) than the couplingꢀprocess, in agreement with the experimental absence of the
protonolysis product.
The formation of the 3ꢀpyrroline 7b follows the equivalent pathway after azacyclization
(Figure 2), namely, loss of HCl, ligand displacement from the metal coordination sphere,
insertion of the C=C bond of the allylic halide into Pd–alkyl bond to yield a σꢀC–Pd
intermediate, and finally, a βꢀelimination step. Firstly, the lower activation barrier for
the HCl release is likely due to the lower steric distortion needed to attain the transition
state, whereas the higher stability of I2 is due to the strong Hꢀbond of the formed HCl
5
with the amine group. This complex can coordinate the allyl bromide to produce the
ꢀ1
allylꢀpalladium complex I35Br, 0.5 kcal mol less stable than I36Br, through a transition
ꢀ1
structure TS35Br showing an energy 0.4 kcal mol higher than the dihydropyran
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
counterpart (TS36Br). This intermediate species evolves into complex I45Br by formation
ꢀ1
of the new carbon–carbon bond via the transition state TS45Br, 0.4 kcal mol less stable
than TS46Br. Finally, the elimination step leads to the complexed 3ꢀpyrroline 7bꢀPd in a
ꢀ1
exergonic step (by –11.7 kcal mol ) by overcoming a low activation barrier.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
If we compare the free energy profiles for the formation of 6b and 7b, a similar profile
for the crossꢀcoupling process is found. However, the rateꢀdetermining step for 6b is the
cyclization, although very close to the coupling event, while for 7b the crossꢀcoupling is
ꢀ
proposed as the rate determining step, involving a transition structure only 0.1 kcal mol
1
higher than the heterocyclization for 6b. These results are supported by the
competitive formation of 6b and 7b.
ꢀ1
Figure 2. Calculated freeꢀenergy profile [M06/def2ꢀQZVP/PCM, in kcal mol ] for the
Pdꢀcatalyzed heterocyclization/crossꢀcoupling reaction of 1b with allyl bromide to form
6b (blue) and 7b (black).
18
ACS Paragon Plus Environment

Page 19 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Following the same analysis for the crossꢀcoupling reaction with the αꢀallenic esters
8
to afford the dihydropyrans 10 (R = Ph), the calculations reveal critical differences
between the formation of the dihydropyran and 3ꢀpyrroline frameworks.
Recently, results on the palladiumꢀcatalyzed crossꢀcoupling process between two
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
8
different allenes (allenol and αꢀallenic ester) to form dihydropyrans were reported.
The computational analysis of the reaction mechanisms suggested a kinetic control of
the chemoꢀ and regioselective intramolecular cyclization. Moreover, the calculations
accounted for the absence of the plausible homoꢀcoupling, other crossꢀcoupling modes
or Pd–C protonolysis products.
6
The alkenyl palladium(II) intermediate I2 can coordinate an αꢀallenic ester ligand
to form the alleneꢀPd complex I36Al (Figure 3) through an associative mechanism step
ꢀ1
involving a very low energy barrier (0.9 kcal mol ) via TS3 . This species evolves
6
Al
into complex I46Al via the transition state TS46Al, which shows the incipient formation
of a CꢀC bond with the central carbon of the allene (C–C distance = 2.107 Ǻ). This step
#
ꢀ1
proceeds with a low energy barrier (ꢀG = 2.7 kcal mol ) and is strongly exergonic (ꢀG
ꢀ1
=
–37.7 kcal mol ), thus suggesting an irreversible process. It should be noted that
2
19
TS46Al shows a C(sp )ꢀH…π interaction with the aromatic core (2.504 Ǻ). Other
ꢀ1
conformations for TS46Al and I46Al lacking this interaction are 2.8 and 2.7 kcal mol
higher in energy, respectively (see additional computational results in the Supporting
Information).
Finally, a transꢀβꢀdeacetoxypalladation step drives to the formation of the transꢀ
4
II
butaꢀ1,3ꢀdienylꢀdihydropyran 10aꢀPd, η ꢀcoordinated to Pd . This step takes place
through the transition structure TS56Al, which shows a C–O cleavage (2.146 Ǻ) and the
ꢀ1
initial coordination of the ester to the metal (2.951 Ǻ), with an energy 19.8 kcal mol
ꢀ1
below 1bꢀPd . The overall process is highly exergonic (–30.1 kcal mol ).
D
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A close inspection of the crossꢀcoupling process from de azacyclic intermediate I25
reveals crucial differences. The corresponding computed reaction profiles are depicted
in Fig. 3, which collects the relative free energy values calculated at 298 K (ꢀG298).
ꢀ
Thus, our calculations put forward that the formation of I35Al is kinetic (by 1.9 kcal mol
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
ꢀ1
) and thermodynamically disfavored (by 5.3 kcal mol ) over the formation of the
oxacycle I3 . Then, the coupling step occurs via the transition state TS4 , involving
6
Al
5Al
ꢀ1
the CꢀC bond formation (2.043 Ǻ), with a free energy 5.5 kcal mol higher in energy
ꢀ1
than TS46Al in a strong exergonic transformation (ꢀG298 = –39.0 kcal mol ). It should
3
2
be noted that TS45Al shows a C(sp )ꢀH…π (2.798, 3.211 Ǻ), weaker than a C(sp )ꢀH…π
in TS4 , which could account for, in part, the higher energy. Finally, we locate the
6
Al
transition state TS55Al, a saddle point associated with the transꢀβꢀdeacetoxypalladation
step, which leads to the complexed dihydropyrrole product, 21ꢀPd. Therefore, based on
the computed data, it can be concluded that the formation of 10a seems to be the most
favored process for the palladiumꢀcatalyzed heterocyclization/crossꢀcoupling sequence
between αꢀallenols and αꢀallenic esters, taking place with regioꢀ and chemoselectivity,
which nicely agrees with the experimental findings.
20
ACS Paragon Plus Environment

Page 21 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
Figure 3. Comparison of the freeꢀenergy profiles [M06/def2ꢀQZVP/PCM, in kcal mol
1
]
for the Pdꢀcatalyzed heterocyclization/crossꢀcoupling reaction of 1b with αꢀallenic
acetate 8a to yield 10a (blue) and the undetected dihydropyrrole product 21 (black).
To account for the differences found between the coupling reagents, we have
performed a NBO analysis of the coupling processes of allyl bromide and αꢀallenic
2
C
acetates. In the intermediate complexes I3, the σꢀdonating (from the endocyclic C sp
lone pair partially overlapping the 5s orbital on palladium) and πꢀbackdonating (a
highly polarized dπꢀtoꢀpπ donation from the metal to the empty pπ orbital on C
C
)
interactions between the heterocyclic scaffold and the catalyst generate a short
Pd−endocyclic carbon bond and a high Wiberg bond order (these values exceed the
maximum value of 0.5 for a σ bond, see additional computational results in the
Supporting Information). This effect is also reflected in the analysis of partial NPA
charges, which shows a less NPA negative charge at the PdCl fragment with increasing
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
C
C
ꢀPd bond order. Hence, the charge flow from the π–backbonding interaction of the
metal to the alkene/allene moiety should be lower and the interaction weaker. For the
coupling process with αꢀallenic acetate, a comparison of the NBO analysis for I36Al and
I35Al, reveals a higher NPA charge at the catalyst (–0.10 vs +0.01e), a lower NPA
positive charge at the allene fragment (+0.13 vs +0.03e, respectively), and lower
Wiberg bond orders for the Pd–Cꢀallene interactions for I35Al (0.222 and 0.198 vs 0.186
and 0.155, respectively), which is also confirmed by the lower interaction energy for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ1
I35Al than for I36Al (ꢀE = 7.7 kcal mol ). Additionally, in the coupling transition state,
2
2
TS45Al, the analysis also reveals a weaker interaction energy than for TS46Al (ꢀE = 6.2
ꢀ1
kcal mol ) between the donor π orbital and the empty developing orbital on the
C
endocyclic C atom.
In contrast, a comparison of the NBO analysis for I36Br and I35Br, shows a
similar negative NPA charge at the catalyst (–0.11 vs –0.12e), an analogous NPA
positive charge at the incoming alkene fragment (+0.17 vs +0.19e, respectively), and
equivalent Wiberg bond orders for the Pdꢀalkene interaction (0.246 and 0.218 vs 0.264
and 0.244, respectively). This similarity is reflected by the small difference of
ꢀ1
interaction energy (slightly lower for I35Br, ꢀE = 1.3 kcal mol ). Expectedly, the
2
coupling transition structures also reveal similar values for the interaction energy
between the donor orbital and the empty developing orbital on the endocyclic C atom.
In summary, the results suggest that the interplay of electronic effects at the heterocycle
and coupling reagent controls the competitive formation of dihydropyrans and
dihydropyrroles in the Pdꢀcatalized heterocyclization/crossꢀcoupling reactions of allenic
amino alcohols.
CONCLUSIONS
22
ACS Paragon Plus Environment

Page 23 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
In conclusion, the controlled preparation of 2,5ꢀdihydroꢀ1Hꢀpyrroles, 3,6ꢀdihydroꢀ2Hꢀ
pyrans, and pyrroles has been achieved through swithchable chemoꢀ and regioselectivity
in the metalꢀcatalyzed heterocyclization reactions of allenic amino alcohols. Chemoꢀ
and regioselectivity depend on both the linker elongation as well as the type of catalyst.
Furthermore, in order to justify this different demeanor, Density Functional Theory
calculations have been implemented.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
1
13
General Methods. H NMR and C NMR spectra were recorded on 300 MHz
spectrometers. NMR spectra were recorded in CDCl solutions, except otherwise stated.
3
1
13
Chemical shifts are given in ppm relative to TMS ( H, 0.0 ppm) or CDCl ( C, 77.0
3
ppm). Low and high resolution mass spectra were performed on a QTOF LC–MS
spectrometer using the electrospray mode (ES) unless otherwise stated. Specific rotation
–1
2
–1
[
D
α] is given in 10 deg cm g at 20 °C, and the concentration (c) is expressed in
grams per 100 mL. All commercially available compounds were used without further
purification. Flash chromatography was performed by using silica gel 60 (230−400
mesh) or neutral alumina. Products were identified by TLC (silica gel). UV light (λ =
254nm) and a solution of phosphomolybdic acid in EtOH (1 g of phosphomolybdic acid
hydrate, 100 mL EtOH) was used to develop the plates.
Computational Details. The calculations were performed with the Gaussian 09
2
0
suite of programs. The structures were optimized with Truhlar´s dispersionꢀcorrected
meta hybrid exchangeꢀcorrelation functional M06. This hybrid functional has been
21,22
recommend for structures involving transition metals.
The quadrupleꢀζ quality plus
23
polarization def2ꢀQZVP basis set of Ahlrichs and coworkers have been applied for all
24
atoms except Au and Pd, which were treated with the SDD ECPs. The stationary
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
points were characterized by means of harmonic analysis. For all the transition
structures, the normal mode related to the imaginary frequency corresponds to the
nuclear motion along the reaction coordinates under study. Additionally, in some cases,
we carried out intrinsic reaction coordinate calculations (IRC) to verify that the
transition structures connect with reactants and products. The Polarizable Continuum
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
25
Model (PCM) was applied to consider the solvents effects. Finally, to estimate the
changes in the Gibbs energies in the presence of CH Cl or DMF as solvent, singleꢀ
2
2
point calculations on the optimized geometries were carried out. The natural bond
2
6
orbital analysis (NBO) were performed on the optimized structures.
General Procedure for the Deprotection of Allenyloxazolidines 3, 4a, and 4b.
Preparation of
allenyloxazolidine 3, 4a, or 4b (1.00 mmol) was dissolved in acetonitrile/water (10 mL,
5:5) and the solution was cooled to 0 °C. BiCl (0.05 mmol) was added and the
mixture was stirred at RT until disappearance of the starting material (TLC, typically
14–24 h). The reaction mixture was poured into a satd. aq. solution of NaHCO . The
NꢀBoc Allenic Amino Alcohols 1a, 2a, and 2b. The corresponding
9
3
3
aqueous layer was extracted with ethyl acetate (3 x 20 mL), and the combined organic
extracts were dried (MgSO ) and concentrated in vacuo to afford allenic amino alcohols
4
1a, 2a, and 2b. Further purification was not necessary. Spectroscopic and analytical
data for compounds 1a, 2a, and 2b follow.
NꢀBoc Allenic Amino Alcohol (–)ꢀ1a. From 100 mg (0.42 mmol) of
allenyloxazolidine (–)ꢀ3, compound (–)ꢀ1a (87 mg, quantitative yield) was obtained as a
1
o
colorless oil; [α]
D
= –16.2 (c 1.2, CHCl
3
); HꢀNMR (300 MHz, CDCl
3
, 25 C) δ: 1.43 (s,
9H), 2.86 (br s, 1H), 3.62 (dd, 1H, J = 11.1, 5.5 Hz), 3.70 (dd, 1H, J = 11.1, 4.1 Hz),
4.23 (br s, 1H), 4.90 (dd, 2H, J = 6.7, 3.4 Hz), 4.99 (d, J = 7.5 Hz), 5.23 (dd, 1H, J =
24
ACS Paragon Plus Environment

Page 25 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
1
3
o
12.0, 6.5 Hz); CꢀNMR (75 MHz, CDCl
3
, 25 C) δ: 207.2, 156.0, 89.8, 79.8, 78.5, 65.4,
–1
5
0.7, 28.3; IR (CHCl
3
, cm ): ν 3401, 3343, 1959, 1692, 1170; HRMS (ES): calcd for
+
C
10
H
17NO
3
[M] : 199.1208; found: 199.1212.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
ꢀBoc Allenic Amino Alcohol (–)ꢀ2a. From 100 mg (0.34 mmol) of
allenyloxazolidine (–)ꢀ4a, compound (–)ꢀ2a (86 mg, quantitative yield) was obtained as
1
o
a colorless oil; [α]
D
= –85.5 (c 0.5, CH
2
Cl
2
); HꢀNMR (300 MHz, C
6
D
6
, 25 C) δ: 1.43
(
s, 9H), 1.66 (t, J = 3.1 Hz, 3H), 2.98 (s, 3H), 3.60 (dd, J = 11.4, 4.0 Hz, 1H), 3.73 (d, J
1
3
= 6.5 Hz, 1H), 3.88 (m, 2H), 4.56 (d, J = 1.8 Hz, 2H), 5.02 (d, J = 8.5 Hz, 1H); Cꢀ
o
NMR (75 MHz, C
6
D
6
, 25 C) δ: 207.8, 155.9, 96.5, 83.9, 79.1, 74.8, 62.5, 56.8, 53.3,
–1
2
8.5 (3C), 13.5; IR (CHCl
3
, cm ): ν 3447, 1958, 1695, 1172; HRMS (ES): calcd for
+
C
13 4
H23NO [M] : 257.1627; found: 257.1628.
NꢀBoc Allenic Amino Alcohol (–)ꢀ2b. From 100 mg (0.28 mmol) of
allenyloxazolidine (–)ꢀ4b, compound (–)ꢀ2b (89 mg, quantitative yield) was obtained as
1
o
a colorless oil; [α]
D
= –23.5 (c 0.8, CH
2
Cl
2
); HꢀNMR (300 MHz, C
6
D
6
, 25 C) δ: 1.44
(
s, 9H), 2.98 (s, 3H), 3.49 (dd, J = 11.3, 2.7 Hz, 1H), 3.99 (m, 2H), 4.67 (br s, 1H), 4.88
(m, 2H), 5.53 (d, J = 7.6 Hz, 1H), 7.02 (t, J = 7.3 Hz, 1H), 7.21 (t, J = 7.7 Hz, 2H), 7.72
13
o
(
d, J = 7.6 Hz, 2H); CꢀNMR (75 MHz, C C
6
D
6
, 25 C) δ: 208.8, 156.0, 134.4, 129.0
(
2C), 127.5, 127.1 (2C), 104.6, 82.4, 80.0, 79.1, 61.8, 57.6, 54.0, 28.5 (3C); IR (CHCl3,
–1
+
4
cm ): ν 3437, 1940, 1703, 1174; HRMS (ES): calcd for C18H25NO [M] : 319.1784;
found: 319.1773.
III
Procedure for the La ꢀCatalyzed Cleavage of NꢀBoc Allene 1a. Preparation
of Allenic Amino Alcohol 1b. La[N(SiMe ) ] (0.05 mmol) was added to a stirred
3
2 3
solution of the NꢀBoc allenic amino alcohol 1a (200 mg, 1.0 mmol) in toluene (10.0
mL) under argon. The resulting mixture was stirred at reflux temperature until
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
disappearance of the starting material (TLC, 2 h). The reaction was then filtered through
a celite plug before being concentrated under reduced pressure. Chromatography of the
residue eluting with ethyl acetate/hexanes (1:2) gave analytically pure adduct 1b.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Allenic Amino Alcohol (+)ꢀ1b. From 200 mg (1.0 mmol) of NꢀBoc allenic
amino alcohol (–)ꢀ1a, compound (+)ꢀ1b (99 mg, quantitative yield) was obtained as a
1
o
colorless oil; [α]
D
= +9.4 (c 0.5, CHCl
3
); HꢀNMR (300 MHz, CDCl
3
, 25 C) δ: 4.19
(dd, 1H, J = 8.5, 5.6 Hz), 4.42 (dd, 1H, J = 13.7, 6.9 Hz), 4.57 (t, 1H, J = 8.3 Hz), 4.97
1
3
(
dt, 2H, J = 6.4, 1.7 Hz), 5.22 (dd, 1H, J = 13.3, 6.7 Hz), 5.32 (br s, 1H); CꢀNMR (75
o
–1
MHz, CDCl
3
, 25 C) δ: 208.1, 90.6, 78.8, 70.1, 51.6; IR (CHCl
3
, cm ): ν 3279, 1956,
+
1
228; HRMS (ES): calcd for C H NO [M] : 99.0684; found: 99.0688.
5
9
General Procedure for the IndiumꢀPromoted Reaction between 3ꢀ
Substituted Propꢀ2ꢀynyl Bromides and Garner’s Aldehyde. Preparation of αꢀ
Allenic Alcohols 4aꢀOH and 4bꢀOH. 1ꢀBromoꢀ2ꢀbutyne or 1ꢀbromoꢀ3ꢀphenylꢀ2ꢀ
propyne (3.0 mmol) was added to a well stirred suspension of Garner’s aldehyde (1.0
o
mmol) and indium powder (6.0 mmol) in THF/NH Cl (aq. sat.) (1:5, 5 mL) at 0 C.
4
After disappearance of the starting material (TLC) the mixture was extracted with ethyl
acetate (3 x 5 mL). The organic extract was washed with brine, dried (MgSO ) and
4
concentrated under reduced pressure. Flash chromatography of the residue on
deactivated silica gel using ethyl acetate/hexanes mixtures gave analytically pure
compounds.
αꢀAllenol (–)ꢀ4aꢀOH. From 680 mg (2.96 mmol) of Garner’s aldehyde, and after
chromatography of the residue using hexanes/ethyl acetate (4:1) as eluent gave
compound (–)ꢀ4aꢀOH (374 mg, 44%) as a colorless oil; [α] = –41.3 (c 1.2, CH Cl );
D
2
2
1
o
HꢀNMR (300 MHz, C D , 25 C) δ: 4.63 (m, 2H), 3.70 (m, 4H), 3.36 (t, J = 9.8 Hz,
6
6
13
o
1
H), 1.85 (t, J = 3.1 Hz, 3H), 1.36 (m, 15H); CꢀNMR (75 MHz, C D , 25 C) δ: 207.9,
6
6
155.2, 98.8, 97.4, 79.0, 75.0, 74.8, 63.2, 46.8, 28.8, 28.5 (3C), 19.9, 13.7; IR (CHCl3,
26
ACS Paragon Plus Environment

Page 27 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
–
1
+
cm ): ν 3478, 1957, 1738; HRMS (ES): calcd for C15
4
H25NO [M] : 283.1784; found:
2
83.1771.
αꢀAllenol (–)ꢀ4bꢀOH. From 500 mg (2.18 mmol) of Garner’s aldehyde, and
after chromatography of the residue using hexanes/ethyl acetate (5:1) as eluent gave
compound (–)ꢀ4bꢀOH (411 mg, 54%) as a colorless oil; [α] = –39.1 (c 1.0, CH Cl );
HꢀNMR (300 MHz, C D , 25 C) δ: 7.35 (m, 2H), 7.17 (m, 3H), 5.16 (m, 2H), 3.99 (m,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
D
2
2
1
o
6
6
13
o
4H), 1.43 (m, 15H); CꢀNMR (75 MHz, C
6
D
6
, 25 C) δ: 208.1, 154.9, 132.0, 128.7
(
2C), 128.2, 127.2 (2C), 109.1, 99.4 (2C), 80.0, 72.0, 63.3, 48.0, 28.5, 28.4 (3C), 20.2;
–1
+
IR (CHCl
3
, cm ): ν 3476, 1971, 1735; HRMS (ES): calcd for C20
H27NO
4
[M] :
3
45.1940; found: 345.1929.
General Procedure for the Synthesis of Methoxy Allenes 4. Tetrabutyl
ammonium iodide (cat), 50% aqueous sodium hydroxide (18 mL) and dimethyl sulfate
0.60 mmol) were sequentially added at room temperature to a solution of the
(
corresponding αꢀallenol (0.92 mmol) in dichloromethane (18 mL). The reaction was
stirred until disappearance of the starting material (TLC). Then aqueous ammonia
(
30%) was added (2.5 mL), before being partitioned between dichloromethane and
water. The aqueous phase was extracted with dichloromethane (3 x 15 mL), the
combined organic extracts were dried (MgSO ) and concentrated under reduced
4
pressure. Flash chromatography of the residue on deactivated silica gel using ethyl
acetate/hexanes mixtures gave analytically pure compounds.
Methoxy Allene (–)ꢀ4a. From 175 mg (0.62 mmol) of αꢀallenol (–)ꢀ4aꢀOH, and
after chromatography of the residue using hexanes/ethyl acetate (6:1) as eluent gave
1
compound (–)ꢀ4a (124 mg, 67%) as a colorless oil; [α]
D
= –38.9 (c 0.9, CH
2
Cl
2
); Hꢀ
o
NMR (300 MHz, C
6
D
6
, 25 C) δ: 4.54 (m, 2H), 3.79 (m, 4H), 3.21 (s, 3H), 1.49 (m,
13
o
1
8H); CꢀNMR (75 MHz, C
6
D
6
, 25 C) δ: 207.9, 155.1, 98.8, 97.4, 79.0, 74.9, 75.0,
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 36
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
–
1
6
3.2, 56.8, 46.8, 28.8, 28.4 (3C), 19.8, 13.7; IR (CHCl , cm ): ν 1956, 1737; HRMS
3
+
(ES): calcd for C16
H27NO
4
[M] : 297.1940; found: 297.1921.
Methoxy Allene (–)ꢀ4b. From 170 mg (0.49 mmol) of αꢀallenol (–)ꢀ4bꢀOH, and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
after chromatography of the residue using hexanes/ethyl acetate (5:1) as eluent gave
1
compound (–)ꢀ4b (107 mg, 61%) as a colorless oil; [α]
D
= –37.1 (c 1.5, CH
2
Cl
2
); Hꢀ
o
NMR (300 MHz, C
6
D
6
, 25 C) δ: 7.43 (m, 2H), 7.22 (m, 3H), 5.14 (m, 2H), 4.04 (m,
13
o
4
1
4
H), 3.30 (d, J = 9.3 Hz, 3H), 1.37 (m, 15H); CꢀNMR (75 MHz, C
6
D
6
, 25 C) δ: 208.0,
54.9, 131.9, 128.6 (2C), 128.2, 127.2 (2C), 109.0, 99.3 (2C), 79.9, 71.9, 63.2, 57.3,
–
1
8.0, 28.4, 28.3 (3C), 20.2; IR (CHCl , cm ): ν 1955, 1739; HRMS (ES): calcd for
3
+
C
21
H
29NO
4
[M] : 359.2097; found: 359.2077.
III
General Procedure for the Au ꢀCatalyzed Cyclization of αꢀAminoꢀβꢀ
hydroxyallenes 1. Preparation of 2,5ꢀDihydroꢀ1 ꢀpyrroles 5. AuCl (0.025 mmol)
H
3
was added to a stirred solution of the appropriate allenic amino alcohol 1 (0.50 mmol)
in dichloromethane (5 mL) under argon. The resulting mixture was stirred at RT until
disappearance of the starting material (TLC, typically 3 hours). The reaction was then
quenched with brine (0.5 mL), the mixture was extracted with ethyl acetate (3 x 5 mL),
and the combined extracts were washed twice with brine. The reaction was concentrated
under reduced pressure. Chromatography of the residue eluting with hexanes/ethyl
acetate mixtures gave adducts 5. Spectroscopic and analytical data for pure forms of 5
follow.
2,5ꢀDihydroꢀ1
amino alcohol (–)ꢀ1a, and after chromatography of the residue using hexanes/ethyl
acetate (4:1) as eluent gave compound (–)ꢀ5a (34 mg, 77%) as a colorless oil; [α] = –
, 25 C) δ: 1.50 (s, 9H), 3.11 (br s, 1H),
Hꢀpyrrole (–)ꢀ5a. From 44 mg (0.221 mmol) of NꢀBoc allenic
D
1
o
3 3
58.1 (c 0.7, CHCl ); HꢀNMR (300 MHz, CDCl
3.58 (dd, 1H, J = 11.2, 7.0 Hz), 3.79 (dd, 1H, J = 11.4, 2.3 Hz), 4.09 (ddt, 1H, J = 15.6,
28
ACS Paragon Plus Environment

Page 29 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
5.4, 2.2 Hz), 4.21 (d, 1H, J = 15.4 Hz), 4.71 (br s, 1H), 5.63 (d, 1H, J = 4.4 Hz), 5.83 (s,
13
o
1
H); CꢀNMR (75 MHz, CDCl
3
, 25 C) δ: 156.6, 126.7, 126.6, 80.6, 67.6, 67.3, 54.2,
–1
+
2
8.4 (3C); IR (CHCl , cm ): ν 3247, 1748, 1207; HRMS (ES): calcd C H NO [M] :
3
10 17
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
199.1208; found: 199.1205.
2,5ꢀDihydroꢀ1
hydroxyallene (+)ꢀ1b, and after chromatography of the residue using hexanes/ethyl
acetate (2:1) as eluent gave compound (+)ꢀ5b (25 mg, 81%) as a colorless oil; [α]
Hꢀpyrrole (+)ꢀ5b. From 31 mg (0.31 mmol) of αꢀaminoꢀβꢀ
D
=
1
o
3 3
+5.4 (c 0.5, CHCl ); HꢀNMR (300 MHz, CDCl , 25 C) δ: 3.83 (m, 1H), 4.25 (dd, 1H,
J = 8.5, 5.1 Hz), 4.40 (ddt, 1H, J = 15.6, 3.0, 2.1 Hz), 4.61 (t, 1H, J = 8.6 Hz), 4.74 (m,
13
1
H), 5.91 (dt, 1H, J = 5.9, 2.9 Hz), 6.06 (ddd, 1H, J = 5.9, 3.7, 2.1 Hz); CꢀNMR (75
o
–1
MHz, CDCl
3
, 25 C) δ: 130.9, 128.9, 68.7, 64.6, 54.8; IR (CHCl
3
, cm ): ν 3325, 1247;
+
HRMS (ES): calcd C
5
H
9
NO [M] : 99.0684; found: 99.0687.
II
General Procedure for the Pd ꢀCatalyzed Cyclization of αꢀAminoꢀβꢀ
hydroxyallenes 1 in Presence of Allyl bromide. Preparation of Dihydropyrans 6
and Dihydropyrroles 7. Palladium(II) chloride (0.005 mmol) was added to a stirred
solution of the corresponding allenic amino alcohol 1 (0.10 mmol) and allyl bromide
(0.25 mmol) in N,Nꢀdimethylformamide (1.0 mL). The reaction was stirred under argon
atmosphere until disappearance of the starting material (TLC, 2 h). Water (0.5 mL) was
added before being extracted with ethyl acetate (3 x 4 mL). The organic phase was
4
washed with water (2 x 2 mL), dried (MgSO ) and concentrated under reduced pressure.
Chromatography of the residue eluting with hexanes/ethyl acetate mixtures gave
analytically pure adducts 6 and 7.
Preparation of Dihydropyran (–)ꢀ6a and Dihydropyrrole (+)ꢀ7a. From 40 mg
(0.20 mmol) of αꢀaminoꢀβꢀhydroxyallene (–)ꢀ1a, and after chromatography of the
29
ACS Paragon Plus Environment