Oxidation of aromatic alkynes with nitrate radicals (NO3 ?): An experimental and computational study on a synthetically highly versatile radical

Article abstract of DOI:10.1071/CH07045

Addition of electro- and photochemically generated nitrate radicals, NO₃^?, to the C?C triple bond of aromatic alkynes 9a?9h leads to formation of 1,2-diketones 10a?10h. Surprisingly, benzophenones 11a?11h are obtained as by-products, which formally result from loss of a carbon atom. Density functional studies performed with the BHandHLYP method in combination with various basis sets revealed that 1,2-diketones result from 5-endo cyclization of the initially formed vinyl radical and loss of NO^?. The key step to benzophenone formation is a ?-cleavage at the stage of the vinyl radical with release of NO₂^?, followed by Wolff rearrangement of the resulting ?-oxo carbene. CSIRO 2007.

Full text of DOI:10.1071/CH07045

Corrigendum
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2007, 60, 547
•
Oxidation of Aromatic Alkynes with Nitrate Radicals (NO ):
3
An Experimental and Computational Study on a Synthetically
Highly Versatile Radical
Uta Wille and Jilliarne Andropof
(
Vol. 60, No. 6, pp. 420–428)
The published paper contained an error on p. 424, Table 2. The
correct values in column 18 are:
−1
•
3
Table 2. Relative energies ꢀE [kJ mol ] for stationary points in the NO -induced oxidation of alkynes 9 to 1,2-diketones 10 and ketones 11
Data in italics include the zero-point vibrational energy correction (ZPE)
12
13
14
15
10
16
17
18
19
20
11
(
a) Aliphatic alkynes: reaction of 9i with NO•
3
BHandHLYP/6-311G**
BHandHLYP/cc-pVDZ
BHandHLYP/cc-pVTZ
BHandHLYP/aug-cc-pVTZ
0.0
.0
0.0
.0
0.0
.0
0.0
.0
b) Allylic alkynes: reaction of 9j with NO
72.4
68.0
71.4
67.0
71.8
67.5
69.8
65.6
−121.5
−118.2
−123.3
−120.4
−125.4
−121.7
−127.6
−123.6
−114.2
−118.0
−117.0
−120.9
−115.8
−119.6
−117.5
−121.1
−419.0
−430.0
−410.5
−421.2
−407.8
−418.6
−408.8
−419.5
60.9
50.3
61.2
50.4
65.7
54.9
64.7
53.9
47.0
26.6
51.5
31.2
46.2
26.0
42.1
22.0
74.3
57.4
83.4
66.2
64.6
48.1
56.6
40.4
58.0
38.2
64.4
44.6
57.6
37.6
54.1
34.3
−208.1
−217.6
−199.0
−208.8
−205.2
−214.8
−206.5
−215.9
−409.1
−433.3
−392.2
−416.3
−399.4
−423.2
−400.5
−424.4
0
0
0
0
•
3
(
BHandHLYP/6-311G**
0.0
.0
0.0
.0
97.5
95.2
95.0
92.7
•
−102.3
−95.1
−107.0
−99.8
−96.5
−96.1
−368.7
−373.2
−362.9
−366.7
61.3
50.6
60.9
50.2
56.3
44.9
59.6
48.2
90.3
77.6
97.1
84.5
87.0
72.0
93.1
77.9
−181.9
−187.0
−175.2
−180.3
−362.4
−379.4
−347.1
−364.2
0
BHandHLYP/cc-pVDZ
−102.1
−101.5
0
(
c) Aromatic alkynes: reaction of 9a with NO3
BHandHLYP/6-311G**
0.0
.0
0.0
.0
82.6
80.3
81.1
78.4
−87.7
−82.3
−90.4
−85.8
−81.0
−82.6
−84.6
−86.7
−388.2
−393.6
−381.5
−387.1
61.1
51.4
60.6
50.8
35.9
24.4
39.7
27.9
72.5
58.8
79.8
65.7
69.6
54.1
77.0
60.9
−187.3
−194.9
−179.0
−187.3
−381.4
−399.2
−365.8
−384.1
0
BHandHLYP/cc-pVDZ
0
Reaction of 9d with NO•
3
BHandHLYP/6-311G**
0.0
.0
0.0
.0
83.6
81.2
82.0
79.2
60.0
50.3
59.8
49.9
34.1
22.6
37.6
25.8
69.8
54.3
77.0
60.7
0
BHandHLYP/cc-pVDZ
0
Reaction of 9k with NO•
3
BHandHLYP/6-311G**
0.0
.0
0.0
.0
79.9
77.1
79.1
75.8
61.5
51.8
61.5
51.6
16.7
5.9
20.8
9.6
69.8
54.3
66.5
50.3
0
BHandHLYP/cc-pVDZ
0
©
CSIRO 2007
10.1071/CH07045
0004-9425/07/070547

Full Paper
CSIRO PUBLISHING
Aust. J. Chem. 2007, 60, 420–428
www.publish.csiro.au/journals/ajc
•
3
Oxidation of Aromatic Alkynes with Nitrate Radicals (NO ):
An Experimental and Computational Study on a Synthetically
Highly Versatile Radical
Uta WilleA,B,C and Jilliarne AndropofA,B
A
School of Chemistry, University of Melbourne, Melbourne VIC 3010, Australia.
Bio21 Molecular Science and Biotechnology Institute, 30 Flemington Road,
B
Melbourne VIC 3010, Australia.
Corresponding author. Email: uwille@unimelb.edu.au
C
•
3
Addition of electro- and photochemically generated nitrate radicals, NO , to the C≡C triple bond of aromatic alkynes
9
a–9h leads to formation of 1,2-diketones 10a–10h. Surprisingly, benzophenones 11a–11h are obtained as by-products,
which formally result from loss of a carbon atom. Density functional studies performed with the BHandHLYP method
in combination with various basis sets revealed that 1,2-diketones result from 5-endo cyclization of the initially formed
•
vinyl radical and loss of NO . The key step to benzophenone formation is a γ-cleavage at the stage of the vinyl radical
•
2
with release of NO , followed by Wolff rearrangement of the resulting α-oxo carbene.
Manuscript received: 14 February 2007.
Final version: 10 May 2007.
Introduction
diastereoselective 5-exo radical cyclization 4→5. The reaction
•
2
[1,2]
•
is terminated by release of NO , in 5.
Compared with NO ,
3
Self-terminating radical oxygenations are a concept in radi-
cal chemistry recently discovered by us: C≡C triple bonds in
cyclic and open-chain alkynes are transformed into ketones in
a radical addition/cyclization sequence through reaction with
O-centred radicals, for example the nitrate radical NO . The
example given in Scheme 1 shows that reaction of the cycloalkyl
•
2
NO is far less reactive and does not act as radical chain carrier
in this transformation (hence ‘self-terminating’).
•
3
To our surprise, when NO was reacted with the aromatic
•
3
alkyne 6, the expected cyclization to benzofuran 7 did not occur;
instead, the 1,2-diketone 8 was formed as the major product
[3]
•
3
(Scheme 2).
tethered alkyne 1 with NO yields the anellated tetrahydrofuran
Intrigued by this finding, we asked ourselves: (a) Can NO•
2
in a cascade consisting of radical addition to the C≡C triple
3
generally be used as a reagent for the oxidation of aromatic
alkynes to 1,2-diketones, thereby providing a convenient and
non-toxic alternative to the existing methods that often require
bond, followed by 1,5-hydrogen transfer (1,5-HAT), 3→4, and a
O
R
[4]
R
heavy metal compounds, and (b) why does the aliphatic alkyne
H
•
3
1
react differently with NO than the aromatic alkyne 6?
•
NO3
Rꢀ
•
ꢁ
NO2
O
O
Rꢀ
H
Bun
1
2
•
2
O N
•
ꢁNO
NO3
2
O NO
2
ONO2
R
• R
•
O
Et
NO3
H
6
•
•
Rꢀ
NO3
O
O
Rꢀ
3
H
O
Bun
O
5
O
Bun
ONO2
R
O N
2
1
,5-HAT
O2N
5
-exo
Et
•
O
O
Et
O
4
Rꢀ
8
7
Scheme 1. Proposed mechanism for a NO• induced self-terminating
radical oxygenation.
Scheme 2. Formation of 1,2-diketone 8 in the reaction of the aromatic
3
•
alkyne 6 with NO .
3
©
CSIRO 2007
10.1071/CH07045
0004-9425/07/060420

Oxidation of Aromatic Alkynes
421
•
To explore the generality of NO in transforming aromatic
of the respective chain-shortened ketones 11a–11h, that is, reac-
tion conditions and electron density at the C≡C triple bond. In
the second part, density functional calculations (DFT) on the
reaction of aromatic (9a, 9d, 9k) and aliphatic alkynes (9i, 9j)
3
alkynes into 1,2-diketones, we treated diphenylacetylene 9a with
NO . Indeed, formation of the corresponding 1,2-diketone 10a
occurred, but, interestingly, significantamountsofthebenzophe-
none 11a, which arises through formal loss of one carbon atom,
were also found (Scheme 3).
•
3
•
with NO were undertaken to elucidate the mechanism of this
3
oxidation and to reveal the reason for the different behaviour of
•
3
This accumulation of unexpected outcomes showed that reac-
tions of NO with alkynes are not fully understood. In this paper
alkynes 1 and 6 towards NO , as illustrated in Schemes 1 and 2.
•
3
we therefore wish to present the results of our endeavours to shed
Results and Discussion
•
light onto the various pathways that NO can undergo in its reac-
3
Experimental Studies
tion with C≡C triple bonds. In the first part, experimental studies
•
3
•
3
NO was generated in preparative amounts in the presence of
were performed to reveal the potential of NO in transforming
the respective alkynes 9a–9h by anodic oxidation of lithium
nitrate in a non-degassed solvent mixture of acetonitrile, water,
and an ether, typically methyl tert-butyl ether (MTBE), Eqn 1,
and by photo-induced electron transfer from ceric(iv) ammo-
the various different aromatic alkynes 9a–9h into 1,2-diketones
1
0a–10h and to explore the factors that may influence formation
O
•
O
NO3
R2
nium nitrate (CAN) in acetonitrile under argon at λ = 350 nm,
R1
R2
ꢂ
[5]
R1
Eqn 2.
R1
11a–11k
g: R1 ꢃ R2 ꢃ 4-MeC6H4
R2
O
− anode
•
−
NO → NO + e
(1)
(2)
9
a–9k
10a–10k
3
3
hν
•
3
a: R1 ꢃ R2 ꢃ C6H5
(NH ) Ce(NO ) → NO + (NH ) Ce(NO )
4 2
3 6
4 2
3 5
1
2
1
2
b: R ꢃ C6H5, R ꢃ 4-ClC6H4
h: R ꢃ 4-MeC6H4, R ꢃ 4-MeOC6H4
(CAN)
1
2
1
2
c: R ꢃ C6H5, R ꢃ 4-AcC6H4 i: R ꢃ R ꢃ Me
1
2
1
2
d: R ꢃ R ꢃ 4-ClC6H4
j: R ꢃ R ꢃ Allyl
1
2
1
2
The results of the most significant experiments are listed in
Table 1. The products 10a/11a were identified by NMR and MS,
and compared with authentic samples. 10b–10h and 11b–11h
were identified by GC/MS and comparison of the MS data with
e: R ꢃ C6H5, R ꢃ 4-MeC6H4 k: R ꢃ R ꢃ 4-HOC6H4
f: R ꢃ C6H5, R2 ꢃ 4-MeOC6H4
1
Scheme 3.
Table 1. Experimental conditions and results of the reaction of alkynes 9a–9h with NO• and electron
3
transfer reagents
•
3
Yield [%]A
Entry
Alkyne 9
NO precursor or electron transfer reagent
10
11
B
(
(
a) Photochemically generated NO•
3
1
2
3
a: 6 mM
a: 4 mM
a: 6 mM
CAN: 16 mM
CAN: 4 mM
CAN: 19 mM
23
9C
7
39
17C
30
D
•
E
b) Electrochemically generated NO
3
F
9C
46
8
52
47
35
32
16
6
<1C
13
11
7
4
5
6
7
8
9
1
1
1
a: 40 mM
a: 40 mM
b: 22 mM
c: 22 mM
d: 22 mM
e: 22 mM
f: 22 mM
g: 22 mM
h: 22 mM
LiNO3: 100 mM
LiNO3: 100 mM
LiNO3: 56 mM
LiNO3: 56 mM
LiNO3: 56 mM
LiNO3: 56 mM
LiNO3: 56 mM
LiNO3: 56 mM
LiNO3: 56 mM
7
12
27
10
0
1
2
<1
(
c) Control experiments with electron transfer reagents
G
+
−
C
C
C
13
14
15
16
17
a: 21 mM
a: 17 mM
a: 170 mM
a: 230 mM
a: 25 mM
(BrC6H4)3N SbCl : 19.2 mM
K2S2O8: 43 mM Fe EDTA(SO4)2: 58 mm
Tl(NO3)3: 378 mM
CAN: 500 mM
Anode
<3
<3
<3
–
6
H
II
–
–
I
B,J
E,K
17C
16
C
3
–
A
B
C
D
E
F
Determined by quantitative GC (see Experimental).
In CH3CN, unless otherwise stated; reaction time 20–80 min (see Experimental).
Incomplete conversion of 9.
In CH3CN/H2O (45/1).
In CH3CN/H2O/MTBE (7/1/2), unless otherwise stated; reaction time 2 h.
In CH3CN/MeOH (9/1).
In CH3CN/H2O (4/1); reaction time 1 day.
In CH3CN/H2O (3/4); reaction time 1 day.
G
H
I
In CH3CN/H2O (1/1); reaction time 1 day.
Exclusion of light; reaction time 4 days.
J
K
Electrolysis in CH3CN/H2O/Et2O (5/1/2), with LiCl (45 mM) as supporting electrolyte; reaction time 5.5 h.

4
22
U. Wille and J. Andropof
literature.[6] The yields were determined by quantitative GC (for
details see Experimental).
9
•
ꢂNO
3
Diphenylacetylene 9a was used to optimize the reaction con-
•
•
O
N
O
ditions. Table 1 reveals that reactions with photogenerated NO•
(
O
N
3
O
O
N
entries 1–3) always resulted in a lower yield of the diketone 10a,
5-endo
γ-fragmentation
O
O
O
O
compared to the benzophenone 11b, than those with electrogen-
•
•
3
erated NO (entry 5). A possible explanation for this finding
R1,2
R1,2
R1,2
R1,2
R1,2
R1,2
is the fact that the deep yellow 10a absorbs light at the same
wavelength as CAN and may therefore undergo photochemi-
cal decomposition (this pathway was not explored further), thus
13
12
16
•
NO2
ꢁ
•
3
rendering the photochemical NO source incompatible with this
•
O
N
O
reaction. Irrespective of this, comparison of entry 1 with entry 2
•
O
•
R1,2
?
•
3
shows that an excess of NO is essential for complete conversion
O
O
R1,2
R1
R2
of 9a. Also, in order to exclude formation of ionic intermedi-
ates, H2O was added, which could either act as nucleophile by
trapping cations, or as proton source by reacting with anions.
However, these experiments only resulted in a lower combined
yield of 10a/11a (entry 3), but not in formation of new products,
which confirms the non-ionic nature of this transformation.
R1
R2
17
18
1
4
R1,2
O
•
O
N
R1,2
Under electrochemical conditions, the NO• precursor,
O
O
•
−NO
3
19
LiNO3, acts also as supporting electrolyte and must be there-
fore in excess. It turned out that presence of ethers, e.g. MTBE,
in the solvent was beneficial for the performance of the reaction
R1
R2
15
R1
R2
10
O
(
entry 5), whereas methanol, a common solvent in electroorganic
[
7]
20
synthesis, leads to significant reduction of the overall reaction
rate and yield (entry 4), which may, at least in part, be due to a
competing reaction of NO with methanol. Electrolysis per-
[O]
CO2
•
[8]
ꢁ
3
formed under an oxygen atmosphere only resulted in a slower
overall reaction, but had no influence on the products formed
11
Scheme 4. Proposed mechanism for the formation of 1,2-diketones 10
and the chain-shortened ketones 11 in the reaction of NO with alkynes 9.
(
data not shown).
•
3
To explore whether the ratio of 1,2-diketone/benzophenone,
Charges are omitted for the sake of clarity.
1
0a/11a, could be influenced by the electron density at the aro-
•
matic rings, NO was reacted with the various electron-deficient
3
and electron-rich diaryl alkynes 9b–9d and 9e–9h, respectively
[
9]
(
entries 6–12). All reactions resulted in formation of the
led to formation of 10a (entry 17), but the low yield rules this
method out as a general procedure for the oxidation of aromatic
alkynes to 1,2-diketones. Therefore, not single electron transfer
but radical addition to the C≡C triple bond must be the initial
respective diketones 10b–10h and benzophenones 11b–11h in
generally moderate to good combined yield.
the reaction of NO with the electron-deficient diaryl alkynes
c and 9d leads to less formation of 11. However, since the
[10]
It seems, as if
•
3
9
•
3
step in the NO induced oxidation of alkynes.
yields vary, we reason that no unequivocal dependence between
the product ratio and electron density at the aryl rings is appar-
ent.To conclude, our experiments strongly indicate that 10a–10h
and 11a–11h are formed via two competing intramolecular path-
ways, which obviously cannot be influenced by either external
forces, such as reaction conditions, nor by internal parameters,
like electron density at the C≡C triple bond.
Based on these results, formation of 1,2-diketones 10 and
•
ketones 11 in the reaction of alkynes 9 with NO could be
3
•
explained by the mechanism proposed in Scheme 4. NO3
addition leads to vinyl radical 12, which undergoes 5-endo
cyclization via transition state 13 to the five-membered ring
intermediate 14. The latter can undergo loss of nitrogen oxide,
•
NO , through transition state 15 to give the 1,2-diketone 10.
Control experiments were performed using the exemplary
On the other hand, formation of the chain-shortened ketone
•
3
reaction of 9a in order to reveal the role of NO in this oxi-
11 could occur by γ-fragmentation in 12 through transition
dation. Besides its ability to undergo addition to π-systems
•
state 16, resulting in an α-oxo carbene 17 and NO . The for-
•
◦
2
(
(
see Scheme 1), NO is also a strong one electron oxidant (E
[13]
3
mer may possibly be formed via an intermediate oxirene 18.
•
−
[11]
NO /NO ) = 2.0V versus SCE in acetonitrile).
However,
3
3
Wolff rearrangement through transition state 19 leads to ketene
reaction of 9a with various single electron oxidants, such as
•
3
2
0. Subsequent oxidation of 20 by excess NO (or directly
tris(4-bromophenyl)ammoniumyl hexachloroantimonate (entry
at the anode in the electrochemical experiments), followed by
decarboxylation would then give 11.
1
3), sulfate radical anions, SO• , which were in situ generated
−
4
[12]
through a Fenton reaction (entry 14),
as well as thallium(iii)
nitrate (entry 15) resulted in only minute amounts of 10a, no
formation of 11a, and mostly unconsumed starting alkyne. The
Computational Studies
•
reaction with the one electron oxidant and NO precursor CAN
In order to support the mechanism suggested in Scheme 4, DFT
computations were performed for the reaction of NO with sym-
3
•
3
was very slow, and some 10a and 11a were formed (entry 16).
It cannot be excluded, however, that traces of light may have
metrical alkynes possessing dialkyl (9i), diallyl (9j), and diaryl
substitution (9a, 9d, 9k) at the C≡C triple bond.The calculations
were carried out using the Gaussian03 program.[ Geometry
optimizations were performed with the hybrid density functional
•
produced small amounts of NO , which then reacted with 9a.
3
14]
Direct electrochemical oxidation of 9a at the anode performed
through replacement of LiNO3 by LiCl as supporting electrolyte

Oxidation of Aromatic Alkynes
423
Formation of 1,2-diketones Formation of ketones
• O NO
ONO2
O NO O•
O NO
O
2
2
2
NO
3
12
•
10
•
•
1,2 ꢁNO2
R
R1,2
1,2 ꢁNO2
R1,2
R1,2
R1,2
R
13
18
19
24
25
16
17
2
H O
HO
OH
HO
R1,2
O
12
•
[O]
ꢁ
NO
2
10
R1,2
R1,2
R1,2
26
27
Scheme 7. Possible alternative mechanism for the formation of 1,2-
diketones 10.
20
15
14
−
1
•
100 ± 10 kJ mol (data not shown). In Table 2 are compiled the
calculated energies of all stationary points, ꢀE, which are given
relative to the respective vinyl radicals 12. A computed potential
energy surface for the competing processess in vinyl radical 12
is shown in Scheme 5.
ꢁ
NO
[
O]
11
10
−CO2
Scheme 5. Potential energy surface for the formation of 1,2-diketones 10
and ketones 11 from the vinyl radicals 12 obtained after NO addition to the
•
3
Table 2 does not reveal any significant basis set dependencies
on the calculated energies for the reaction of NO with 2-butyne
i, which was used to benchmark the calculated results. This is
C≡C triple bond in alkynes 9 (BHandHLYP/6-311G** level of theory, ZPE
•
3
included).
8
in agreement with a recent computational study performed by
us on a closely related system using a large variety of differ-
ent ab initio and DFT methods, which showed that the energies
obtained with the BHandHLYP method in combination with
both the 6-311G** and cc-pVDZ basis set, respectively, agreed
very well with energies obtained from single-point calculations
at the CCSD(T)/cc-pVTZ//BHandHLYP/aug-cc-pVTZ level of
theory. Oxirene 18i, however, was the only case, where larger
basis sets led to an increasing stability. The relevance of this
result will be discussed below.
O
N
O
N
•
,214
O
O
O
O
(
R1 ꢃ aryl)
•
Ar
Ar
Ar
Ar
NO
2
1
•
ꢁ
O
[16]
O
N
10
N
O
O
O
O
In Fig. 1 are presented the structures and selected geometrical
parameters of key intermediates and transition states, as well as
the imaginary frequencies of the latter, νimag, calculated at the
BHandHLYP/6-311G** level of theory.[ For all ground and
transition states, the computed geometries around the central
vinyl moiety are very similar and not dependent on the nature
of the substitutents R1 . The only exception from this is vinyl
radical 12. Whereas the dialkyl substituted 12i is a σ-radical with
•
Ar
•
R
Ar Ar
23
22
17]
ꢁ
ꢁ
CO
NO
•
11
,2
Scheme 6. Possible alternative mechanism for the formation of 1,2-
diketones 10 and the chain-shortened ketones 11. Charges are omitted for
the sake of clarity.
1
the spin density being located at C and which exists with both E
[18]
or Z configuration,
the diallyl substituted vinyl radical 12j is
a linear π-radical. The diaryl substituted vinyl radicals 12a,d,k
are also mostly of π-type, but they possess a Z-configured double
◦
BHandHLYP in combination with the 6-311G**, cc-pVDZ, cc-
pVTZ and aug-cc-pVTZ basis sets (see below). All ground and
transition states were verified by vibrational frequency analysis,
and all identified transition states showed only one imaginary
bond with a clearly straightened angle α of 154–160 , whereas E-
configurated or linear isomers were not found. The spin density
in 12a,d,j,k is delocalized over the π system with 50% being at
1
1
C and the remaining 50% at the substituent R .
2
frequency. The spin expectation value, < s >, was very close
In all reactions under investigation, transition state 13 for the
5-endo cyclization of the vinyl radicals 12a, 12d, and 12i–12k
to the respective five-membered ring intermediate 14 is always
higher in energy than the transition state for the competing γ-
fragmentation and cleavage of NO , 16 (see Scheme 5). The
latter energy barrier of about 60 kJ mol (or about 50 kJ mol
to 0.75 after spin annihilation, except for transition state 16,
an electronically complex structure with presumably more than
one unpaired electron, where values ranging from 0.805 (16i) to
•
1
.002 (16k) after spin annihilation were obtained. The Gaussian
2
−
1
−1
output files for all structures shown in Schemes 4, 6, and 7 are
available as an Accessory Publication.
,
if zero-point vibrational energy correction, ZPE, is included), is
independent of the substitution pattern in vinyl radical 12, as
is the distance of 1.83–1.85 Å of the cleaving O–N bond. In
contrast to this, the activation energy required to reach transi-
tion state 13 shows a noticeable dependence on the nature of
the substituents R1 . The distance between the attacking oxy-
gen atom and the radical centre at the C=C double bond varies
from a minimum of 1.938 Å in 13j and a maximum of 2.011 Å in
13i.According to the BHandHLYP/6-311G** computations, the
The computations were performed for the competing path-
ways 12→14→10 and 12→17/18→20. For the bischloro- and
bishydroxy-phenyl substituted alkynes 9d and 9k, only the vinyl
radicals 12d/12k, transition states 13d/13k and 16d/16k, and
ground states 17d/17k and 19d/19k were calculated. Oxidative
transformation of ketenes 20 into ketones 11 has been reported
,2
[
15]
in literature and was not calculated in this work.
It should be
noted that formation of vinyl radicals 12 through addition of NO•
3
−
energy difference ꢀꢀE between 13 and 16 is about 18 kJ mol
1
to the various alkynes 9 was computed to be exothermic by about

4
24
U. Wille and J. Andropof
−1
•
3
Table 2. Relative energies ꢀE [kJ mol ] for stationary points in the NO -induced oxidation of alkynes 9 to 1,2-diketones 10 and ketones 11
Data in italics include the zero-point vibrational energy correction (ZPE)
12
13
14
15
10
16
17
18
19
20
11
(
a) Aliphatic alkynes: reaction of 9i with NO•
3
BHandHLYP/6-311G**
BHandHLYP/cc-pVDZ
BHandHLYP/cc-pVTZ
BHandHLYP/aug-cc–pVTZ
0.0
.0
0.0
.0
0.0
.0
0.0
.0
b) Allylic alkynes: reaction of 9j with NO
72.4
68.0
71.4
67.0
71.8
67.5
69.8
65.6
−121.5
−118.2
−123.3
−120.4
−125.4
−121.7
−127.6
−123.6
−114.2
−118.0
−117.0
−120.9
−115.8
−119.6
−117.5
−121.1
−419.0
−430.0
−410.5
−421.2
−407.8
−418.6
−408.8
−419.5
60.9
47.0
26.6
51.5
31.2
46.2
26.0
42.1
22.0
−121.5
−118.2
−123.3
−120.4
−125.4
−121.7
56.6
58.0
38.2
64.4
44.6
57.6
37.6
54.1
34.3
−208.1
−217.6
−199.0
−208.8
−205.2
−214.8
−206.5
−215.9
−409.1
−433.3
−392.2
−416.3
−399.4
−423.2
−400.5
−424.4
0
50.3
61.2
50.4
65.7
54.9
64.7
53.9
0
0
0
40.4
•
(
(
3
BHandHLYP/6-311G**
0.0
.0
0.0
.0
97.5
95.2
95.0
92.7
•
3
−102.3
−95.1
−107.0
−99.8
−96.5
−96.1
−368.7
−373.2
−362.9
−366.7
61.3
50.6
60.9
50.2
56.3
44.9
59.6
48.2
90.3
77.6
97.1
84.5
87.0
72.0
93.1
77.9
−181.9
−187.0
−175.2
−180.3
−362.4
−379.4
−347.1
−364.2
0
BHandHLYP/cc-pVDZ
−102.1
−101.5
0
c) Aromatic alkynes: reaction of 9a with NO
BHandHLYP/6-311G**
0.0
.0
0.0
.0
82.6
80.3
81.1
78.4
−87.7
−82.3
−90.4
−85.8
−81.0
−82.6
−84.6
−86.7
−388.2
−393.6
−381.5
−387.1
61.1
51.4
60.6
50.8
35.9
24.4
39.7
27.9
72.5
58.8
79.8
65.7
69.6
54.1
77.0
60.9
−187.3
−194.9
−179.0
−187.3
−381.4
−399.2
−365.8
−384.1
0
BHandHLYP/cc-pVDZ
0
Reaction of 9d with NO•
3
BHandHLYP/6-311G**
0.0
.0
0.0
.0
83.6
81.2
82.0
79.2
60.0
50.3
59.8
49.9
34.1
22.6
37.6
25.8
69.8
54.3
77.0
60.7
0
BHandHLYP/cc-pVDZ
0
Reaction of 9k with NO•
3
BHandHLYP/6-311G**
0.0
.0
0.0
.0
79.9
77.1
79.1
75.8
61.5
51.8
61.5
51.6
16.7
5.9
20.8
9.6
69.8
54.3
66.5
50.3
0
BHandHLYP/cc-pVDZ
0
(
9
including ZPE) in the reaction of the dialkyl substituted alkyne
i. In the case of the aromatic alkynes 9a, 9d, and 9k, transition
to give 10, or a neophyl rearrangement with subsequent opening
of the cyclopropanone ring 22→23 followed by decarbonylation
−
1
•
state 13 is about 28 kJ mol (including ZPE) higher in energy
than 16, which shows that varying electron densities in the aryl
rings have no influence on the barrier height of 13. Interestingly,
for the diallyl-substituted alkyne 9j, transition state 13j is ener-
getically higher by even 45 kJ mol (including ZPE) than 16j.
A possible explanation for this finding will be given below.
Once transition state 13 is overcome, formation of the respec-
tive 1,2-diketones 10 is energetically straightforward. Compared
to the starting vinyl radicals 12, the five-membered ring inter-
mediates 14 are thermodynamically favourable, with ꢀE val-
and NO cleavage to give 11.
BHandHLYP calculations (using the 6-311G** and cc-pVDZ
basis set) performed for the exemplary reaction system 14a
revealed, however, that this neophyl rearrangement is energet-
ically uphill (the optimized structures are not shown). Although
the half-cleaved radical 21 could be located at this level of the-
−
1
[19]
−
1
ory, which is about 160 kJ mol more stable than 14a (analogue
aliphatic derivatives were not found), 21 possesses an extremely
−
1
weak O–N bond that requires only 1 kJ mol activation bar-
•
rier (ZPE included) to release NO and to give 10 in a strongly
−
1
ues ranging from −118 kJ mol in the case of 14i to about
exothermic reaction. On the other hand, ꢀE between 21 and
−1
−1
−
82 kJ mol in the case of 14a (ZPE included). The frag-
the acyl radical 23 is 70 kJ mol endothermic. Thus, the neo-
mentation into the respective diketones 10 and release of the
phyl rearrangement cannot compete with the pathway leading to
1,2-diketone formation.
•
comparably stable NO through transition state 15 is a rapid pro-
cess requiring only few kJ mol , independent of the nature of
R
ner (see below). The O–N bonds to be broken in 15 are only
slightly elongated, compared to 14, which is in accordance with
the Hammond postulate that exothermic reactions have early
−
1
Therefore, we assume that formation of 11 must proceed
through the initially suggested pathway, out of vinyl radical 12,
1
,2
. This process may proceed in a one-step or step-wise man-
•
2
through γ-cleavage of the comparatively stable NO via tran-
sition state 16 leading to the α-oxo carbenes 17, which is a
distinctly endothermic process for all vinyl radicals. Accord-
ing to the BHandHLYP calculations using the 6-311G** or the
cc-pVDZ basis set, initial formation of oxirene intermediates 18,
which subsequently undergo isomerization into α-oxo carbenes
17 is highly unlikely, because the oxirenes are even higher in
energy than transition state 16.With the larger cc-pVTZ and aug-
cc-pVTZ basis sets, it appears, however, that at least the dialkyl
substituted oxirene 18i becomes more stable than transition state
[
20]
transition states.
•
Loss of NO to give 1,2-diketones appears to be the only
energetically feasible reaction for the cyclized radical 14. A pos-
sible alternative mechanism, at least for aromatic alkynes, which
could lead from 14 to both 1,2-diketones 10 and ketones 11 is
shown in Scheme 6. This includes a stepwise ring opening to the
resonance-stabilized radical 21, followed by either loss of NO•

Oxidation of Aromatic Alkynes
425
aR1C1C2
a: 160.0 (Z)
d: 159.9 (Z)
i: 141.7 (E)
nimag
a: 884i
d: 894i
i: 809i
j: 957i
k: 852i
qC1C2ON j: 179.1
a: 100.1 k: 153.9 (Z)
d: 98.4
a: 1.537
i: 1.534
j: 1.530
a: 1.431
i: 1.436
j: 1.440
q
1
nimag
ONOC
a: 150.0
i: 148.7
j: 150.7
a: 888i
i: 884i
j: 812i
i: 104.7
j: 106.3
k: 97.0
aR1C1C2
a: 150.7
d: 150.5
i: 147.5
j: 149.6
k: 150.9
a: 1.976
d: 1.968
i: 2.011
j: 1.938
k: 2.000
12
2
a: 1.844
d: 1.849
i: 1.831
j: 1.844
k: 1.835
1
4
15
1
3
aR2C1C2 nimag
a: 1.430
d: 1.431
i: 1.427
j: 1.432
k: 1.431
a
1
1
C
2
R
C
qR1C1C2O
a: 98.4
d: 97.6
i: 100.4
j: 103.7
k: 96.1
qR1C1C2O
a: 78.9
d: 77.4
i: 80.0
a: 88.2 a: 287i
nimag
a
R C C
1 1 2
a: 164.3
i: 164.4
j: 163.8
a: 509i a: 158.2
d: 499i d: 159.0
i: 534i i: 140.4
j: 508i j: 178.2
k: 530i k: 154.0
d: 87.7
i: 90.3
j: 88.0
k: 85.5
d: 297i
i: 305i
j: 333i
k: 337i
j: 80.0
a: 1.467
i: 1.484
j: 1.459
a: 1.369
i: 1.254
j: 1.271
k: 78.3
a: 1.597
d: 1.600
i: 1.622
j: 1.597
k: 1.603
a: 2.057
d: 2.048
i: 2.115
j: 2.053
k: 2.005
16
17
18
19
◦
◦
1
2
Fig. 1. Geometrical parameters (bond length [Å], angles α [ ], and dihedral angles θ [ ]; data involving R /R refer to the carbon directly
−
1
attached to the central vinyl moiety) of important ground and transition states (imaginary frequencies, νimag [cm ]) in the oxidation of alkynes
9a,d,i,j,k to 1,2-diketones 10a,d,i,j,k and ketones 11a,d,i,j,k, determined at the BHandHLYP/6-311G** level of theory.
1
6i. We have made no attempts to resolve this discrepancy, but
for the bisallyl and bisaryl substituted vinyl radicals 12a, 12d,
12j, and 12k than for the aliphatic 12i. This can be explained by
the resonance stabilization of the unpaired electron in the former
vinyl radicals, which is sacrificed by cyclization to 14, in which
the spin density is only delocalized between the two ring oxygen
atoms (11% each), the nitrogen atom (46%), and the exocyclic
since 18i is still significantly less stable than the α-oxo carbene
7i, we reason that oxirenes are not formed as intermediates in
1
this reaction.
A crucial step in the proposed mechanism is the Wolff
rearrangement of the α-oxo carbenes 17 to give the thermo-
dynamically favourable ketenes 20.[ Reaching transition state
21]
[16]
oxygen atom (30%). On the other hand, the aliphatic, not res-
1
9, which was located for all systems under investigation,
onance stabilized vinyl radical 12i benefits energetically from
an increased delocalization of the unpaired electron in 14i.
Alternatively to the suggested 5-endo cyclization/
fragmentation pathway for the formation of 1,2-diketones 10
shown in Scheme 4, a mechanism, which involves the barrier-
−
1
requires an activation energy of around 12 kJ mol in the case
−
1
of the dialkyl substituted 17i, and 30–40 kJ mol in the case of
the diallyl and diaryl substituted α-oxo carbenes 17a, 17d, 17j,
and 17k (ZPE included). 19 is an unsymmetrical, early transition
state, in which the C –C –R bonds form a right angle triangle
with the to be formed R –C bond being longer by ca 0.4 Å than
1
2
2
•
less recombination of vinyl radical 12 with a second NO to
3
2
1
the closed-shell dinitrate 24, followed by stepwise elimina-
2
2
•
2
the to be broken R –C bond (Fig. 1).
tion of NO , is also possible (Scheme 7). The computations
Under the assumption that the subsequent oxidation of ketene
0 and decarboxylation to ketone 11 is fast, our calculations
(BHandHLYP with the 6-311G** and cc-pVDZ basis set)
performed exemplary for the system 12a revealed that the
first homolytic fragmentation, 24→25, although being slightly
2
would predict preferred formation of 11 rather than the 1,2-
diketone 10, which contrasts the experimental outcomes shown
inTable 1. Closer inspection of the relative energies ꢀE (Table 2)
reveals that in all cases, except of the aliphatic and bishydroxy-
phenyl substituted α-oxo carbenes 17i and 17k, respectively,
Wolff rearrangement through transition state 19 requires slightly
−
1
exothermic (−14 kJ mol , including ZPE), requires an activa-
−
1
tion energy of 110 kJ mol (including ZPE), which is signifi-
cantly higher than any other barrier calculated for this reaction
•
system (optimized geometries not shown).The second NO loss,
2
−
1
25→10, requires on the other hand only 4 kJ mol . Because of
•
2
•
2
more energy than the reverse addition of NO to the carbonyl
the large energy barrier for the first NO cleavage, we believe
oxygen in 17 via transition state 16, to regenerate vinyl radical
that such a stepwise radical fragmentation pathway is unlikely.
However, it is also possible that the dinitrate 24 could undergo
hydrolysis to the enediol 26 and its tautomer 27, which could
then be oxidized to give 10. Since this pathway would require
the presence of water, the reactions of 9 with photogenerated
1
2.Thus, we believe that the bottleneck of the pathway leading to
•
ketone 11 is not loss of NO and formation of the α-oxo carbene
2
[22]
1
7, but Wolff rearrangement of the latter to the ketene 20.
The ratio of the carbonyl compounds 10 and 11 formed in the
•
3
•
3
reaction of NO with alkynes 9 reflects therefore the difference
NO , which were performed in acetonitrile, should not give any
in barrier heights between transition states 13 and 19, rather than
the energy difference between 13 and 16.
As mentioned earlier, the activation barrier leading to tran-
sition state 13 for the 5-endo cyclization is significantly higher
diketones 10, whereas those performed in the presence of water
should lead to a higher yield of 10.The results inTable 1 (entries
1 and 3) show that the opposite was the case. In the photochem-
ical experiments using acetonitrile as the solvent, the sample

4
26
U. Wille and J. Andropof
came into contact with water only during work-up. Since we
have no indication from our GC/MS experiments for the forma-
tion of any 26 and/or 27, a pathway involving a dinitrate 24 must
be excluded.
Formation of oxirenes 18 as intermediates is unlikely, because
they are generally even higher in energy than transition state
16. Wolff rearrangement of the substituent R1 in 17 leads to
ketenes 20, which are believed to undergo further oxidation by
,2
•
How can we understand why aliphatic alkynes of type 1 react
either NO or directly at the anode. Subsequent decarboxylation
3
•
3
with NO in a radical cyclization sequence to yield anellated
leads to the observed chain-shortened ketones 11. The Wolff
rearrangement 17→20 requires generally slightly more energy,
tetrahydrofurans 2 (see Scheme 1), whereas aromatic alkynes of
type 6 give 1,2-diketones (see Scheme 2)? Computations per-
formed at the BHandHLYP/6-311G** level of theory revealed
for the 1,5-HAT in the σ-type vinyl radical 3 an activation barrier
•
2
than the reverse addition of NO to the carbonyl oxygen in 17,
which regenerates vinyl radical 12. We therefore reckon that he
experimentally observed product ratio 10/11 reflects the energy
difference between the transition states 13 and 19, rather than
the energy difference between transition states 13 and 16.
−
1 [23]
of about 40 kJ mol
.
This is significantly below transition
state 13i for 5-endo cyclization of the σ-type vinyl radical 12i
to the radical intermediate 14i or transition state 16i for the γ-
•
3
ReactionofNO withaliphaticalkynesoftype1isdifferentto
•
2
cleavage of NO (see Scheme 4). Thus, in aliphatic alkynes of
that of aromatic alkynes of type 6, because the activation barrier
for a 1,5-HAT in the σ-type vinyl radical 3 is significantly lower
than the barrier for both 5-endo cyclization to a five-membered
ring intermediate of type 14 and for γ-fragmentation to the α-
oxo carbene 17. On the other hand, a 1,5-HAT in the π-type
type 1 formation of 1,2-diketones is energetically less favourable
than an intramolecular 1,5-HAT that ultimately leads to the
tetrahydrofuran 2. On the other hand, the orbital of the reso-
nance stabilized unpaired electron in the π-type vinyl radical
•
3
•
3
derived from NO addition to the aromatic alkyne 6 (not shown)
vinyl radical derived from NO addition to the C≡C triple bond
is perpendicular to the ether side chain. The 1,5-HAT requires
overlap of the SOMO with the σ-orbital of a C–H bond at the
methylene carbon atom positioned α to the ether oxygen atom,
which is geometrically impossible.
in 6 is geometrically not possible, because the resonance stabi-
lized orbital containing the unpaired electron and the σ orbital
of a suitable C–H bond in the ether side chain are not correctly
aligned.
•
3
In this work, we have shown that NO is synthetically excep-
tionally versatile, as it cannot only transfer one, but even two
of its oxygen atoms to alkynes. In the pathways leading to 1,2-
Conclusions
Reactions of NO• with alkynes are remarkably complex.
Whereas with aliphatic alkynes of type 1, initial addition of NO
diketones 10 and ketones 11, either NO or NO , respectively,
•
•
3
2
•
3
are released, which are highly unreactive radicals, compared to
•
to the C≡C triple bond is followed by an intramolecular radi-
NO , and do not act as radical chain carriers under the exper-
3
[24]
•
3
cal cyclization, formation of 1,2-diketones is the predominant
imental conditions.
Therefore, the described NO induced
•
3
pathway in the reaction of NO with aromatic alkynes of type
oxidation of aromatic alkynes can be considered as a new exam-
ple of self-terminating radical oxygenations. In addition, this
work reveals an entirely new method to generate carbene inter-
mediates through γ-fragmentation in vinyl radicals. We are
currently exploring the generality of this methodology in our
laboratory.
•
3
6
. Exploration of the general applicability of NO in transform-
ing C≡C triple bonds into dicarbonyl compounds revealed that
diaryl-substituted alkynes 9a–9h can indeed be oxidized to the
respective 1,2-diketones 10a–10h in moderate to good yields,
but they were always accompanied by ketones 11a–11h, which
arearisingthroughformallossofonecarbonatom. Electrochem-
ical generation of NO appeared to be the preferable method,
since the photochemical conditions for NO production led to
destruction of the 1,2-diketones.
•
3
Experimental
•
3
Diphenylacetylene 9a, CAN, and LiNO3 were supplied from
The experiments showed that (a) the oxidation is initiated
Aldrich. The mono- and di-substituted alkynes 9b–9h were pre-
•
[25]
by NO addition to the C≡C triple bond and not by single
pared according to literature.
The photochemical reactions
3
electron transfer, and that (b) the 1,2-diketones 10 and the
chain-shortened ketones 11 must be formed through compet-
ing pathways of the initial vinyl radicals 12, which cannot be
influenced by changing the reaction conditions or by varying the
were performed using a TQ150 lamp (Heraeus Hanau). The
electrochemical reactions were performed in an undivided cell
equipped with a cyclindrical platin net anode and a centred
corkscrew-shaped platin counter electrode at potentials of 2.0–
2.4V using a home-made potentiostat. The reaction cell was
cooled with a water bath. Electrolysis was usually carried out
until TLC indicated complete consumption of the alkyne. Reac-
tion analysis and product identification was performed by GC
and GC/MS and comparing the MS data with authentic samples
(in the case of 10a/11a) and literature data.[ GC: Varian CP
3380, column CP-Sil 5 CB, 30 m; and Supelco SPB-1701, 30 m;
temperature program (both columns) 1005→2505 heating rate
electron density at the C≡C triple bond. However, since 10 and
•
3
1
1 can be easily separated by column chromatography, the NO -
induced oxidation of aromatic alkynes to 1,2-diketones could be
an alternative in those cases, where the existing synthetic meth-
•
3
ods are not suitable. Since the electrochemical NO generation
6]
can be considered as a reagent-free technique, this new method
is also environmentally highly interesting.
DFT computations revealed that 1,2-diketones 10 result from
-endo cyclization of the vinyl radicals 12 to five-membered ring
◦
−1
5
5 C min . GC/MS: Shimadzu GC-17A/Shimadzu QP5050.
The flame ionization detector of the GC was calibrated
for quantitative analysis using authentic samples of 9a, 10a,
and 11a and a standard (n-hexadecane or cyclododecanone).
The obtained peak area correction factors, fGC(a), for 9a, 10a,
and 11a, respectively, were compared with correction values,
fECAN(a), obtainedforthesamecompoundsusingtheincremental
ECAN (effective carbon number) method,[ to give the corre-
lation factor fc for each compound, which is fc = fGC(a)/fECAN(a).
intermediates 14, which fragment into 10 and the comparatively
•
stable NO in a very fast subsequent step. The ketones 11, on
the other hand, are formed through γ-fragmentation and loss
•
2
of NO already at the stage of the vinyl radicals 12 to yield at
first an intermediate α-oxo carbene 17 in an endothermic step.
The activation energy required for the γ-cleavage is not only
independent of the substitution pattern in the vinyl radicals 12,
but is also always lower than the competing transition state 13.
26]

Oxidation of Aromatic Alkynes
427
The respective peak area correction factors for 9b–9h, 10b–
[7] T. Shono, Electroorganic Chemistry as a New Tool in Organic
Synthesis 1984 (Springer: Berlin).
1
0h and 11b–11h, fGC(b-h), were then determined from this
[
8] O. Ito, S. Akiho, M. Iino, J. Org. Chem. 1989, 54, 2436.
correlation factor according to fGC(b−h) = fc × fECAN(b−h).
doi:10.1021/JO00271A038
[9] We have also studied the reaction of NO• with mixed aryl–alkyl,
•
Reaction of 9a with Photogenerated NO (Typical
3
3
dialkyl, and terminal aromatic alkynes. However, numerous products
were obtained in each case, which were not further investigated.
Experimental Procedure)
In a pyrex test tube equipped with a rubber septum and an argon
balloon, 90 mg (164 µmol) CAN and 9.9 mg (55 µmol) 9a were
dissolved in 10 mL of acetonitrile and degassed in an ultrasound
bath for 10 min under a constant flow of argon. With argon
gas flowing, the sample was fixed at the cooling mantle of the
lamp with rubber bands and irradiated for 80 min. The irradi-
ation time was shorter (20–30 min), when smaller amounts of
CAN were used. After addition of a weighed amount of standard
•
3
[
10] The very low yield obtained in the reaction of NO of with 9h could be
explained by a competing direct one electron oxidation of the electron-
•
3
rich aromatic rings by NO .
[11] E. Baciocchi, T. Del Giacco, S. M. Murgia, G. V. Sebastiani, J. Chem.
Soc. Chem. Commun. 1987, 1246. doi:10.1039/C39870001246
12] U. Wille, Org. Lett. 2000, 2, 3485. doi:10.1021/OL006527Y
13] (a) E. G. Lewars, Chem. Rev. 1983, 83, 519. doi:10.1021/
CR00057A002
[
[
(
b) J. E. Fowler, J. M. Galbraith, G. Vacek, H. F. Schaefer, III, J. Am.
Chem. Soc. 1994, 116, 9311. doi:10.1021/JA00099A057
c) M. Torres, J. L. Bourdelande, A. Clement, O. P. Strausz, J. Am.
(
n-hexadecane), the solution was poured into brine and extracted
with diethyl ether. The combined organic fractions were dried
over MgSO4, concentrated, and analyzed by GC.
(
Chem. Soc. 1983, 105, 1698. doi:10.1021/JA00344A072
14] M. J. Frisch, G. W.Trucks, H. B. Schlegel, G. E. Scuseria, M.A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C.
Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Men-
nucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji,
M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,
T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E.
Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J.V. Ortiz, Q. Cui,A. G. Baboul, S. Clif-
ford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C.Y.
Peng,A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian 03, Revision
D.01 2004, (Gaussian, Inc.: Wallingford, CT).
[
•
Reaction of 9a with Electrogenerated NO (Typical
Experimental Procedure)
3
3
47 mg (5.03 mmol) LiNO3 and 358 mg (2.01 mmol) 9a were
dissolved in 50 mL of acetonitrile/water/MTBE (7/1/2) and elec-
trolyzed until TLC showed complete consumption of 9a (2 h).
After addition of a weighed amount of standard (n-hexadecane),
the solution was poured into brine and extracted with diethyl
ether. The combined organic fractions were dried over MgSO4,
concentrated and analyzed by GC.
Accessory Publication
The accessory publication is available from the author or, until
June 2012, the Australian Journal of Chemistry.
Acknowledgments
[15] (a)T.W. Koenig,T. Barklow,Tetrahedron 1969, 25, 4875. doi:10.1016/
S0040-4020(01)83028-3
This work was supported by theAustralian Research Council under theARC
Centres of Excellence program, theVictorian Institute for Chemical Sciences
High Performance Computing Facility, and the Australian Partnership for
Advanced Computing. We thank Hakki Turupçu and Christine Duke for
their support.
(
b) S. Dayan, I. Ben-David, S. Rozen, J. Org. Chem. 2000, 65, 8816.
doi:10.1021/JO001101I
[
[
16] U. Wille, J. Org. Chem. 2006, 71, 4040. doi:10.1021/JO0520543
17] The geometries computed at the various levels of theory were found
to vary only insignificantly.
[
18] Computations performed by us have shown that the E-configured
References and Notes
−1
vinyl radical 12i is less stable by about 4 kJ mol than the Z isomer:
[
1] (a) U. Wille, L. Lietzau, Tetrahedron 1999, 55, 10119. doi:10.1016/
(a) U. Wille, T. Dreessen, J. Phys. Chem. A 2006, 110, 2195.
doi:10.1021/JP0454772J.
S0040-4020(99)00564-5
(
(
b) U. Wille, L. Lietzau, Tetrahedron 1999, 55, 11456.
c) L. Lietzau, U. Wille, Heterocycles 2001, 55, 377.
(b) J. Fossey, D. Lefort, J. Sorba, Free Radicals in Organic Chemistry
1995 (John Wiley & Sons: Chichester). For convenience, we have
used in this work the E-configured vinyl radical 12i as energetic ref-
erence, because no configurational change at the C=C double bond
is required in going from 12i to 14i via transition state 13i.
[
2] (a) U. Wille, Chem. Eur. J. 2002, 8, 340. doi:10.1002/1521-
765(20020118)8:2<340::AID-CHEM340>3.0.CO;2-4
b) T. Dreessen, C. Jargstorff, L. Lietzau, C. Plath, A. Stademann, U.
3
(
Wille, Molecules 2004, 9, 480.
[19] Solvation effects, which were calculated using the PCM model as
implemented in Gaussian03 did not result in significant changes of
the barrier heights of any transition state 13 and 16, respectively.
[20] G. S. Hammond, J. Am. Chem. Soc. 1955, 77, 334. doi:10.1021/
JA01607A027
[21] This rearrangement has been calculated using methods involving
high levels of electron correlation and basis sets; see for example:
(a) A. P. Scott, R. H. Nobes, H. F. Schaefer, III, L. Radom, J. Am.
Chem. Soc. 1994, 116, 10159. doi:10.1021/JA00101A039
(b) P. J. Wilson, D. J. Tozer, Chem. Phys. Lett. 2002, 352, 540.
doi:10.1016/S0009-2614(01)01496-8
[
[
3] L. Lietzau, Ph.D. Thesis 1999 (University of Kiel: Kiel).
4] (a) N. S. Srinivasan, D. G. Lee, J. Org. Chem. 1979, 44, 1574.
doi:10.1021/JO01323A049
(
b) D. G. Lee, V. S. Chang, W. D. Chandler, J. Org. Chem. 1985, 50,
306, doi:10.1021/JO00222A021
c) Z. L. Zhu, J. H. Espenson, J. Org. Chem. 1995, 60, 7728.
doi:10.1021/JO00129A010
d) M. S. Yusubov, G. A. Zholobova, S. F. Vasilevsky, E. V. Tretyakov,
D. W. Knight, Tetrahedron 2002, 58, 1607. doi:10.1016/S0040-
020(02)00025-X
4
(
(
4
[
[
5] U. Wille, C. Plath, Liebigs Ann./Recueil 1997, 111, and references
therein.
6] Mass Spectra in NIST Chemistry WebBook, NIST Standard Refer-
ence Database Number 69 (Eds P. J. Linstrom, W. G. Mallard) 2005
(c) G. Vacek, J. M. Galbraith, Y. Yamaguchi, H. F. Schaefer, III,
R. H. Nobes, A. P. Scott, L. Radom, J. Phys. Chem. 1994, 98, 8660.
doi:10.1021/J100086A013
•
[22] We did not investigate whether the reaction of NO with alkynes in the
3
•
2
(
National Institute of Standards and Technology: Gaithersburg, MD).
presence of excess NO leads to exclusive formation of 1,2-diketones,
http://webbook.nist.gov
in order to support the assumed reversibility of the γ-fragmentation

4
28
U. Wille and J. Andropof
•
•
1
2→17. Under such conditions, NO is trapped by NO to give N2O5,
[26] (a) J. C. Sternberg, W. S. Gallaway, D. T. L. Jones, Gas Chromatog-
raphy (Eds N. Brenner, J. E. Callen, M. D. Weiss) 1962 (Academic
Press: New York, NY).
3
2
thus rendering this experiment insignificant.
23] U. Wille, T. Dreessen, unpublished.
24] NO is a much weaker oxidant than the excess compound NO . Sec-
ondary reactions arising from oxidation processes initiated by NO
are therefore highly unlikely.
25] Z. Novák, P. Nemes, A. Kotschy, Org. Lett. 2004, 6, 4917.
[
[
•
2
•
3
(b) G. Guichon, C. L. Guillemin, J. Chromatogr. Libr. 1988, 42, 445.
(c) J. T. Scanlon, D. E. Willis, J. Chromatogr. Sci. 1985, 23, 333.
(d) A. D. Jorgensen, K. C. Picel, V. C. Stamoudis, Anal. Chem. 1990,
62, 683. doi:10.1021/AC00206A007
•
2
[
doi:10.1021/OL047983F