Dimerization of (+)-lysergic acid esters

Article abstract of DOI:10.3987/COM-07-11011

Dimer isomer mixtures, characterized by a bridgehead C8-C8′ bond, (6a-7a; 6b-7b) were obtained from (+)-lysergic acid methyl or ethyl ester (1b; 1c) in a solution of methanol or ethanol. The isomers were separated, and their structures were determined by detailed NMR measurements and X-ray analysis. Density functional theory was applied to provide insight into the reaction mechanism. Based on an extended examination and the theoretical calculations, a plausible reaction sequence leading to dimers is also presented. The proposed mechanism has been verified by detecting the formation of the superoxide radical anion (O₂^·-).

Full text of DOI:10.3987/COM-07-11011

HETEROCYCLES, Vol. 71, No. 5, 2007
1075
HETEROCYCLES, Vol. 71, No. 5, 2007, pp. 1075 - 1094. © The Japan Institute of Heterocyclic Chemistry
Received, January 24, 2007, Accepted, 9th April, 2007, Published online, 10th April, 2007. COM-07-11011
DIMERIZATION OF (+)-LYSERGIC ACID ESTERS§
István Moldvai,*,a Eszter Gács-Baitz,b Eszter Termesvári-Major,a Luca
Russo,c Imre Pápai,d Kari Rissanen,c Éva Szárics,e Julianna Kardos,e
and Csaba Szántaya,f
aChemical Research Center, Hungarian Academy of Sciences, Institute of
Biomolecular Chemistry, Department of Natural Organic Compounds, H-1525
Budapest, POB 17, Hungary; bCRC, HAS, Institute of Structural Chemistry, NMR
Laboratory, H-1525 Budapest, POB 17, Hungary; cNanoScience Center,
University of Jyväskylä, Department of Chemistry, POB 35, 40014 JYU, Finland;
dCRC, HAS, Institute of Structural Chemistry, Department of Theoretical
e
Chemistry, H-1525 Budapest, POB 17, Hungary; CRC, HAS, Institute of
Biomolecular Chemistry, Department of Neurochemistry, H-1525 Budapest, POB
17, Hungary; fResearch Group for Alkaloid Chemistry, HAS, Budapest University
of Technology and Economics, H-1521 Budapest, Gellért tér 4, Hungary
e-mails:
imoldvai@chemres.hu
egacs@chemres.hu
krissane@cc.jyu.fi
papai@chemres.hu
jkardos@chemres.hu szantay@mail.bme.hu
Abstract – Dimer isomer mixtures, characterized by a bridgehead C8-C8’ bond,
(6a-7a; 6b-7b) were obtained from (+)-lysergic acid methyl or ethyl ester (1b; 1c)
in a solution of methanol or ethanol. The isomers were separated, and their
structures were determined by detailed NMR measurements and X-ray analysis.
Density functional theory was applied to provide insight into the reaction
mechanism. Based on an extended examination and the theoretical calculations, a
plausible reaction sequence leading to dimers is also presented. The proposed
mechanism has been verified by detecting the formation of the superoxide radical
anion (O2• −).
INTRODUCTION
Studies on ergot alkaloids have been in the focus of interest for about two centuries.1 Conscious
applications of the natural products and their derivatives in the regular medicine2 represented the first
essential step. On the other hand, from a chemical point of view, the isolation and preparation of the first
homogeneous ergot alkaloid performed by Sandoz’s researchers3 can be regarded as a milestone in this
area. The majority of the ergot alkaloids contain (+)-lysergic acid 1a or (+)-isolysergic acid 2a as the
principal constituent of the molecule (Figure 1).
§A few aspects of this work have been partly presented on “1st European Chemistry Congress” (27-31
August, 2006; Budapest Hungary; N-PO-183).

1076
HETEROCYCLES, Vol. 71, No. 5, 2007
(S)
8
(R) CO2R
CO2R
H
H
9
7
12
H
H
10
N
N
13
14
Me
Me
6
11
(R)
5
4
15
16
3
2
a R=H
b R=Me
c R=Et
HN
1a, 1b, 1c
HN
1
2a, 2b, 2c
Figure 1 Structure of (+)-Lysergic acid (1a) and (+)-Isolysergic acid (2a) and their Esters.
After isolation of 1a in 1934,4 the preparation of a few simple derivatives, e.g. its methyl ester (1b)4 was
performed in the earlier years. The structure determination of 1a and its derivatives, including the
stereochemistry, was terminated in the early sixties years.2 During the last seven decades, various special
properties of 1a and its derivatives have been presented. Isomerization to 2a by heating a solution of 1a in
several types of solvents (water, alcohols) was first described by Wellcome’s group.5 Esters 1b and 2b
were prepared from acids 1a and 2a by methylation.5 Mutarotation of 1b in MeOH6 was also observed
and was supposed to occur through a double bond migration.7 The relationship between 1a and 2a and
their derivatives was explained later as C8-epimerization via an “enolic intermediate” 3 (Figure 2).8
HO
X
CO2H
H
NHMe
N
Me
3 X= OH, OMe, NR1R2
HN
HN
4
Figure 2 Structures of hypothetic compounds 3 and 4.
Wellcome’s group presented the second remarkable property of 1a in 1936. On treating 1a with aqueous
o
barium hydroxide solution under harsh reaction conditions (sealed-tube, 150 C, 4 h) racemic product
(±)-1a was obtained.6 The racemization was interpreted by Woodward as occurring via an achiral
intermediate 4 (Figure 2), and his hypothesis inspired several groups to elaborate their total syntheses2,9 or
approaches2,9 to (±)-1a, applying, however, 2,3-dihydroindole derivatives.
The third curiosity, the so-called indole-naphthalene isomerization, was described in 1953.10 Acid (±)-1a
was esterified with EtOH/HCl at 0 oC, then the solution of the (±)-1c obtained was heated in a sealed-tube.
After then, the reaction mixture was acylated by Ac2O, and at the end of the transformations naphthalene
5a was isolated. The same result was obtained in MeOH. This irreversible isomerization was interpreted
by simple energy calculations: naphthalenes proved to be more stable than indoles. This isomerization

HETEROCYCLES, Vol. 71, No. 5, 2007
1077
highly influenced research groups dealing with the total synthesis of (±)-1a, the majority of them tried to
avoid the approaches with indole derivatives, but rather they used dihydroindoles.
X
9
(a) EtOH or MeOH / HCl, 0 oC,
then 100 oC in a sealed-tube, 20 min
(b) K2CO3 / Ac2O
8
7
10
1
10a
a, b
N
6a
(±)-1a
or
2
3
Me
10b
3b
6
1a
3a
Ac
5a
5
N
4
5c X=(8R)-CO2Et; 56%
5d X=(8R)-CO2Me; 48%
5a X=CO2Et
5b X=CO2Me
Figure 3 Indole-Naphthalene Isomerization.
Finally, we have to mention the formation of the so-called “lumi-derivatives”; C10-OH or C10-OMe
compounds could be obtained from amides of 1a in H2O/H+ or MeOH/H+ under irradiation of the
solutions with UV light.11,12
RESULTS AND DISCUSSION
1) Synthetic aspects - Considering the above-presented extensive and lingering studies, one might hardly
expect any new transformations in this field. Nevertheless, as a part of our continued interest in the
reactions of the ergot derivatives, we have re-examined a few aspects of the above properties of 1a and its
derivatives. During this work an unexpected reaction of 1b and 1c, an unprecedented dimerization in the
ergot alkaloids field was discovered under mild conditions.
To revisit and complete some of the earlier studies cited above, the indole-naphthalene isomerization was
repeated starting with optically active 1a. Naphthalene 5c and 5d were obtained via the Stoll’s route
reserved the optical activity {5c: [α]D +62o (c 1.99, CHCl3), 5d: [α]D +16o (c 0.25, CHCl3)} showing that
no racemization via C-8 epimerization occurred in the circumstances applied (Fig. 3). The absolute
configuration of both 5c and 5d at C-8 must be (R). In the dominant conformation of N-Ac- naphthalenes
(5a-5d) exists in a quasi-equatorial position, which is supported by the measured coupling constant value
(9.8 Hz) between vicinal protons H-7β and H-8, a characteristic value for trans-diaxial relative
arrangement.
Next the racemization of 1a was re-examined. Repeating the original procedure we found that (±)-1a did
really formed in good yield when 1a was heated in sealed-tube in barium hydroxide solution.
Furthermore we found that barium hydroxide solution has no distinctive role; potassium hydroxide
solution gave the same result (yield: 78%). During the recrystallization, a small crop (10%) of pure (±)-2a
has also been isolated and characterized by spectroscopic methods. With pure (±)-2a in our hand, we tried
to transform it into (±)-2b with diazomethane but the product obtained proved to be a mixture of (±)-1b
and (±)-2b (ratio: ≈ 11:9).#

1078
HETEROCYCLES, Vol. 71, No. 5, 2007
#In our first direct synthesis of 1a, the structures of both 2a and 2b were determined by NMR spectroscopy in
mixtures.13 The formation of epimer mixture as end product under the alkylation can be rationalized in the
following way: Acid (±)-2a dissolves poorly in acetone. The epimerization of (±)-2a into the equilibrium
containing (±)-2a and (±)-1a is faster then the methylation.
As only epimerization of 1a was investigated and published earlier,5 we also tried a similar
transformation of ester 1b into 2b in MeOH solution. On boiling a solution of 1b (MeOH, overnight) or
o
heating it at 40 C for 48 h, completely unexpected dimer isomers (6a and 7a) were formed. The same
result was obtained when the solution of 1b was allowed to react at room temperature for about 5-6 days.ε
Me
Me
H
H
N
N
14
1
RO2C
5'
RO2C
8'
NH
NH
1'
HN
HN
8
5
CO2R
CO2R
14'
N
N
H
H
Me
Me
7a R=Me; 7b R=Et
6a R=Me; 6b R=Et
part B
part A
Figure 4 Structure of Dimers 6a-7a and 6b-7b.
After purification the reaction mixture, an isomer mixture of 6a and 7a was obtained in a 68% yield
(6a:7a ≈ 6:4). The separation of dimer isomers by a second chromatography afforded 6a (30%) and 7a
(12%) in pure form. The ethyl ester analogs (6b, 7b) were obtained in the same way, however in lower
yields (6b + 7b: 52%; ratio: ≈ 7:3; 6b: 23%, 7b: 15%). The bulkier ethyl ester group did not substantially
affect the reaction sequence leading to dimer isomers. On the other hand, we found that the presence of an
ester group is a requirement for the dimerization. The solution of 1a in MeOH was refluxed for 2-3 days
but only the earlier described epimerization at C8 was observed, and no traces of dimers could be
detected. Alcohols as solvent were found essential for the formation of dimers. In all other solvents
examined (CHCl3, CDCl3, CD2Cl2, DMSO-d6, acetone, CDCl3 + MeOD-d4: 4:2), 1b gave only epimer
mixtures. It is worth mentioning here that the position of equilibrium significantly depends on the
character of the ester group. While a 3:2 ratio of equilibrium between 1b and 2b can be measured by the
NMR method after keeping the solution in CDCl3 for two weeks at room temperature, at the same time
the portion of 2c in the 1c ' 2c equilibration was found to be as small as 7%. A few entirely other types
dimers in the ergot alkaloids field have been presented in the literature.14
εAll reactions were protected against the air, light and humidity applying the usual experimental protocol. The
reaction was repeated about 6-7 times on different batches of MeOH. Compound 1b was prepared several times
and applied as starting material for the dimerization. The result was always the same. The reaction was also
performed in a NMR tube to verify that the dimers did not form during the isolation.

HETEROCYCLES, Vol. 71, No. 5, 2007
1079
2) Structure elucidation of dimers - a) NMR-methods: The formation of the dimer structures was
assumed from the presence of a quaternary carbon at ca. 53 ppm and by the absence of vicinal couplings
on the H-7 proton resonances in the 13C and 1H NMR spectra of compounds 6a, 6b, 7a, and 7b.
The NMR spectra displayed single sets of spectral parameters for 6a and 6b, which correspond to
symmetrical dimer forms while for 7a and 7b we observed two chemically non-equivalent sets of 1H and
13C resonances. Consequently, asymmetric dimer structures were assigned to these compounds, where the
two halves differ in the configuration at C-8. Decoupling and heterocorrelated NMR experiments helped
to assign each set of protons and carbons of the two halves. However, the similarity of the spectral
parameters did not afford a direct configurational assignment of the two halves.
The spectral parameters of one half of the asymmetric dimers (part B of 7a and 7b) are quite similar to
those of the symmetric dimers 6a and 6b, while in the other half (part A) these parameters differ from
their values of part B. It can be assumed that, similarly to diethyl amide derivatives of 1a and 2a, the
C8-epimers exist in different conformations (Figure 5). 15
The upfield shift of C-7 and C-4 carbons of part
NOE
A, relative to their values in part B, can be
H
H
H
5
4
C11
C11
C3
H
C3
H
accounted for the γ-gauche interaction existing in
the conformer II. The differences of the H-7
geminal couplings and some of the long-range
couplings in the two halves of the above dimers
can also be connected with slight conformational
differences of ring D. However, the contribution
of conformer II can not be significant, since no
NOE’s were found between H-4α and H-7α
protons.
X
H
H
..
(β)
MeN
Me
N
H
7
X
H
H
Y
..
NOE
8
H
Y
H
conformer-II
conformer-I
moiety of 1a: X=CO2R, Y=part B
moiety of 2a: X=part A, Y=CO2R
Figure 5 Relationships of Dimers in Ring D.
In the absence of vicinal proton-proton couplings in ring D, the alternative approach in assigning the C-8
configuration is to use the stereodependence of the vicinal carbon-proton coupling values. As follows
3
from the Karplus-type dihedral angle relationship, maximum JCH values are expected for 0o and 180o
while minimum (zero) values for the near 90o dihedral angles.16 In monomers 1b and 1c, according to
molecular models, the dihedral angle between the carboxy carbon and H-9 proton is ca. 20o.
Consequently, this carbon correlated relatively strongly with the H-9 proton in the long-range
heterocorrelated experiments. Similar long-range hetero-correlations were found between the H-9 proton
and the carboxy carbon in part A of the asymmetric dimers 7a and 7b, while no such correlations were
obtained between the in symmetric dimers 6a and 6b and in part B of the asymmetric dimers 7a and 7b.

1080
HETEROCYCLES, Vol. 71, No. 5, 2007
In the symmetric dimers, and in part B of the asymmetric dimers, on the basis of the observed geometric
dependence of 3J(H9,COO), it can be concluded that the orientation of the ester groups corresponds to that of
the isolysergic acid while part A of the asymmetric dimers are displaced as in the lysergic acid. The
preferred orientation about the C8–C8’ bond in 7a can also be established by the NOESY spectrum. The
NOE connectivities of some ring D protons between the two halves reveal that 7a, similarly to 6a, exists
preferably in an extended form. In particular, the H-9 proton of part A showed NOE correlations both
with H-7α and H-7β protons of part B, while the H-9 proton of part B correlated only with the H-7α proton
of part A. This difference is the consequence of the different C-8 stereochemistry of the two halves,
where, according to the stereomodels, the H-7β proton of part A can not get into the vicinity of the part B
protons. The relative arrangement of the two halves in 7a is corroborated by the long-range connectivities
observed between the protons and carbons of the two D rings. The molecular models of the suggested
stereochemistry reflect a ca. 20o dihedral angle between H-9(B) and C-8(A) and a ca. 90o dihedral angle
between H-9(A) and C-8(B). Accordingly, intense heterocorrelation was detected in the HMBC and
INEPT-long-range spectra between H-9(B) and C-8(A), while no such correlation existed between
H-9(A) and C-8(B).
b) X-Ray analysis: Tiny crystals of 6a, suitable for single crystal X-ray analysis,17 were obtained from a
mixture of diisopropylether and acetone (7/3); despite several attempts we could not obtain single crystals
of any of the asymmetric isomers (namely 6b and 7b) (Figure 6).
C38
C39
37
38 O
H
C13
26N
C2 4
C2 7
C23
C12
C14
C15
C22
O
C9
C25
C37
C8
C28
21NH
C16
C1 0
C31
C11
C30
C32
HN1
C36
C29
C17
C35
C5
C7
C2
C34
C3
N6
O17
C33
C4
O
H
18
C18
C19
Figure 6 The ORTEP plot and crystallographic numbering scheme for the first of the independent
molecules of 6a (second molecule labelled with corresponding 40+n labels).
The obtained dimer has a unique structure. The search in the CSD18 (Cambridge Structural Database) for
a similar molecular structure, defined as two generic six-membered rings (with C, N or O atoms) joined
together by an acyclic C(sp3)-C(sp3) bridgehead bond, indicates that none of such structure has been
previously reported with ester functions attached to the bridgehead atoms. Also C(sp3)-C(sp3) bonds
between tertiary (bonded to either C, N or O atoms) carbons substituted with an ester function are not
very common in the CSD. In addition to the unique connection of the dimer parts, two other interesting
structural features occur. The C(sp3)-C(sp3) bridgehead bond is very long and the molecule consists of

HETEROCYCLES, Vol. 71, No. 5, 2007
1081
two symmetry equivalent parts (viz. 2 x 2b). The crystallographic results for 6a confirm the presence of
symmetry within the dimer. The asymmetric unit of the unit cell (space group P212121) contains two
independent dimers having slightly different torsion angles between the ester groups, yet the same
intramolecular symmetry persists. In each of the two independent dimers one half of the molecule is
symmetry-equivalent with the other half, to which it is related by a non-crystallographic twofold axis. The
twofold symmetry is unaffected by the conformational freedom around the bridgehead C−C bond of the
dimer allowing the two independent molecules to adopt slightly different torsion angles around the
bridgehead C−C bond. In this respect, crystallographic results confirm the NMR evidence. In addition to
the intriguing molecular structure, the bridgehead C−C bond of the dimer manifests a striking feature of
the molecular structure of 6a and of the whole class of dimers described in this work. The two large parts
of the dimer are structurally surprisingly unaffected by the close proximity of the ester groups. Even the
very short intramolecular cis-conformational carbonyl carbon to carbonyl carbon [C(17)–C(37) and
C(57)−C(77)] distances, 2.866(8) and 2.939(9) Å respectively, do not alter the normal sp3 C−C and C−N
bond distances and angles around the bridgehead carbons [C(8), C(28), C(48) and C(68)]. As the sum of
van der Waals radii of carbon is 3.4 Å, the measured contact distances are 0.534/0.461 Å shorter. The
bridgehead bond lengths, measured from the two independent dimers, are 1.603(9) Å and 1.612(9) Å
[C(8)−C(28) and C(48)–C(68)], respectively. These apparently high values are not surprising considering
that 86 examples of acyclic C(sp3)−C(sp3) bonds between tertiary (bonded to carbon atom) carbons longer
than 1.6 Å can be found in the CSD. As mentioned before, the value of the torsion angle between the ester
functions across the bridgehead C−C bond is different in the two crystallographically independent
molecules, being 51.2(7)° for the first molecule [C(17)−C(8) −C(28)−C(37)] and 59.3(7)° for the second
[C(57)−C(48)−C(68)−C(77)].
3) A plausible interpretation of reaction mechanism leading to dimers – A few similar types of
dimerization by oxidative coupling from more simple compounds in the presence of metal ions, can be
found in the literature.19 Furthermore, we have performed our original reaction with a few modifications
to obtain details from the character of the dimerization. The solution of 1b in MeOH was treated as
o
described above (stirring at 40 C) in the presence of a radical scavenger (3,5-di-tert-butylphenol, 1
equivalent). The scavenger completely altered the original result; no dimers were formed. This experience
shows that the dimerization may be a radical process. In another run, Na2S2O4 also inhibited the
dimerization. This indicates that the traces of air-oxygen also may play a role in the reaction pathway. In
both cases, the epimer mixture of 1b and 2b could be recuperated from the reaction mixture. When the
o
mixture was directly irradiated with white light the dimerization took place perfectly for 20 h at 40 C.
This indicates that the light accelerated the dimerization. In the final experiment radical initiator 10 mol%
of AIBN was added to the reaction mixture but it did not change the rate of the dimerization at all.

1082
HETEROCYCLES, Vol. 71, No. 5, 2007
In order to provide further support for the radical mechanism, we examined the electronic structure and
thermodynamical stability of the radical 1b• assumed to act as an intermediate in the dimerization
reaction, by applying density functional theory at the B3LYP/6-31G* level.20 As expected, the radical has
a nearly planar equilibrium structure, which allows a high degree of delocalization of the unpaired
electron. In Figure 7, we depicted the isodensity surface of the singly occupied molecular orbital of 1b•
along with the spin density of the radical, which both indicate that the electron delocalization is extended
almost to the entire molecular frame including the allylic unit, the COOMe group, and the two aromatic
rings as well.
a)
b)
Figure 7 Surface plots of a) singly occupied MO and b) spin density of radical 1b• (isodensity values
are 0.05 and 0.01 e/bohr3, respectively).
R•
D298 (calc) D298 (exp)
•
110.6
102.7
109.1
104.8
93.4
111
107
106
105
95.8
88
C6H5
•
CF3
•
•
CH2H3
•
CH3
•
CMe3
88.2
PhCH2
HCO•
87.0
87
•
84.9
86
CH2=CH-CH2
1b•
68.2
Mean absolute deviation: 1.4 kcal/mol
Table 1 Calculated and experimental C-H bond dissociation enthalpies (in kcal/mol) for a set of R-H
compounds. Experimental data are taken from Ref. 21.

HETEROCYCLES, Vol. 71, No. 5, 2007
1083
These results point to the enhanced stability of the 1b• radical; therefore, we estimated the homolytic
bond dissociation enthalpy (BDE) of the C−H bond in 1b. The BDE values of R−H compounds are
generally considered to provide a reasonable measure of the relative stability of free radicals,21 including
•
•
stable radicals such as CH2=CHCH2 or PhCH2 , the experimental data range between 80-90 kcal/mol.
Our calculations (Table 1) predict a significantly lower C−H bond dissociation enthalpy for 1b, indicating
that the associated radical might be unusually stable and probably long lived, allowing for dimerization.
Although the calculated BDE value falls into the energy range of UV-visible light, it is quite unlikely that
radical 1b• is produced via direct photolytic excitation. We anticipate that the reaction pathways depicted
in Figure 8 may represent a plausible mechanism.
On the basis of our findings that two isomers are produced in the reaction, we argue that an anionic
species (i) formed via deprotonation of 1b is a common intermediate in the epimerization (route a) and
dimerization pathways. The former reaction route can proceed through the enolic form (ii) of 1b, whereas
the radical formation likely involves an oxidation process leading either directly to radical (1b• = iv) or
involving a peroxo radical intermediate (iii). The relative rates along these reaction channels might be
rather sensitive to the stability of the anionic intermediate, which should be closely related to the solvent
medium. Our observation that the dimerization only takes place in protic solvents suggests that the
presence of an H-bonded network might be a key factor for the preference of routes (b) or (c). Further
theoretical and also kinetic studies were required to gain more insight into the mechanistic details.
CO2Me
O
OMe
N
CO2Me
H
N
H
Me
epimerization
1b
as base
+
H
N
Me
a
( )
Me
1b
1b
2b
and
ii
( )
i
( )
O2
c
( )
b
( )
CO2Me
O O
CO2Me
H
H
N
N
d imer ization
1b
Me
Me
6a 7a
+
+
- O2
iii
(
)
iv
(
1b
) =
Figure 8 A plausible reaction sequence leading to dimers.

1084
HETEROCYCLES, Vol. 71, No. 5, 2007
2.5
2.0
1.5
1.0
0.5
0.0
Figure 9 (+)-Lysergic acid methyl ester-induced
• −
formation of O2 in MeOH. The time-course of
ethidium fluorescence intensity changes (%) was
recorded logarithmically, comprising 1000 data
point acquired automatically. The concentration of
(+)-1b: 0.1 mM.
0.1
1
10
Time (s)
To identify the primary reactive oxygen species requires a method with sufficient temporal resolution.
Therefore measurements were conducted in the 0.05-10 s range of time applying the techniques of fast
chemical kinetics in combination with fluorescence detection22 (Figure 9). Hydroethidine, a blue
fluorescent dye until oxidized to the red fluorescent ethidium was used to selectively monitor superoxide
radical anion (O2• −) formation.23 These measurements helped us to choose pathway c as the more
probable transformations leading to the dimers.
CONCLUSIONS
We have shown that the simple derivatives of (+)-lysergic acid, (+)-lysergic acid methyl or ethyl ester,
afforded dimer isomers under mild reaction conditions in alcohols. The dimerization can be regarded as
fourth special reaction in this field, beside the well-known discovered particular properties of ergot
derivatives. An observation at the early stage − mutarotation of 1b in methanol − has taken a final and
correct explanation. The isomers of the dimer were separated and their structures were disclosed by
detailed NMR examination and X-ray analysis. Our proposed reaction pathway leading to the dimers
attempts to delineate the C8-epimerization and dimerization via a common intermediate, a carbanion
species that is formed in the first step. In the following transformations unusually stable radical could be
formed by catalysis of traces of air-oxygen, and their recombination can give the dimer product and a
superoxide radical anion. Experimental data on (+)-lysergic acid methyl ester-induced superoxide radical
anion formation provided evidence for the proposed mechanism. The feasibilities of the C8-epimerization
and dimerization routes have been related to solvent effects.
EXPERIMENTAL
General: All reactions were carried out under an atmosphere of nitrogen or argon in dried glassware.
Methanol and ethanol were dried Grignard-reaction and distillation or with a simple distillation.
Chloroform and dichloromethane were distilled from phosphorous pentoxide. Acetone was distilled from
potassium permanganate. Column chromatography was performed with Merck silica gel (0.040-0.063

HETEROCYCLES, Vol. 71, No. 5, 2007
1085
mm). Analytical thin-layer chromatography was performed with commercial silica gel plates (Merck or
Polygram of Macherey-Nagel, plastic sheets), and the plots were visualized under UV light or developed
by iodine atmosphere and immersion in a solution of o-toluidine. Melting points were obtained on a Carl
Zeiss apparatus equipped with microscope and are uncorrected. Elemental analyses (C, H, N) were
carried out by Vario EL III (Elementar Analysen System Gmbh) automatic microanalyzer. IR data were
recorded on potassium bromide plates on a Nicolet-7795 FT-IR spectrometer. Optical rotations were
measured on an AA-10R automatic polarimeter (Optical Activity Ltd.) using 1.0 dm cells and the sodium
D line (589 nm) at ambient temperature. MS spectra were run on an AEI-MS-902 (70 eV; direct
insertion) and on a Kratos-MS-902 mass spectrometer. FAB-MS spectra were measured on a ZAB 2SEQ
spectrometer. High resolution NMR measurements were carried out on a Varian Unity Inova 400 and
Varian NMR System 600 spectrometers, equipped with 5 mm indirect detection probes, operating at
399.9 and 599.9 MHz for 1H and at 100.5 and 150.9 MHz for 13C nuclei. All spectra were acquired at 30
o
or 35 C. The proton chemical shifts were measured relative to tetramethylsilane internal standard, and
the carbon chemical shifts were reported using solvents as internal standards (CDCl3 at 77.2 ppm, Py-d5
at 150.0 ppm). The structural and spectroscopic assignments were made by the combined use of 1D and
2D experiments, such as HSQC and HMBC heterocorrelation techniques, using standard Varian software.
One-dimensional long-range 1H-13C correlations were obtained in a sequence of selective
INEPT-long-range experiments with delay values optimized for 3, 6, or 8 Hz couplings. Phase sensitive
NOESY spectrum with 2048 x 128 blocks has been measured using 0.5 s mixing time. Data processing
was performed using three-fold linear prediction. Rutin-like 1H and 13C measurements were carried out on
a Varian Gemini 200 spectrometer.
1) Synthesis - Ester derivatives 1b and 1c were prepared applying the Sandoz’s procedure10 (HCl/alcohol)
but using 1a as starting material instead of (±)-1a. Physical constants and spectroscopic data of 1b were
in full agreement with the reported in our earlier publication.13 Ester 1c can be regarded as a new
compound without physical and spectroscopic characterization. Similarly, naphthalene 5a and 5b has
been known as racemic derivatives without any spectroscopic data. We have followed the official IUPAC
rules for the name of dimers but for the NMR assignment the traditional ergoline-numbering was applied
to a better comparison of dimers and monomers.
(±)-Isolysergic acid [(±)-2a] (Racemization of 1a): A suspension of 1a (3.0 g, 11.2 mmol) in aqueous
KOH (1.5%, 50 mL) was passed into a steely sealed-tube and closed. The sealed-tube was putted into an
oil-bath heated to 150 oC and kept steady for 4 h. After cooling to rt, the solution was put into a flask, the
sealed-tube was washed with hot water (2 x 25 mL). The pH of the solution was adjusted to about 6 with
aqueous HCl solution (20%, 25-30 mL). The precipitated dark crystals were filtered off, washed with

1086
HETEROCYCLES, Vol. 71, No. 5, 2007
water to yield (±)-1a as grey semisolid crystals (2.34 g, 78%), which proved to be optically inactive. An
analytical sample (0.5 g) was recrystallized from hot water with charcoal to afford (±)-2a (50 mg, 10%)
beside the well known (±)-1a. Acid (±)-2a: mp 230-255 oC (tawny colour crystals, decomp.); TLC: Rf of
(±)-2a: 0.21 (CHCl3 + MeOH + NH4OH, 5/5/0.1 mL); [TLC: Rf of (±)-1a: 0.28 (CHCl3 + MeOH +
NH4OH, 5/5/0.1 mL)]. IR νmax (KBr): 3415, 3117, 3055, 1600, 1450, 1378, 1344 cm-1. HRMS (EI-MS):
1
calcd for: C16H16N2O2; 268.1212, found 268.1216. H NMR (Pyr-d5, 400 MHz): δ 2.52 (s, 3 H, NCH3),
2.80 (dd, J = 11.4, 4.2 Hz, 1 H, H-7β), 2.90 (m, J = 14.4, 11.5, 1.6 Hz, 1 H, H-4α) 3.33 (m, 1 H, H-5),
3.52 (m, 1 H, H-8), 3.54 (dd, J = 14.4, 5.9 Hz, 1 H, H-4β), 3.60 (dd, J = 11.4, 2.4 Hz, 1 H, H-7α), 7.01 (dd,
J = 5.9, 2.2 Hz, 1 H, H-9), 7.20 (t, J = 1.6 Hz, 1 H, H-2), 7.33 (dd, J = 7.5, 7.3 Hz, 1 H, H-13), 7.40–7.45
(m, 1 H, H-12, H-14), 11.65 (d, J = 1.6 Hz, 1 H, NH) ppm. 13C NMR (Py-d5, 100 MHz): δ 27.7 (C-4),
42.4 (C-8), 43.5 (NCH3), 54.5 (C-7), 63.6 (C-5), 110.6 (C-3), 110.6 (C-12), 111.3 (C-14), 119,7 (C-9),
120.0 (C-2), 123.6 (C-13), 127.5 (C-16), 129.2 (C-11), 135.4 (C-15), 137.1 (C-10), 175.1 (CO2H) ppm.
(+)-Lysergic acid methyl ester (1b): Ester 1b was prepared as described in the below examples. For the
crystallization dry benzene was applied. The NMR spectra have been given to a better comparison of
dimer subunits. 1H NMR (CDCl3, 400 MHz): δ 2.59 (s, 3 H, NCH3), 2.68 (m, J = 14.5, 11.5, 1.8 Hz, 1 H,
H-4α), 2.69 (dd, J = 11.4, 10.7 Hz, 1H, H-7β), 3.17 (m, J = 11.5, 5.6, 3.5, 2.1 Hz, 1 H, H-5), 3.26 (m, J =
11.4, 5.3, 1.2 Hz, 1 H, H-7α), 3.50 (dd, J = 14.5, 5.6 Hz, 1 H, H-4β), 3.71 (m, J = 10.7, 5.3, 3.5, 2.3 Hz, 1
H, H-8), 3.77 (s, 3 H, OCH3), 6.59 (m, J = 2.3, 2.1, 1.2 Hz, 1 H, H-9), 6.87 (t, J = 1.8 Hz, 1 H, H-2),
7.15–7.23 (m, 3 H, H-12, H-13, H-14), 8.12 (d, J = 1.8 Hz, 1 H, NH) ppm. 13C NMR (CDCl3, 100 MHz):
δ 27.4 (C-4), 42.4 (C-8), 43.9 (NCH3), 52.2 (OCH3), 55.1 (C-7), 63.1 (C-5), 109.9 (C-12), 111.0 (C-3),
112.7 (C-14), 118.2 (C-9), 118.5 (C-2), 123.5 (C-13), 126.5 (C-16), 128.3 (C-11), 134.2 (C-15), 136.4
(C-10), 173.0 (CO2) ppm.
(+)-Lysergic acid ethyl ester (1c): A suspension of 1a (3.0 g, 11.2 mmol) in dry EtOH (100 mL) was
o
cooled with an ice-bath to −5 C. Dry HCl gas was bubbled through the suspension, which gradually
became to a dark solution for about 2-3 h while the temperature was kept on 0-5 oC. (The progress of the
esterification was checked with TLC; eluent: a) CHCl3+MeOH; 10/1; b) CHCl3 + MeOH + NH4OH,
5/5/0.1). The reaction mixture was poured into a mixture of CHCl3 and ice-water (300 + 100 mL), then
the pH was adjusted about 8 with NH4OH solution (25%, 50 mL). The phases were separated and the
organic layer was washed with water (2 x 50 mL), dried. The filtrate was evaporated under reduced
pressure to yield brown oil (3.9 g). The oil obtained was crystallized (cyclohexane + AcOEt + acetone,
20/0.5/0.5 mL) to afford the title compound as tawny colour crystals (2.41 g, 73%); Rf 0.8 (CHCl3 +
MeOH, 10/1); mp 119-121 oC; [α]D +62o (c 0.32, CHCl3). IR νmax (KBr): 3398, 2959, 1728, 1447, 1300,
1183 cm-1. HRMS (FAB-MS; matrix: glycerol): calcd for [M+H]+: 297.1603, found 297.1600; comp.:

HETEROCYCLES, Vol. 71, No. 5, 2007
1087
1
M=296; C18H20N2O2. H NMR (CDCl3, 400 MHz): δ 1.33 (t, J = 7.1 Hz, 3 H, CH2CH3), 2.59 (s, 3 H,
NCH3), 2.68 (dd, J = 11.4, 10.7 Hz, 1 H, H-7β), 2.69 (m, J = 14.5, 11.6, 1.7 Hz, 1 H, H-4α), 3.16 (m, J =
11.6, 5.6, 3.4, 2.0 Hz, 1 H, H-5), 3.27 (m, J = 11.4, 5.4, 1.2 Hz, 1 H, H-7α), 3.50 (dd, J = 14.5, 5.6 Hz, 1
H, H-4β), 3.70 (m, J = 10.7, 5.4, 3.4, 2.3 Hz, 1 H, H-8), 4.22 (q, J = 7.1 Hz, 2 H, OCH2), 6.60 (m, J = 2.3,
2.0, 1.2 Hz, 1 H, H-9), 6.88 (dd, J = 1.8, 1.7 Hz, 1 H, H-2), 7.15–7.23 (m, 3 H, H-12, H-13, H-14), 8.13
(d, J = 1.8 Hz, 1 H, NH) ppm. 13CNMR (CDCl3, 100 MHz): δ 14.5 (CH3), 27.4 (C-4), 42.5 (C-8), 43.9
(NCH3), 55.1 (C-7), 61.1 (OCH2), 63.2 (C-5), 109.9 (C-12), 111.0 (C-3), 112.7 (C-14), 118.5 (C-9), 118.5
(C-2), 123.5 (C-13), 126.4 (C-16), 128.3 (C-11), 134.2 (C-15), 136.3 (C-10), 172.6 (CO2) ppm. Anal.
Calcd for C18H20N2O2: C, 72.95, H, 6.80, N, 9.45. Found: C, 72.90, H, 6.87, N, 9.47.
(+)-4-Acetyl-7-methyl-4,5,7,8,9,10-hexahydroindolo[4,3-fg]quinoline 9-carboxylic acid ethyl ester
(5c; Indole-Naphthalene Isomerization): Acid 1a (0.75 g, 2.79 mmol) was transformed into ethyl ester as
described above but the reaction mixture was transformed further in the following way. The ethanolic
solution saturated with HCl was transfused into a sealed-tube, closed, and it was putted into an oil-bath
o
heated to 100 C and kept steady for 0.5 h. After cooling to rt, the reaction mixture was poured to a
mixture of CHCl3, ice water and saturated Na2CO3 solution (300 + 100 + 60 mL) while stirring and
cooling the mixture. In the second step acetic anhydride (7.5 mL, 79 mmol) and potassium carbonate
(15.0 g, 108 mmol) were added to the mixture and stirred further for 0.5 h. The phases were separated and
the organic layer was washed with water (3 x 75 mL), dried. The filtrate was evaporated under reduced
pressure to yield a black oil (787 mg) which was purified by column chromatography (hexane + AcOEt +
CHCl3; 300/300/6.0, then 400/400/12) to yield the title compound as black colour crystals (530 mg,
56%); Rf 0.42 (CHCl3 + acetone, 20/1). An analytical sample was recrystallized from a mixture of
benzene + diisopropyl ether + acetone applying charcoal; mp 178-180 oC (white crystals); [α]D +61.8o (c
1.99, CHCl3). IR νmax (KBr): 1734, 1661, 1659, 1628, 1600 1415 cm-1. HRMS (EI-MS): calcd for
C20H22N2O3; 338.1630, found 338.1632. 1H NMR (CDCl3, 400 MHz; two sets of amide rotamers): δ 1.30
(t, J = 7.1 Hz, 3 H, CH2Me), 2.24, 2.43 (2 s, 3 H, COMe), 2.98, 3.02 (2 s, 3 H, NMe), 3.04–3.14, 3.25 (2
m, 3 H, H-9, H-10α, H-10β), 3.36 (dd, J = 11.4, 9.8 Hz, 1 H, H-8β), 3.49 (m, J = 11.4, 3.8, 1.8 Hz, 1 H,
H-8α), 4.22 (q, J = 7.1 Hz, 2 H, OCH2), 4.93 + 5.11 (2 s, 2 H, H-5), 6.69 + 6.80 (2 s,1 H,H-6), 6.68 + 7.63
(2 dd, J = 7.3, 0.5 Hz, 1 H, H-3), 7.18 + 7.20 (2 dd, J = 8.4, 0.5 Hz, 1 H, H-1), 7.31 + 7.34 (2 dd, J = 8.4,
7.3 Hz, 1 H, H-2) ppm.13C NMR (CDCl3, 100 MHz): δ 14.5 (CH2Me), 24.0 + 25.0 (COMe), 26.0 (C-10),
38.2 + 38.3 (C-9), 40.6 + 40.8 (NMe), 52.9 (C-8), 54.4 + 55.2 (C-5), 61.0 (OCH2), 103.6 + 107.2 (C-3),
105.4 + 105.8 (C-6), 109.6 + 110.3 (C-10a), 114.2 + 114.4 (C-1), 123.9 + 124.3 (C-3b), 129.1 + 130.0
(C-2), 130.8 + 131.4 (C-10b), 134.2 + 134.4 (C-5a), 142.5-143.4 (C-6a), 145.7 + 146.3 (C-3a), 168.6 +
168.8 (NCO), 173.8 (CO2) ppm. Anal. Calcd for C20H22N2O3: C, 70.99, H, 6.55, N, 8.28. Found: C, 70.90,

1088
HETEROCYCLES, Vol. 71, No. 5, 2007
H, 6.48, N, 8.17.
(+)-4-Acetyl-7-methyl-4,5,7,8,9,10-hexahydroindolo[4,3-fg]quinoline-9-carboxylic acid methyl ester
(5d): Naphthalene derivatives 5d was prepared as described above. Rf 0.4 (CHCl3 + acetone, 20/1); mp
155-157 oC (from diisopropyl ether); [α]D +16o (c 0.25, CHCl3). IR νmax (KBr): 1734, 1661, 1659, 1628,
1600 1415; 1739, 1666, 1628, 1600, 1500, 1410 cm-1. MS (EI-MS): 324 (M+, 100), 309 (5, M-Me), 293
(4), 282 (13), 263 (12), 221 (23). 1H NMR (CDCl3, 400 MHz): 2.29, 2.51 (2 s, 3 H, COMe), 3.00, 3.04 (2
s, 3 H, NMe), 3.03–3.15 + 3.27 (2 m, 3 H, H-9, H-10α, H-10β), 3.37 (dd, J = 11.5, 9.8 Hz, 1 H, H-8β),
3.48 (m, J = 11.5, 3.7, 1.6 Hz, 1 H, H-8α), 3.76 (s, 3 H, OMe), 5.05 + 5.15 (2 s, 2 H, H-5), 6.72 + 7.65 (2
dd, J = 7.3, 0.5 Hz, 1 H, H-3), 6.78 + 6.85 (2 s, 1 H, H-6), 7.20 + 7.21 (2 dd, J = 8.4, 0.5 Hz, 1 H, H-1),
7.33 + 7.36 (2 dd, J = 8.4, 7.3 Hz, 1 H, H-2) ppm. 13C NMR (CDCl3, 100 MHz): δ 24.1 + 25.0 (COMe),
26.0 (C-10), 38.2 + 38.3 (C-9), 40.6 + 40.7 (NMe), 52.2 (OMe), 52.9 + 53.0 (C-8), 54.4 + 55.3 (C-5),
103.6 + 107.2 (C-2), 105.4 + 105.8 (C-4), 109.5 + 110.2 (C-10), 114.2 + 114.4 (C-12), 124.0 + 124.4
(C-16), 129.2 + 130.1 (C-2), 130.9 + 131.5 (C-10b), 134.2 + 134.4 (C-5a), 142.6 + 143.5 (C-6a), 146.0 +
146.4 (C-3a), 168.6 + 168.8 (NCO), 174.2 (CO2) ppm. Anal. Calcd for C19H20N2O3: C, 70.35, H, 6.21, N,
8.64. Found: C, 70.41, H, 6.19, N, 8.57.
(6aR,6’aR,9R,9’R)-Dimethyl-7,7’-dimethyl-4,4’,6,6a,6’,6a’,7,7’,8,8’,9,9’-dodecahydro-9,9’-biindolo[
4,3-fg]quinoline-9,9’-dicarboxylate (6a) and
(6aR,6’aR,9R,9’S)-dimethyl-7,7’-dimethyl-4,4’,6,6a,6’,6a’,7,7’,8,8’,9,9’-dodecahydro-9,9’-biindolo-
[4,3-fg]quinoline-9,9’-dicarboxylate (7a): Ester 1b (1.5 g, 5.3 mmol) was dissolved in dry MeOH (150
o
mL) under nitrogen atmosphere and the solution was stirred for 48 h at 40 C then rt for 24 h. After
evaporation of the solvent, the residue was purified by chromatography (eluent: CHCl3+MeOH, 20/1; Rf
of 6a+7a > 1b but differences could not be observed between 6a and 7a in this solvent mixture). The
isomer mixture (1.01 g, 68%; ratio of 6a:7a = 63/37%, determined by NMR integrals) was separated with
a second chromatography (eluent: hexane +AcOEt + MeOH, 6/4/0.5; Rf of 7a > 6a).
o
Dimer 6a (443 mg, 30%), Rf 0.385 (hexane + AcOEt + MeOH, 5/5/1); mp 173-177 C (nearly white
crystals from a mixture of diisopropyl ether + acetone, 7/3); [α]D +170o (c 0.5, CHCl3). IR νmax (KBr):
3446, 3399, 2949, 1733, 1606, 1445, 1431, 1236 cm-1. HRMS (FAB-MS; matrix: NOBA): calcd for
1
C34H34N4O4 562.2580, found 562.2610; [M+H]+: 563.2678; ESIpoz (turbo spray) [M+H]+: 563.4. H
NMR (CDCl3, 400 MHz; to a better comparison, the chemical shifts of dimers are given applying the
ergoline-numbering): δ 2.53 (s, 2 x 3 H, NCH3), 2.63 (m, J = 14.6, 11.3, 1.6 Hz, 2 x 1 H, H-4α), 2.79 (d, J
= 11.3 Hz, 2 x 1 H, H-7β), 3.09 (m, J = 11.3, 5.7, 1.9 Hz, 2 x 1 H, H-5), 3.46 (dd, J = 14.6, 5.7 Hz, 2 x 1
H, H-4β), 3.49 (dd, J = 11.3, 1.2 Hz, 2 x 1 H, H-7α), 3.69 (s, 2 x 3 H, OCH3), 6.82 (dd, J = 1.9, 1.2 Hz, 2 x

HETEROCYCLES, Vol. 71, No. 5, 2007
1089
1 H, H-9), 6.88 (dd, J = 1.6, 1.8 Hz, 2 x 1 H, H-2), 7.18 (dd, J = 6.7, 6.9 Hz, 2 x 1 H, H-13), 7.22 (dd, J =
6.7, 1.2 Hz, 2 x 1 H, H-12), 7.25 (dd, J = 6.9, 1.2 Hz, 2 x 1 H, H-14), 7.99 (d, J = 1.8 Hz, 2 x 1 H, NH)
ppm. 13C NMR (CDCl3, 100 MHz): δ 27.3 (2 x C-4), 43.9 (2 x NCH3), 52.5 (2 x OCH3), 53.1 (2 x C-8),
57.2 (2 x C-7), 63.4 (2 x C-5), 110.0 (2 x C-12), 111.2 (2 x C-3), 112.7 (2 x C-14), 118.6 (2 x C-2), 119.9
(2 x C-9), 123.6 (2 x C-13), 126.7 (2 x C-16), 128.3 (2 x C-11), 134.2 (2 x C-15), 136.5 (2 x C-10), 173.3
(2 x CO2) ppm. Anal. Calcd for C34H34N4O4: C, 72.58, H, 6.09, N, 9.96. Found: C, 72.61, H, 6.14, N,
9.91.
o
Dimer 7a (176 mg, 12%), Rf 0.414 (hexane + AcOEt + MeOH, 5/5/1); mp 153-177 C (decomp.; light
brawn crystals; from a mixture of cyclohexane + AcOEt, 3/2); [α]D +190o (c 0.263, CHCl3). IR νmax
(KBr): 3414, 2950, 2848, 1728, 1604, 1447, 1230 cm-1. HRMS (FAB-MS; matrix: NOBA): calcd for
1
C34H34N4O4 562.2580, found 562.2594; [M+H]+: 563.2672; ESIpoz (turbo spray) [M+H]+: 563.4. H
NMR (CDCl3, 600 MHz,) part A: δ 2.49 (s 3 H, NCH3), 2.73 (m, J = 14.6, 11.4, 1.4 Hz, 1 H, H-4α), 2.83
(d, J = 12.4 Hz, 1 H, H-7β), 3.16 (m, J = 11.4, 5.4, 1.6 Hz, 1 H, H-5), 3.39 (dd, J = 14.6, 5.4 Hz, 1 H,
H-4β), 3.52 (dd, J = 12.4, 0.9 Hz, 1 H, H-7α), 3.61 (s, 3 H, OCH3), 6.67 (dd, J = 1.6, 0.9 Hz, 1 H, H-9),
6.87 (dd, J = 1.8, 1.4 Hz, 1 H, H-2), 7.14–7.25 (m, 3 H, H-12, H-13, H-14), 7.90 (d, J=1.8 Hz, 1 H, NH)
13
ppm. C NMR (CDCl3, 150.9 MHz): δ 26.1 (C-4), 43.0 (NCH3), 52.2 (OCH3), 53.9 (C-8), 55.2 (C-7),
62.2 (C-5), 110.1 (C-12), 111.4 (C-3), 112.6 (C-14), 118.6 (C-2 ), 119.1 (C-9), 123.4 (C-13), 126.8
(C-16), 128.8 (C-11), 134.2 (C-15), 138.6 (C-10), 172.8 (CO2) ppm, part B: 2.45 (s, 3 H, NCH3), 2.59 (m,
J = 14.3, 11.3, 1.7 Hz, 1 H, H-4α), 2.88 (d, J = 11.5 Hz, 1 H, H-7β), 3.00 (m, J = 11.3, 5.7, 1.7 Hz, 1 H,
H-5), 3.41 (dd, J = 14.3, 5.7 Hz, 1 H, H-4β), 3.62 (dd, J = 11.3, 1.2 Hz, 1 H, H-7α), 3.72 (s, 3 H, OCH3),
6.72 (dd, J = 1.7, 1.2 Hz, 1 H, H-9), 6.85 (dd, J = 1.7, 1.7 Hz, 1 H, H-2), 7.14–7.25 (m, 3 H, H-12, H-13,
H-14), 7.91 (d, J=1.7 Hz, 1H, NH) ppm. 13C NMR (CDCl3, 150.9 MHz): δ 27.2 (C-4), 43.8 (NCH3), 52.5
(OCH3), 53.5 (C-8), 58.7 (C-7), 63. 4 (C-5), 109.9 (C-12), 111.3 (C-3), 112.7 (C-14), 118.6 (C-2), 121.3
(C-9), 123.5 (C-13), 126.7 (C-16), 128.6 (C-11), 134.2 (C-15), 136.1 (C-10), 173.4 (CO2). Anal. Calcd
for C34H34N4O4: C, 72.58, H, 6.09, N, 9.96. Found: C, 72.49, H, 6.00, N, 9.89.
(6aR,6’aR,9R,9’R)-Diethyl-7,7’-dimethyl-4,4’,6,6a,6’,6a’,7,7’,8,8’,9,9’-dodecahydro-9,9’-biindolo[4,
3-fg]quinoline-9,9’-dicarboxylate (6b) and
(6aR,6’aR,9R,9’S)-diethyl-7,7’-dimethyl-4,4’,6,6a,6’,6a’,7,7’,8,8’,9,9’-dodecahydro-9,9’-biindolo[4,3
-fg]quinoline-9,9’-dicarboxylate (7b): Ester 1c (1.0 g, 3.37 mmol) was dissolved in dry EtOH under
nitrogen atmosphere and the soluton was stirred for 4 days at 50 oC. After evaporation of the solvent, the
residue was purified by chromatography (eluent: CHCl3 + acetone, 20/1; Rf of 6b+7b > 1c but differences
could not be observed between 6b and 7b in this solvent mixture). The isomer mixture (514 mg, 52%;
ratio of 6b:7b = 68/32%, determined by NMR integrals) was separated with a second chromatography

1090
HETEROCYCLES, Vol. 71, No. 5, 2007
(eluent: hexane + AcOEt + MeOH, 6/4/0.5; Rf of 7b > 6b).
Dimer 6b (227 mg, 23%), Rf 0.41 (hexane + AcOEt + MeOH, 5/5/1); mp 172-190 oC (pale brown crystals,
from cyclohexane, 5 mL); [α]D +59.8o (c 0.167, CHCl3). IR νmax (KBr): 3382, 1727, 1706, 1447, 1377,
1231 cm-1. HRMS (FAB-MS; matrix: glycerol): calcd for [M+H]+: 591.2971, found 591.2986; comp.:
M=590; C36H38N4O4. 1H NMR (CDCl3 + DMSO-d6, 400 MHz): δ 1.24 (t, J = 7.1 Hz, 2 x 3 H, CH3), 2.50
(s, 2 x 3H, NCH3), 2.62 (m, J = 14.6, 11.3, 1.7 Hz, 2 x 1 H, H-4α), 2.81 (d, J = 11.4 Hz, 2 x 1 H, H-7β),
3.09 (m, J = 11.3, 5.7, 1.9 Hz, 2 x 1 H, H-5), 3.44 (dd, J = 14.6, 5.7, 2 x 1 H, H-4β), 3.46 (dd, J = 11.4,
1.2, 2 x 1 H, H-7α), 4.12 + 4.22 (m, J = 10.8, 7.1 Hz, 2 x 2 H, OCH2), 6.82 (dd, J = 1.9, 1.2 Hz, 2 x 1 H,
H-9), 6.88 (dd, J = 1.7, 1.8 Hz, 2 x 1 H, H-2), 7.17 (dd, J = 6.9, 7.2 Hz, 2 x 1 H, H-13), 7.21 (dd, J = 6.9,
1.2 Hz, 2 x 1 H, H-12), 7.24 (dd, J = 7.2, 1.2 Hz, 2 x 1 H,H-14), 7.82 (d, J = 1.8 Hz, 2 x 1 H,NH) ppm.
13C NMR (CDCl3 + DMSO-d6, 100 MHz): δ 14.5 (2 x CH3), 27.4 (2 x C-4), 43.9 (2 x NCH3), 52.7 (2 x
C-8), 57.5 (2 x C-7), 61.2 (2 x OCH2), 63.5 (2 x C-5), 109.8 (2 x C-3), 110.4 (2 x C-12), 111.9 (2 x C-14),
119.3 (2 x C-2), 120.1 (2 x C-9), 122.8 (2 x C-13), 126.7 (2 x C-16), 128.0(2 x C-11), 134.4 (2 x C-15),
136.5 (2 x C-10), 172.6 (2 x CO2) ppm. Anal. Calcd for C36H38N4O4: C, 73.20, H, 6.48, N, 9.48. Found: C,
73.36, H, 6.51, N, 9.35.
Dimer 7b: (148 mg, 15%), Rf 0.49 (hexane + AcOEt + MeOH, 5/5/1 mL); mp 143-146 oC (decomp., pale
brown crystals; from cyclohexane + Et2O); [α]D +134.5o (c 0.22, CHCl3). IR νmax (KBr): 3411, 2978,
2849, 1724, 1656, 1604, 1448, 1226 cm-1. HRMS (FAB-MS; matrix: glycerol): calcd for [M+H]+:
591.2971, found 591.2978; comp.: M=590; C36H38N4O4. 1H NMR (CDCl3, 400 MHz) part A: δ 1.18 (t, J
= 7.1 Hz, 3 H, CH3), 2.49 (s, 3 H, NCH3), 2.74 (m, J = 14.4, 11.3, 1.6 Hz, 1 H, H-4α), 2.87 (d, J = 12.4
Hz, 1 H, H-7β), 3.20 (m, J = 11.3, 5.4, 1.8 Hz, 1 H, H-5), 3.35 (dd, J = 14.4, 5.4 Hz, 1 H, H-4β), 3.46 (dd,
J = 12.4, 0.7 Hz, 1 H, H-7α), 4.05–4.35 (m, 2 H, OCH2), 6.65 (dd, J = 1.8, 0.7 Hz, 1 H, H-9), 6.85 (dd, J
= 1.6, 1.7 Hz, 1 H, H-2), 7.13–7.24 (m, 3 H, H-12, H-13, H-14), 7.89 (d, J = 1.7 Hz, 1 H, NH) ppm. 13C
NMR (CDCl3, 100 MHz): δ 14.3 (CH3), 25.7 (C-4), 42.9 (NCH3), 53.7 (C-8), 55.0 (C-7), 61.5 (OCH2),
62.0 (C-5), 110.2 (C-12), 111.6 (C-3), 112.6 (C-14), 118.6 (C-2), 119.7 (C-9), 123.4 (C-13), 126.8 (C-16),
129.1 (C-11), 134.1 (C-15), 138.2 (C-10), 172.6 (CO2). 1H NMR (CDCl3, 400 MHz) part B: δ 1.30 (t, J =
7.1 Hz, 3 H, CH3), 2.44 (s, 3 H, NCH3), 2.56 (m, J = 14.2, 11.3, 1.6 Hz, 1 H, H-4α), 2.83 (d, J = 11.3 Hz,
1 H, H-7β), 2.98 (m, J = 11.3, 5.6, 1.9 Hz, 1 H, H-5), 3.37 (dd, J = 14.2, 5.6 Hz, 1 H, H-4β), 3.59 (dd, J =
11.3, 1.1 Hz, 1 H, H-7α), 4.05-4.35 (m, 2H, OCH2), 6.71 (dd, J = 1.6, 1.1 Hz, 1 H, H-9), 6.81 (dd, J = 1.6,
13
1.7 Hz, 1 H, H-2), 7.13-7.24 (m, 3 H, H-12, H-13, H-14), 7.93 (d, J = 1.7 Hz, 1 H, NH) ppm. C NMR
(CDCl3, 100 MHz): δ 14.4 (CH3), 27.3 (C-4), 43.8 (NCH3), 53.4 (C-8), 58.9 (C-7), 61.0 (OCH2), 63.5
(C-5), 109.7 (C-12), 111.5 (C-3), 112.7 (C-14), 118.4 (C-2), 121.8 (C-9), 123.4 (C-13), 126.7 (C-16),
128.9 (C-11), 134.2 (C-15), 135.9 (C-10), 173.0 (CO2). Anal. Calcd for C36H38N4O4: C, 73.20, H, 6.48, N,

HETEROCYCLES, Vol. 71, No. 5, 2007
1091
9.48. Found: C, 73.29, H, 6.53, N, 9.41.
2) X-ray analysis - X-ray single-crystal data collection was performed on a BRUKER-NONIUS
KAPPACCD diffractometer, geared with graphite-monochromated Mo Kα radiation (λ=0.71073 Å) and
provided of an APEXII area detector and an OXFORD CRYOSTREAM cryogenic device. Data
collection: COLLECT.24 Cell refinement and data reduction: EVALCCD.25 Structure solution and
refinement: SHELXTL.26
Table 2 Crystal data and structure refinement for 6a.
Chemical formula: C34H34N4O4
Formula weight: 562.66
Data collection method: Nonius KappaCCD
diffractometer
Temperature: 173(2) K
φ and ω scans
Goodness-of-fit on F2: 1.108
Largest and mean shift/su: 0.000 and 0.000
Largest diff. peak and hole: 0.70 and −0.45 e Å−3
Crystal system, space group: monoclinic, P21
Unit cell parameters:
θ range for data collection: 5.2 to 25.0°
Index ranges: h −12 to 12, k 0 to 29, l 0 to 14
Completeness to θ = 25.0° 97.0 %
Reflections collected: 5579
Independent reflections: 5579 (Rint = 0.0000)
Reflections with F2>2σ: 3846
a = 10.798(2) Å α = 90°
b = 24.899(5) Å β = 93.97(3)°
Absorption correction:
c = 11.878(2) Å γ = 90°
semi-empirical from equivalents
Min. and max. transmission: 0.9827 and 0.9956
R indices (all data): R1 = 0.1329, wR2 = 0.1566
Structure solution: direct methods
Refinement method:
Cell volume: 3185.8(11) Å3
Z: 4
Calculated density: 1.284 g/cm3
Absorption coefficient µ: 0.087 mm−1
F(000):1308
Full-matrix least-squares on F2
Crystal colour and size:
Weighting parameters a, b: 0.0031, 6.6266
colourless, 0.20 × 0.15 × 0.05 mm3
Reflections for cell refinement:
Crystal colour and size:
colourless, 0.20 × 0.15 × 0.05 mm3
Reflections for cell refinement:
2059 (θ range 2.5 to 27.5°)
Data / restraints / parameters: 5579 / 1465 / 820
Final
wR2=0.1305
R indices (all data): R1 = 0.1329, wR2 = 0.1566
R
indices [F2>2σ]:
R1=0.0802,
3) Computational details - Geometry optimizations for 1b and radical 1b• were carried out at the
B3LYP/6-31G* level of density functional theory20 using the Gaussian 03 package.27 The same

1092
HETEROCYCLES, Vol. 71, No. 5, 2007
computational method was used to compute the harmonic vibrational frequencies to estimate the
zero-point energy (ZPE) and thermal contributions to the gas phase enthalpies of the reaction components.
The thermal corrections were calculated for standard conditions (T = 298 K and p = 1.0 atm). The
homolytic C-H bond dissociation contributions to the gas phase enthalpies of the reaction components.
The thermal corrections were calculated for standard conditions (T = 298 K and p = 1.0 atm). The
homolytic C-H bond dissociation energy of 1b was calculated as D298 = H(1b•) + H(H•) − H(1b), where
H refers to the enthalpy of a given species at 298 K.
4) Detection of superoxide radical anion - Rapid mixing (<1 ms) of ice-cold reactants was performed in
a micro-volume stopped-flow UV/fluorescence detection system (SF-61 DX2) supported by the
Kynet-Asist2 software (HiTech Scientific). For monitoring ethidium formation, the end-product of
hydroethidine (Molecular Probes) oxidation, the red ethidium fluorescence (610 nm) was detected using a
535 nm excitation filter (Molecular Probes). The time-course of ethidium fluorescence intensity changes
(%) was recorded logarithmically, comprising 1000 data point acquired automatically. Ice-cold reactants,
reactant 1 (0.2 mM dihydroethidium in methanol) and reactant 2 (methanol) or reactant 3 (0.2 mM
(+)-lysergic acid methyl ester in methanol), were loaded into the reservoirs thermostatted at 30.0±0.1 oC.
After 3 min, aliquots (75 µl) of reactant 1 were mixed with equal amount of reactant 2 or 3. Baseline
reaction (reactant 1 with reactant 2) was repeated 8 times, and the 8 data sets were averaged to get one
trace for the baseline reaction. Measurements on the effect of (+)-lysergic acid methyl ester on ethidium
formation (reactant 1 with reactant 3) were repeated 8 times, also. Transformation of the percentile
fluorescence data into the percent change of fluorescence was performed with subtraction of the trace of
the baseline reaction.
ACKNOWLEDGEMENT
The Grants of 1/A/005/2004 NKFP Medichem2, the National Research Foundation (OTKA; No. T-046015
and K-60549), the Hungarian and the Finnish Academy (KR, LR) are gratefully acknowledged. Special
thanks are due to Dr. Ágnes Gömöry for the MS spectra as well as Ms. Gabriella Hanek and Ms. Mónika
Jeruska for technical assistance. The valuable help of Dr. Gábor Tárkányi is highly appreciated in some
of the NMR experiments. Also, appreciation is due Gedeon Richter Ltd. for providing 1a as a gift.
REFERENCES AND NOTES
1.
2.
A. Stoll, Chem. Rev., 1950, 47, 197.
a) A. Stoll and A. Hofmann, ‘The Ergot Alkaloids,’ The Alkaloids, Vol. 8, ed. by R. H. F. Manske
and H. L. Holmes, Academic Press, New York, London, 1965, pp. 726-783. b) P. A. Stadler and P.

HETEROCYCLES, Vol. 71, No. 5, 2007
1093
Stutz, ‘The Ergot Alkaloids,’ The Alkaloids, Vol. 15, ed. by R. H. F. Manske and H. L. Holmes,
Academic Press, New York, 1975, pp. 1-40 and citations therein.
3.
4.
5.
6.
A. Stoll, Swiss Patent, 79.879 (1918).
W. A. Jacobs and L. C. Craig, J. Biol. Chem., 1934, 104, 547.
S. Smith and G. M. Timmis, J. Chem. Soc., 1936, 1440.
W. A. Jacobs and L. C. Craig, J. Biol. Chem., 1936, 115, 227. The solution of (+)-lysergic acid
methyl ester {[α]D +85° (c 0.3, MeOH)}in MeOH was heated in a sealed-tube for 0.5 h. The optical
rotation change to [α]D +118° (c 0.3, MeOH) and remained at this value while the solution was
refluxed for a further 2 h.
7.
8.
9.
L. C. Craig, T. Shedlovsky, R. G. Gould, and W. A. Jacobs, J. Biol. Chem., 1938, 125, 289.
A. Stoll, A. Hofmann, and F. Troxler, Helv. Chim. Acta, 1949, 32, 506.
a) I. Ninomiya and T. Kiguchi, ‘Ergot Alkaloids’ The Alkaloids, Vol. 38, ed. by A. Brossi,
Academic Press, New York, 1990, pp. 1-148. b) M. Somei, Y. Yokoyama, Y. Murakami, I.
Ninomiya, T. Kiguchi, and T. Naito, ‘Recent Synthetic Studies on the Ergot Alkaloids and Related
Compounds’ The Alkaloids, Vol. 54, ed. by G. A. Cordell, Academic Press: San Diego, CA, 2000;
pp. 197-257; c) T. Kawasaki and K. Higuchi, ’Simple indole alkaloids and those with a
nonrearranged monoterpenoid unit’ Nat. Prod. Rep., 2005, 22, pp. 761-793.
10. A. Stoll and T. Petrzilka, Helv. Chim. Acta, 1953, 36, 1125.
11. L. Bernardi, G. Bosisio, O. Goffredo, and B. Patelli, Gazz. Chim. Ital., 1964, 95, 384.
12. U.S. Pat. 3.585.201 (1971).
13. I. Moldvai, E. Major-Temesvári, M. Incze, É. Szentirmy, E. Gács-Baitz, and Cs. Szántay, J. Org.
Chem., 2004, 69, 5993.
14. a) e.g. U.S. Pat. 3.880.856 (1975; E. Lilly and Co). b) V. Křen, P. Sedmera, M. Polášek, A.
Minghetti, and N. Crespi-Perellino, J. Nat. Prod., 1996, 59, 609. c) V. Křen, L. Weignerová, M.
Kuzma, A. Jegorov, and P. Sedmera, Heterocycles, 2001, 55, 1045. d) K. Jenett-Siems, I. Köhler, C.
Kraft, H. H. Pertz, V. Křen, A. Fišerova, M. Kuzma, J. Uhlrichová, and E. Eich, Bioorg. Med.
Chem., 2004, 12, 817.
15. K. Bailey and A. A. Grey, Can. J. Chem., 1972, 50, 387.
16. J. L. Marshall, Carbon-Carbon and Carbon-Proton NMR Couplings; Methods in Stereochemical
Analysis, Vol. 2, ed. by A. P. Marchand, Verlag Chemie International, Deerfield Beach, Florida,
1983; pp. 22-26.
17. Crystal data for 6a: C34H34N4O4 x ½ H2O x ½ C4H8O2, FW = 615.71, monoclinic P21, a = 10.798(2)
b = 24.899(5), c = 11.878(2) Å, β = 93.97(3)°, dc = 1.284 g/cm3, Reflections collected 5579 of
which 3846 independent, 820 parameters, R1 = 0.0802, wR2 = 0.1305, R1 = 0.1329, wR2 = 0.1566

1094
HETEROCYCLES, Vol. 71, No. 5, 2007
(all data), GOF = 1.108, largest diff. peak and hole 0.705 and -0.450 e Å3. Crystallographic data
(excluding structure factors) for the structures in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication numbers CCDC 608116. Copies of the
data can be obtained, free of charge, on application to CCDC, 12 Union Road, Cambridge CB2 1EZ,
UK. [fax: +44(0)-1223-336033 or e-mail: deposit@ccdc.cam.ac.uk].
18. F. H. Allen, Acta Cryst., 2002, B58, 380.
19. a) B. W. Herten and G. A. Poulton, J. Chem. Soc., Chem. Commun., 1975, 456. b) G. Büchi and R.
M. Freidinger, Tetrahedron Lett., 1985, 26, 5923. c) C. A. Muedas, R. R. Ferguson, S. H. Brown,
and R. H. Crabtree, J. Am. Chem. Soc., 1991, 113, 2233. d) J. H. Ye, J. Xue, K. Q. Ling, and J. H.
Xu, Tetrahedron Lett., 1999, 40, 1365. e) M. C. Kozlowski, E. S. DiVirgilio, K. Malolanarasimhan,
and C. A. Mulrooney, Tetrahedron: Asymmetry, 2005, 16, 3599.
20. a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648. b) C. Lee, W. Yang, and R. G. Parr, Phys. Rev. B,
1988, 37, 785. c) P. J. Stephens, F. J. Devlin, C. F. Chabalowski, and M. J. Frisch, J. Phys. Chem.,
1994, 98, 11623.
21. M. B. Smith and J. March, March’s Advanced Organic Chemistry, Wiley, New York, 2001; p. 234.
22. I. Kovács, É. Szárics, G. Nyitrai, T. Blandl, and J. Kardos, J. Neurochem. Int., 1998, 33, 399.
23. V. P. Bindokas, J. Jordan, C. C. Lee, and R. J. Miller, J. Neurosci., 1996, 16, 1324.
24. R. W. Hooft, COLLECT 1998. Nonius BV, Delft, The Netherlands.
25. A. J. M. Duisenberg, L. M. J. Kroon-Batenburg, and A. M. M. Shreurs, J. Appl. Cryst., 2003, 36,
220.
26. G. M. Sheldrick, SHELXTL 2001, Version 6, Brukes AXS Inc., Madison, Wisconsin, USA.
27. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A.
Montgomery Jr., and co-workers, 2004, Gaussian 03, revision C.02; Gaussian, Inc.: Pittsburgh, PA.