1H, 13C NMR and X-ray crystallographic studies of highly polyhalogenated derivatives of costunolide lactone

Article abstract of DOI:10.1016/j.saa.2005.02.007

The costunolide lactone, a sesquiterpene compound isolated from Zaluzania triiloba species, reacted with several dihalocarbene sources produced by trihaloform-NaOH under successive phase transfer reactions yielding mono-, bis- and tris-dihalocyclopropane adducts. The structures, as well as the configurational assignments of the different derivatives, were established by ¹H and ¹³C NMR spectroscopy and assisted by X-ray crystallographic and molecular modelling studies. The specific shielding of protons in the neighbourhood of different halogens on the cyclopropane moieties was correlated to the pseudocontact interactions.

Full text of DOI:10.1016/j.saa.2005.02.007

Spectrochimica Acta Part A 62 (2005) 604–613
1H, 13C NMR and X-ray crystallographic studies of highly
polyhalogenated derivatives of costunolide lactone
D. Coronaa,∗, E. Dıaz , J.L. Nava , A. Guzman , H. Barrios , A. Fuentes ,
a,∗
a
a
a
b
´
´
S.A. Hernandez-Platab, J. Allardc, C.K. Jankowskic
a
Instituto de Qu´ımica, Universidad Nacional Auto´noma de Me´xico, Circuito Exterior, C.U. Coyoaca´n, 04510 Me´xico D.F., Mexico
b
Facultad de Qu´ımica, Universidad Auto´noma del Estado de Me´xico, Paseo Tolloca´n y Colo´n, Toluca 50000, Toluca, Estado de Me´xico
c
Faculte´ des Sciences, U de Moncton, Moncton, NB, Canada E1A 3E9
Received 7 December 2004; received in revised form 11 February 2005; accepted 11 February 2005
Abstract
The costunolide lactone, a sesquiterpene compound isolated from Zaluzania triiloba species, reacted with several dihalocarbene sources
produced by trihaloform-NaOH under successive phase transfer reactions yielding mono-, bis- and tris-dihalocyclopropane adducts. The
1
structures, as well as the configurational assignments of the different derivatives, were established by H and 13C NMR spectroscopy and
assisted by X-ray crystallographic and molecular modelling studies.
Thespecificshieldingofprotonsintheneighbourhoodofdifferenthalogensonthecyclopropanemoietieswascorrelatedtothepseudocontact
interactions.
© 2005 Elsevier B.V. All rights reserved.
Keywords: 1H and 13C NMR; Polyhalogenated cyclopropane derivatives from costunolide lactone
1. Introduction
In this area, for example, some interesting acetylenic com-
pounds were isolated from L. grandulifera showing the ox-
ocin structures with one or two bromine atoms [6,7]. Fenical
[8,9] reported on the isolation of bromo and chloro acetylenic
alcohols from Chondria oppositiclada algae, and on the iso-
lation from seaweed of the Laurentia species of an intriguing
halogenated vinyl peroxide. Kazlauskas’ [10] investigations
of Antartic and Australian Delisea fimbriata led to the dis-
coveryofiodine-containingsubstancesthatdifferstructurally
from the well-known naturally occurring iodinated products,
in particular the thyroxine-related compounds.
Finally, several years ago Sims et al. [11] identified, in An-
tartic D. fimbriata, a series of C8 ketones containing bromine,
chlorine and iodine atoms, whose structures were confirmed
by spectroscopic studies.
Natural occurring sesquiterpene lactones do not usually
contain halogens. Among the interesting and simple syn-
thetic methodologies developed to build such compounds
the addition of the halocarbenes could be used to obtain
their dihalocyclopropane adducts. Mathias and Weyerstahl
Because of their magnetic sensitivity, their quadrupole
moment and their associated line broadening [1–4] the four
common halogen elements, except 19F, are not often used
in organic NMR studies. For this reason, structural stud-
ies involving natural products containing chlorine, bromine
and iodine, and their polyhalogenated derivatives are typi-
cally performed using 1H and 13C NMR spectroscopy. More
recently, the environmental occurrence of polyhalogenated
compounds originated from the modification of natural prod-
uctsbywidespreadhalogenationhasattractedattentiontothis
newer and still evolving branch of natural product chemistry.
Although their importance in synthetic organic chemistry
is well known, the presence of halogen nuclei in the primary
structures of natural products is less common: their occur-
rence is limited mainly to the marine natural products [5].
∗
Corresponding authors. Tel.: +52 6224421; fax: +52 6162203.
E-mail address: maudiaz@servidor.unam.mx (E. D´ıaz).
1386-1425/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2005.02.007

D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
605
[12] have described the preparation and spectroscopic prop-
erties of hetero-di-halosubstituted styrenes (pair of F and Cl;
Br and I, etc.) and other olefins which display interesting
NMR behaviour as far as the proton chemical shifts and their
proton–proton coupling constants are concerned.
Despite experimental data accumulated in this field during
the last decade, it should be noted that very few dihalocarbene
additions to the double bonds of ␣,␤-unsaturated carbonyls
have been reported.
Our interest in difluoro carbene chemistry led to the
synthesis and spectral description of several unsaturated
sesquiterpene lactones after their exposure to a large excess of
difluoro carbene (the small and easy-to-prepare source of the
cyclopropane ring), and the investigation of the compounds
formed [13–15].
Dichlorocarbenes have also been used as intermedi-
ates for the ␥-lactone terpenes synthesis, in particular for
the sesquiterpene lactone synthesis [16,17]. Dibromocyclo-
propane derivatives of the sesquiterpenes have also been used
by our group for the synthesis of allene derivatives via the
Hiyama red-ox Cr2+-catalyzed reaction [18] and more re-
cently applied to some unusual rearrangements of dichloro-
cyclopropanes into dichloro methylvinylic derivatives and
␣,␤-unsaturated aldehydes and acids [19].
by two mechanism and experimentally observed shielding
may be rationalized as being of either, or both, of two origins.
One of these gives rise to contact interactions as shown in
the equation
ꢀB/B0 = −ange2Be2S(S + 1)/3gNBNKT
whereas the other produces the dipolar or pseudocontact in-
teractions given by equation
ꢀB/B0 = −[(3 cos2 θ − 1)Be2S(S + 1)/3r3KT]F(g)
In the equations presented, the paramagnetic-induced res-
onance shift ꢀB is expressed as a function of the applied
magnetic field B0, the hyperfine interaction constant an is
expressed in Tesla, ge is the rotationally averaged electronic
value g, Be is the Bohr magneton, gn and BN are the corre-
sponding nuclear parameters, and S is the spin of unpaired
electrons.
In the second equation several molecular geometry pa-
rameters appear. For instance, r is the separation between the
resonating nuclei and the unpaired electrons (expressed in
radians), θ is the angle between the vector defined by r and
the principal axis of symmetry of the molecule, and F(g) is a
function of the g tensor component. Substantial decrease and
increase in nuclear shielding may then result from either the
contact or dipolar interactions, or from both.
From these equations, we see that the paramagnetic influ-
ence on the nuclear shielding on protons in the NMR spec-
trum is of great importance when considering the heavy nu-
clei presence in the molecule. Distance and other geometrical
considerations between protons and halogens should then be
addressed, in order to explain the observed chemical shift for
protons in the vicinity of halogen atoms.
Furthermore, since there are very few works reports on
polyhalogenated derivatives of sesquiterpene lactones and in
particular on their detailed NMR characteristics, we decided
to undertake this study and report on the synthesis of poly-
halogenated cyclopropane adducts attached to the costuno-
lide lactone skeleton as well as to discuss the NMR proper-
ties of the protons attached to these structures and which are
under the direct influence of the halogenated cyclopropane
moieties.
As it has been said, the influence of fluorine, chlorine,
bromine and iodide on the chemical shifts of protons or car-
bons is not well understood or explained. In this study, we
used derivatives of sesquiterpenic costunolide lactone rigidi-
fied by introduction of dihalocyclopropane rings, in order to
explain the trend in the shielding of selected proton, induced
by different halogens placed in strictly geometrically defined
positions. As can be noticed the exact interpretation of the
proton spectra and the unambiguous assignment of protons
is then the key starting point to this study. The geometrical
features of the molecules should also be a credible; in this re-
specttheX-rayandmolecularmodellingdatawereconstantly
challenged by the observed chemical shift, as compared to the
estimated value.
Detailedstudiesofthehighfield2DNMRspectrarevealed
a special behaviour of the protons attached to the sesquiter-
pene lactones with one, two or three different dihalocyclo-
propane rings. We linked this behaviour to their specific ge-
ometry and studied it with stereochemical data deduced from
the X-rays and molecular modelling, particularly in the con-
text of the pseudocontact interaction.
The shielding constant of nuclei is usually represented as a
sum of two terms: a positive diamagnetic term and a negative,
paramagnetic term.
The diamagnetic term largely depends on the electronic
density of the nearby nucleus and remains almost constant
in most chemical environments, even for heavy nuclei as the
halogens are. The intramolecular influence [20] of one nu-
cleus on the shielding of the other is called the heavy atom
effect. In this case, the nuclear shielding of a lighter nucleus
is translated into the change in its electronic energy level, on
account of the spin-orbit coupling interactions derived from
the neighbouring heavy nuclei.
On the other hand, the presence of a paramagnetic centre,
either on the molecule itself or in the neighbourhood of the
observed nuclei, can severely influence the observed nuclear
shielding. The induced shielding effect may then be produced
2. Results and discussion
Because of the crucial importance of the correct, unam-
biguous assignment of the proton NMR spectra of the com-
pounds chosen in this study, we present in the Section 5, the
details of the data used to realize this objective. The tables
of NMR data summarize the results used to correlate them to

606
D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
Scheme 1.
1
the geometrical parameters deduced from X-rays or molecu-
lar modelling experiments for this group of compounds.
Several mono-, bis- and tris-dihalocyclopropane costuno-
lide lactone adducts studied are presented in Scheme 1
The proton NMR spectra were analyzed on the basis of
changes in halogen carbon bond distances as well as on the
contribution of pseudocontact effect induced by the distance
and geometrical parameters of these changes.
The complete H and 13C assignments of the molecules
described herein were achieved by means of the 1H/1H
(COSY, NOESY) [21,22] and H/13C correlation methods
1
(HMQC, HMBC) [23–25].
For every adduct described, either 4␣–15␣, 11␣–13␣ or
10␣–14␣ derivatives were obtained. On the ground of the
configuration of the dihalocyclopropane adduct obtained, one
would expect that H-5␣, H-7␣ or H-1␣ should undergo an

D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
607
additional deshielding through the spatial interactions with
the halogen atoms present on the cyclopropane moiety since
the models may suggest a long range deshielding.
The stereochemistry of the dihalocyclopropane at
C4 C15, C10 C14, C11 C13 in these compounds can be as-
certained by the available NMR data and proved through
the X-ray crystallographic studies as performed for the
compounds 6 and 7, as well as for the hexafluorinated
adduct 13 described some years ago [13,14,26]. Deriva-
tive 13 was included in this study in order to afford the
complete series of parameters such as size, bond distances
and bond angles of derivatives containing several fluorine
atoms.
The data such as C Br distance, Br C Br bond angle, etc.
were extracted from the crystallographic studies of structures
6, 7 and 13 and those for the structures 3, 4, 8–12 were ob-
tained from molecular modelling data (Table 2).
3. Tentative semi-empirical relation between
observed chemical shift of the alpha to a
cyclopropane proton and the nature and geometry of
that cyclopropane
The cyclopropane proton data (chemical shifts, angular
parameters)forthesecompoundswerethenexaminedsearch-
ing first for a geometrical relation between the proton chem-
ical shift on the methine ␣ to cyclopropane-ring and the
presence of the gem-dihalogens on the same cyclopropane
ring. Such a semi-empirical relation should correspond to
the already specified requirement of the angular-dihedral (or
longer) angle and distance between halogen and the proton
effect on the proton chemical shift. The correction parameters
should be electronegativity (taking account of the nature of
the halogen atom) and a constant which will serve to the sta-
tistical treatment of the ensemble, and it will combine other
geometrical parameters used either as a multiplicatory or ad-
ditive corrective term.
The 1H NMR spectrum of the adducts 2–12 displayed AB
pattern corresponding to the methylene of the dihalocyclo-
propane attached on C11 C13, C4 C15 and C10 C14 which
was unambiguously assigned through the COSY spectrum.
For example, the dihalocyclopropane attached at C11 C13
in adduct 5 displayed at δ = 2.16 a doublet (J = 8.0 Hz) as-
signed to the exo-proton to the carbonyl lactone group, while
the higher field doublet at δ = 2.02 was assigned to the endo
proton.
The usual AB patterns in all adducts having the dihalocy-
clopropane moiety, either on C11 C13, C4 C15 or C10 C14
double bond, showed the methylene geminal couplings be-
tween 7 and 8.0 Hz which enabled with help of the COSY
spectrum, to assign the cyclopropane protons on C13, C14 or
C15 in the adducts.
ThroughtheCOSYspectrum, itwasalsopossibletoassign
the key protons H-1, H-5 or H-7, since they shown the most
relevant chemical shift changes which enabled to witness the
pseudocontact contribution from the halogen atoms attached
to the molecule.
With the complete proton assignment obtained from the
COSY spectrum, 13C assignments were easily deduced from
HMQC and HMBC spectra.
Our search was oriented first toward the homodihedral
dependence of the chemical shift for the same proton on the
H Ca Cb Cc X homodihedral angle different for each of
two halogens (θHXcis or θHXtrans). In this case the equation
found for 46 data was established as:
Table 1 presents the 1H chemical shifts of the compounds
2–13. In this table is also included some of the 13C chemical
shifts of compounds 1, 3–7 and 10 the assignments performed
using DEPT, HMQC and HMBC experiments.
3 cos2 θ − 1
δH ∼ KA
f(axn constant),
(1)
(1)
r3
A general trend was observed for the proton chemical shift
ofthecyclopropanemoietiesinallofthehaloderivativesstud-
ied. It was commonly observed that those protons on C-14
(compounds 2, 4, 6, 9, 11 and 12; Table 1) displayed a highest
chemical shift as well as the lesser chemical shift difference
between both protons.
The X-ray crystallographic study of derivative 13, previ-
ously reported [13] permitted us to extract several important
parameters such as C F, F C F bonds angles as well as spa-
tialdistancesbetweenH7 F4, H7 F3, H5 F2, H5 F1, H1 F6
and H1 F5 in particular (see Table 2).
xn = electronegativity (Pauling scale)
The coefficient of determination within this group was found
to be only R2 = 48%. Following this track the search for
better correlation enabled us to consider the dependence
of the proton chemical shift on the dihedral angle between
H Ca Cb Cc, θHCc. Two halogens electronegativity was
considered as the correction parameter to the equation.
Derivative 7 was used as model to investigate the crys-
tallographic parameters involved in the diiodo cyclopropane
moiety attached on C11. In such a manner, C16 I2, C16 I1
bond distances, I2 C I1 bond angle as well as H7 I1 and
H7 I2 spatial distances were obtained.

Table 1
Position
Compound
1
δH (2)
3
4
5
6
7
δH (8)
δH (9)
10
δH (11)
δH (12)
δH (13)
δH
δC
46.5
δH
δC
48.3
δH
δC
51.1
δH
δC
47.3
δH
δC
49.1
δH
δC
47.3
δH
δC
48.3
1
2
2.93
1.88
1.80
2.52
2.48
1.96
1.83
1.79
2.55
2.36
3.19
2.05
1.87
2.13
2.08
2.03
1.87
1.80
2.55
2.37
2.99
1.96
1.91
2.56
2.54
2.24
1.90
2.05
2.06
2.15
3.00
1.95
1.91
3.02
2.99
3.21
2.05
1.85
2.10
2.10
2.28
2.04
1.95
2.14
2.08
3.16
1.90
2.0
2.10
2.07
2.20
1.99
1.96
2.11
2.09
2.18
2.04
1.98
2.04
2.04
2.04
2.58
1.92
5.15
29.5
29.2
27.9
30.2
26.0
30.2
29.3
3
32.1
33.6
29.6
32.6
30.8
33.2
31.7
4
5
6
7
8
152.1
51.5
84.7
43.9
30.3
39.1
54.0
82.0
44.1
29.9
148.4
50.6
80.4
43.0
25.6
151.1
52.1
84.6
44.5
31.8
38.9
51.9
81.0
43.2
25.6
151.2
52.0
84.4
47.4
35.1
39.9
51.5
81.9
44.1
29.4
2.86
3.94
2.98
2.25
1.32
2.41
2.13
2.92
3.98
3.03
2.70
1.44
2.34
1.64
2.52
4.35
2.66
1.75
1.28
2.34
2.27
3.02
4.08
2.82
2.24
1.20
2.30
1.62
2.99
4.08
2.70
1.76
1.21
2.43
2.10
2.67
4.37
2.82
1.23
2.26
2.39
1.61
3.0
2.53
4.36
2.39
1.75
1.25
2.35
2.35
2.64
4.38
2.55
2.30
1.22
2.39
1.63
2.52
4.34
2.65
1.28
1.75
2.34
2.28
2.53
4.24
3.03
2.60
1.53
2.41
1.62
2.59
4.35
2.81
2.14
1.26
2.39
1.56
2.20
4.51
2.70
1.76
1.33
2.25
1.49
4.09
2.41
1.74
1.20
2.44
2.07
9
35.8
29.9
27.7
35.7
27.5
35.7
29.3
10
11
12
13
149.7
139.8
169.5
119.9
147.0
36.0
172.0
31.8
32.3
37.0
171.9
31.7
148.7
35.6
172.2
31.2
32.1
36.4
171.9
31.8
148.7
31.9
172.7
32.7
146.8
35.7
171.9
29.5
6.04
5.64
4.86
4.74
5.08
4.99
6.27
6.57
1.60
1.50
5.31
5.08
2.15
2.02
5.07
4.89
2.36
1.66
2.21
2.02
1.56
1.49
5.30
5.10
2.16
2.02
4.90
4.82
5.28
5.09
2.20
2.05
1.57
1.55
2.33
1.67
2.27
2.19
4.89
4.82
5.29
5.08
2.28
2.16
5.08
4.90
2.37
1.62
2.33
2.21
1.57
1.54
2.32
1.69
2.15
2.02
5.05
4.90
1.47
2.17
6.27
5.57
1.41
1.39
2.22
1.43
2.19
2.03
1.36
1.38
2.14
1.48
2.01
1.96
1.16
1.16
2.06
1.59
14
15
111.9
108.2
115.8
30.4
39.8
37.3
112.8
109.7
27.4
37.0
31.0
112.8
109.7
61.2
115.3
28.1
112.5
16
17
18
27.4
36.7
38.4
27.2
36.0
66.3
27.4
Other
OAc, 2.1

D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
609
Table 2
Compound
6a
Bond angle (◦)
Bond distance (A)
Br4 17: 1.935,
H· · ·X distance (A)
Dihedral angle (◦)
˚
˚
Br4 C17 Br3: 110.7,
Br1 C16 Br2: 108.7,
Br5 C18 Br6: 110.1
C
H7· · ·Br4: 2.922,
H7· · ·Br3: 4.303,
H5· · ·Br1: 2.735,
H5· · ·Br2: 4.533,
H1· · ·Br6: 2.798,
H1· · ·Br5: 4.644
H1 C1 C10 C18: −32.9, C1 C10 C18 Br6: −2.7,
H1 C1 C10 C18 Br6: −35.6,
Br3 C17: 1.896,
Br1 C16: 1.917,
Br2 C16: 1.916,
H1 C1 C10 C18 Br5: −190.0, H7 C7 C11 C17
:
36.9, C7 C11 C17 Br4: 9.2, H7 C7 C11 C17 Br4:
46.1, H5 C5 C4 16: 8.2, C16 C4 C5 Br1: 9.7,
Br5
C
18: 1.942,
C
Br6 C18: 1.882
H5 C5 C4 C16 Br1: 1.0, C7 C11 C17 Br3:
143.8, H7 C7 C11 C17 Br3: −105.0,
C1 C10 C18 Br5: −157.8, C5 C4 C16 Br2:
−148.0, H5 C5 C4 C16 Br2: −156.0
H7 C7 C11 C16: 50.2, H7 C7 C11 C16 I1: 55.6,
C7 C11 C16 I1: 5.4, H7 C7 C11 C16 I2: −98.4,
C7 C11 C16 I2: −148.6
7a
I1 C16 I2: 109.1
I1 C16: 2.162,
I2 C16: 2.125
H7· · ·I1: 3.012,
H7· · ·I2: 4.506
13a
F4 C17 F3: 107.9,
F1 C16 F2: 107.0,
F5 C18 F6: 107.1
F3
C
17: 1.342,
H7· · ·F4: 2.580,
H7· · ·F3: 3.840,
H5· · ·F2: 2.425,
H5· · ·F1: 4.075,
H1· · ·F5: 2.488,
H1· · ·F6: 4.135
H5 C5 C4 C16: −15.3, C5 C4 C16 F2: 8.6,
H5 C5 C4 C16 F2: −6.7, H7 C7 C11 C17: 36.9,
C7 C11 C17 F4: 8.2, H7 C7 C11 C17 F4: 45.1,
H1 C1 C10 C18: −34.7, C1 C10 C18 F5: −0.5,
H1 C1 C10 C18 F5: −34.2, C1 C10 C18 F6:
−151.2, C5 C4 C16 F1: −145.8,
F4 C17: 1.344,
F2 C16: 1.360,
F1 C16: 1.354,
F5 C18: 1.357,
F6 C18: 1.344
H5 C5 C4 C16 F1: −160.8,
H7 C7 C11 C17 F3104.8
9b
I1 C17 I2: 111.5,
Br1 C16 Br2:
112.3, Br3 C18 Br4:
112.2
I1
C
17: 2.10, I2 C17
2.10, Br1 C16: 1.909,
Br2 C16: 1.91,
:
H7· · ·I1: 2.996,
H7· · ·I2: 4.488,
H5· · ·Br2: 2.735,
H5· · ·Br1: 4.564,
H1· · ·Br4: 2.80,
H1· · ·Br3: 4.684
H7 C7 C11 C17: 35.6, C7 C11 C17 I1: 9.8,
H7 C7 C11 C17 I1: 45.4, H1 C1 C10 C18 Br4:
−35.6, H1 C1 C10 C18: −32.3, C1 C10 C18 Br4:
Br3
Br4
C
C
18: 1.909,
18: 1.909
−3.3, H5 C5 C4 C16 Br2: 3.1, H5 C5 C4 C16
−8.7, C5 C4 C16 Br2: 9.8, C7 C11 C17 I2:
−144.8, H7 C7 C11 C17 I2: −109.2,
H5 C5 C4 C16 Br1: −155.2, C5 C4 C16 Br1:
−147.9
:
12b
Br1 C17 Br2: 112.6,
Cl1 C16 Cl2: 113.5,
Cl3 C18 Cl4: 113.6
Br1 C17: 1.90,
Br2 C17: 1.908.
Cl1 C16: 1.757,
Cl2 C16: 1.758,
Cl3 C18: 1.757,
Cl4 C18: 1.757
H7· · ·Br1: 2.885,
H7· · ·Br2: 4.292,
H5· · ·Cl1: 4.391,
H5· · ·Cl2: 2.635,
H1· · ·Cl3: 2.70,
H1· · ·Cl4: 4.475
H1 C1 C10 C18: −33.8, Cl3 C18 C10 C1: −1.8,
H7 C7 C11 C16: 37.6, Br1 C16 C11 C7: 9.2,
H7 C16 C11 C7 Br1: 46, C5 C4 C16 Cl2: 9.9,
H5 C5 C14 C16: −9.2, H5 C5C4 C16 Cl2: 0◦,
H1 C1 C10 C18 Cl3: −35.6, C1 C10 C18 Cl4:
−155.8, H1 C1 C10 C18 Cl4: 190.0,
H1 C1 C10 C18 Cl3: −36.2, C5 C4 C16 Cl1:
−147.7, H5 C5 C4 C16 Cl1: −157.4,
C7 C11 C17 Br2: −143.7, H7 C7 C11 C17 Br2:
−106.0
10b
Cl C16 Cl: 113.9,
Br C17 Br: 112.0
Cl1
Cl2
Br1
Br2
C
C
C
C
16: 1.757,
16: 1.757,
17: 1.909,
17: 1.908
H5· · ·Cl1: 2.648,
H5· · ·Cl2: 4.394,
H7· · ·Br1: 2.990,
H7· · ·Br2: 4.323
H5 C5 C4 C16: −17.3, C5 C4 C16 Cl1: 10.6,
H5 C5 C4 C16 Cl1: −4.9, H7 C7 C11 C17: 46.3,
C7 C11 C17 Br1: 7.2, H7 C7 C11 C17 Br1: 53.5,
C5 C4 C16 Cl2: −145.0, H5 C5 C4 C16 Cl2:
−160.5, C7 C11 C17 Br2: −145.6,
H7 C7 C11 C17 Br2: 99.3
3b
Br C17 Br: 112,
Br C16 Br: 111.4
Br C17: 1.909,
Br C17: 1.908,
Br C16: 1.909,
Br C16: 1.909
H7· · ·Br1: 3.043,
H7· · ·Br2: 4.343,
H5· · ·Br3: 2.765,
H5· · ·Br4: 4.598
H7 C7 C11 C17: 47.5, C7 C11 C17 Br1: 6.8,
H5 C5 C4 C16: −15.6, C5 C4 C16 Br3: 10.7,
H5 C5 C4 C16 Br3: −5.0, H5 C5 C4 C16 Br4:
−161.2, H7 C7 C11 C17 Br1: 54.3,
H7 C7 C11 C17 Br2: −98.5, C5 C4 C16 Br4:
−145.6, C7 C11 C17 Br2: −99.3
8b
I
C17 I: 110.9,
I
C17: 2.10, I C:
2.10, Br C16: 1.908,
Br C16: 1.910
H7· · ·I1: 3.133,
H7· · ·I2: 4.524,
H5· · ·Br1: 2.758,
H5· · ·Br2: 4.592
H7 C7 C11 C17: 45.0, C7 C11 C17 I1: 7.4,
H7 C7 C11 C17 I1: 52.4, C5 C4 C16 Br1: 10.6,
H5 C5 C4 C16: −15.2, H5 C5 C4 C16 Br1:
−4.6, H5 C5 C4 C16 Br2: −161.0,
Br C16 Br: 112.6
C5 C4 C16 Br2: −145.8, H7 C7 C11 C17 I2:
−101.2, C7 C11 C17 I2: −146.2
11b
Cl C16 Cl: 113.8,
Cl C17 Cl: 113.5
Cl
Cl
C
C
16: 1.757,
16: 1.757,
H5· · ·Cl3: 2.664,
H5· · ·Cl4: 4.434,
H1· · ·Cl1: 2.832,
H5· · ·Cl2: 4.147
Cl3 C16 C4 C5: 7.8, H5 C5 C4 C16: −18.9,
H5 C5 C4 C16 Cl3: 11.1, H1 C1 C10 C17: 47.2,
C1 C10 C17 Cl1: 8.0, H1 C1 C10 C17 Cl1: 55.2,
H1 C1 C10 C17 Cl2: −100.0, C1 C10 C17 Cl2:
−147.2, H5 C5 C4 C16 Cl4 −166.6,
Cl C17: 1.757,
Cl C16: 1.757
C5 C4 C16 Cl4: −147.7

610
D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
Table 2 (Continued )
˚
˚
Compound
4b
Bond angle (◦)
Bond distance (A)
H· · ·X distance (A)
Dihedral angle (◦)
H1 C1 C10
17: −31.2, C1 C10 C17 Br3: −3.4,
Br C17 Br: 107.5,
Br C16 Br: 112.5
Br C17: 1.909,
Br C17: 1.913,
H7· · ·Br1: 2.906,
H7· · ·Br2: 4.315,
H1· · ·Br3: 2.794,
H1· · ·Br4: 4.644
C
H1 C1 C10 C17 Br3: −34.6, H7 C7 C11 C16
:
Br
C
16: 1.908,
38.6, C7 C11 C16 Br1: 8.4, H7 C7 C11 C16 Br1:
47.2, C7 C11 C16 Br2: −144.6,
H7 C7 C11 C16 Br2: −136.0,
Br C16: 1.909
H1 C1 C10 C17 Br4: −190.0, C1 C10 C17 Br4:
−158.7
a
Method: X-ray.
b
Method: molecular modelling.
In this case we arrived to Eq. (2) with the correlation using
trans to the halogens) remains relatively constant and varies
from 140◦ to 150◦. In the calculations (Hyper Chem 6 Mm+)
the bond distances of Cc X used were for F, Cl, Br and I,
23 proton chemical shifts. The measured data give a confi-
dence of 43%. The use of the electronegativity of the different
halogens enabled us to improve the coefficient of determina-
tion of 59%.
˚
respectively, 1.34, 1.75, 1.91 and 2.16 A.
The cyclopropane proton data for these compounds
(chemical shifts and angular parameters) were then treated
searching for other geometrical relation between the proton
chemical shift on ␣- to cyclopropane ring and the influence
of the gem-dihalogen atoms.
n cos2 θ1 cos2 θ2
δH Kb
f(bxn constant)
(2)
r3
As a result of these two correlations it seems that the chemical
shift of alpha to cyclopropane proton is sensitive to both cis
and trans halogens. This dependence however, could be pre-
sented as the trigonometrical function of the cos2 θH,Ca, the
angle between observed proton and the third carbon which
is holder of two geminal halogens. The correlation with two
individual halogen angles is not as important in this mea-
surement as this first dihedral angle value is. It seems that the
protons are shielding according to the combine effect of two
halogens on this carbon, where both halogens act as a correc-
tive term. The angle between these two halogens is related to
its sp3 character, two bond X Cc X angle vary from 108◦ to
113◦, which is not much different, for the entire halogen se-
ries. The dependence on the distance between the proton and
the single halogen also seems to be already included in the
angular term and from one compound to the other does not
vary much. The partition of the X Cc X angle in projection
to the observed proton ion as “homodihedral” partial angle
which in principle should affect the r−3, seems also to be of
secondary importance because the measured difference be-
tween the homodihedral angles of H Ca Cb Cc Xcis and
for the same proton H Ca Cb Cc Xtrans (proton cis and
Such a semi-empirical relation should correspond to the
already mentioned requirements of the angular (dihedral or
other angle) and distance between halogen and the proton
effects on the proton chemical shift. The correctional pa-
rameter should again be an electronegativity, which takes
into account the nature of the halogen atom (the usual
constant will serve the statistical treatment of the ensem-
ble).
Following these calculations, two more correlations were
made on the ensemble of geometrical parameters, because
of the availability of separately, two homodihedral angles
H Ca Cb Cc Xforeveryproton. Thesemi-empiricalequa-
tion was made with the assumption that, of two halogens
placed in an space around a proton in specific manner, the cis
halogen has different influence than the trans one. Reduction
of this semi-empirical equation allowed to define the term δH
according to Eq. (3) (Table 3) and which for a given set of
23 data, showed an R2 = 16% for cis and R2 = 54% for trans
without any corrective Xn term.
When the electronegativity Xn is applied, as a correc-
tive multiplicatory term according to Eq. (5) (Table 3) the
Table 3
Karplus-type equation between δHx of the ␣- to cyclopropane proton and dihalocyclopropane nature and geometry
Formula of equationa
R2 (%)b
δH ∼ f(An; cos2 θX,Hcis ; C)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
16
54
51
63
49
63
76 (64)c
75 (62)c
δH ∼ f(An; cos2 θX,Htrans ; C)
cis
δH ∼ f(An; cos2 θX,H ; C; Xn)
trans
δH ∼ f(An; cos2 θX,H ; CXn)
trans
δH ∼ f[A1(cos2 θX,H ; C1) + B1(cos2 θX,H ; C2)]
cis
δH ∼ f[A1(cos2 θX,H ; C1) + B1(cos2 θX,H ; C2) + DXn]
trans
cis
δH ∼ f[A1(cos2 θX,H ; C1) + B1(cos2 θX,H ; C2)Xn]
trans
cis
δH ∼ f[A1(cos2 θX,H ; C1) + B1(cos2 θX,H ; C2)](Dσ)
trans
cis
a
A, B, C, D – constants; Xn – electronegativity, Pauling scale; σ – Shoolery shielding effects; θx – angle as defined between X and Hx (Hcis or Htrans); X F,
Cl, Br or I (cis or trans); f = function of . . .. For detailed Eq. (9) for three halogens δH = (2.33 + 2.81 cos2 θX,H
− 18.68 cos2 θX,H )Xn.
trans
cis
b
Coefficient of determination or R2.
For three halogens, except iodine
c

D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
611
agreement reached was 51% but according to Eq. (3) reach
only 16.5%.
for 13C spectra were obtained at 300 K and chemical shifts
1
are expressed relative to TMS (0.00 ppm) for H. For 13C
Finally, the last semi-empirical equations were deduced
with the emphasis on both halogens influence on the mea-
sured proton chemical shift. As a result the coefficient of
determination for all 23 data (23 pair of coordinates) were
at 49% without electronegativity term (Eq. (7)) and up to
63% with Xn additive term (Eq. (6)). The same term used as
multiplicatory correction lead to 64% of correlation (Eq. (9)).
Last attempts were made assuming that the agreement of
data for three of four halogen F, Cl and Br, is much better
than when I is included into calculations. As a result of such
a reduced set of data, of 20 instead 23 (or six values over
46 entries less) for the best equation (Eq. (9)) the agreement
factor found was 76%. When additionally monitored by elec-
tronegativity, used as an additive term the correlation was at
64% and when the electronegativity was used as a corrective
multiplicatory term, the R2 within such a reduced set was at
74%.
From these nine models, the best semi-empirical equation
found was by far Eq. (9) for the three halogens, which means
that the dependence of the observed chemical shift of both
halogens cis and trans to the methine proton in alpha-position
to the dihalocyclopropane ring could be modulated via the
homodihedral angle. When this “Karplus-type” equation was
monitored by Shoolery shielding effect corrections, applied
as multiplicatory term, the best fit of data for four halogens (F,
Cl, Br and I) was found at R2 = 62%. When the same equation
was recalculated on a reduced number of halogens (without
iodine) this fit was improved to R2 = 75% (Eq. (10)).
NMR reference, the centre peak of the 1:1:1 multiplet of the
CDCl3 was assigned the value of 77.0 ppm.
The experiments were performed using an inverse detec-
tion 5 mm probe. The COSY, NOESY, HMQC and HMBC
spectra were recorded using the standard Varian Unity pro-
grams.
The 1H and 13C resonances were identified through inter-
active interpretation of NOESY, COSY, HMQC and HMBC
spectra (Table 1).
For analytical purpose, the mass spectra were recorded
on a JEOL JMS-5X 10217 instrument in EI/PI mode, 70 eV,
200 ◦C, via direct probe. Only the molecular ion (m/z) values
are reported. IR spectra were recorded on a Nicolet Magna
55-X-FT instrument.
In order to prepare some fluorinated adducts derived
from costunolide lactone, we performed two different unsuc-
cessful reactions: CF3HgPh/C6H6/reflux (Seyferth reaction)
[27–29]; or ClCF2COONa /diglima/reflux. Both procedures
failed due to the fact that costunolide lactone underwent un-
desirable fast polymerization.
Because we considered necessary to have available
1
high resolution H NMR data of tris-difluoro cyclopropane
adduct to compare we have used as model the tris-
difluorocyclopropane derivative of Zaluzanin C acetate (13)
previously described [26] which enabled us to get the proton
chemical shift sought.
The dihalocarbene, generated by phase transfer reaction
using HCX3 (X = Cl, Br, I)/NaOH/Triton B has already been
used to synthesize sesquiterpenic dihalocyclopropane lac-
tones as well as the corresponding mono-, bis- and tris-
dihalocyclopropane adducts (see Scheme 1). Thus, when cos-
tunolide lactone 1 was reacted with HCBr3/NaOH/Triton B,
the mono-dibromocyclopropane adducts 2 and 5, were ob-
tained as well as the correspondent bis-dibromocyclopropane
3 and 4 and tris-dibromocyclopropane 6 derivatives. The
structures of these compounds were characterized mainly by
1H and 13C NMR spectra (see Table 1).
Compound 7 is a diiodine derivative obtained with the
phase transfer reaction of the costunolide lactone 1 with
CHCI3. When this diiodocyclopropane derivative was later
reacted with HCBr3/NaOH/Triton B a new diiodo, dibromo
adduct 8 as well as diiodotetrabromo adduct 9 were obtained.
Chlorinated derivatives were obtained when the source of
the dihalocarbene was HCCl3. Therefore, when costunolide
lactone 1 reacted with HCCl3/NaOH/Triton B one only tetra-
chlorinated adduct was obtained.
4. Conclusions
The interpretation of these results is leading to the conclu-
sion that the induced proton chemical shift of the protons H1,
H5 and H7 is depending on the dihedral angle between halo-
gen atoms in the vicinity as well as on the proton–halogen dis-
tance, according to the pseudocontact Eq. (2) where the later
relationship shown a dependence on r−3, and on the cos2 θ
(homodihedral). As can be observed in Table 1, the protons
H1, H5 and H7 shown a paramagnetic shielding following the
relationship I > Br > Cl > F which is opposite to the expected
relationship coming from electronegativity (F > Cl > Br > I).
It is proposed a semi-empirical correlations in which the
use of the corrective terms as electronegativity and the Shool-
ery shielding effects [30] only affords small improvements
to such approaches.
Likewise, adducts possessing in the molecule chlorine as
well as bromine were obtained by two steps dicarbohalo-
genation (e.g. from compound 5 and compound 11). When
dibromo derivative 5 reacted with CHCl3/NaOH/Triton B a
new dibromodichloro adduct 10 was isolated. On the other
hand, when the tetrachloro adduct 11 was reacted with
HCBr3/NaOH/Triton B a dibromotetrachloro derivative 12
was obtained.
5. Experimental
1H and 13C NMR spectra were recorded on a Varian Unity
1
300 operating at 300.0 MHz for H and 75.0 MHz for 13C.
High resolution and 2D NMR spectra were recorded on a Var-
ian Unity 500 operating at 500.0 MHz for 1H and 125.0 MHz

612
D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
Compound 2, m.p. 187–189 ◦C; C16H18Br2O2; MW 400.
acteristics: first, as an additive term to the equation or the
second one as multiplicatory term. Both approaches are used
in a series of such equations developed after seminal works
of Karplus [31].
MS m/z 400 [M+], 402 [M+ + 2], 404 [M+ + 4], 321, 323, 256
(1 0 0).
Compound 3, m.p. 140–141 ◦C; C17H18O2Br4; MW 570.
MS m/z 570 [M+], 572 [M+ + 2], 574 [M+ + 4], 576 [M+ + 6],
578 [M+ + 8], 491, 493, 495, 497, 411, 413, 415, 388, 171,
256, 129, 57 (1 0 0); IR νmax (cm−1) 2952, 2871, 1776, 1143.
Compound 4, m.p. 203–204 ◦C; C17H18O2Br4; MW 570.
MS m/z 570 [M+], 572 [M+ + 2], 574 [M+ + 4], 576 [M+ + 6],
578 [M+ + 8], 495, 493, 491, 467, 413, 390, 386, 388 (1 0 0),
307, 227; IR νmax (cm−1) 2958, 2929, 1778, 1309, 1207,
1141, 1055.
Acknowledgements
Our special thanks go to MSC Maria Isabel Chavez (I. De
Quimica, UNAM) for the NMR determinations. D. Corona
and S.A.H.P. thank SNI-Conacyt for research fellowship.
Compound 5, m.p. 110–111 ◦C; C16H18O2Br2; MW 400,
MS EI m/z 400, 402 [M+ + 2], 404 [M+ + 4]; IR νmax (cm−1
)
References
3086, 3006, 2943, 1772, 1641, 1323, 1240, 1152.
Compound 6, m.p. 241–243 ◦C; C18H18Br6O2; MW 740.
MS m/z FAB MH+ 741, 743 [MH+ + 2], 745 [MH+ + 4], 747
[MH+ + 6], 749[MH+ + 8], 751[MH+ + 10], 753[MH+ + 12];
665, 664, 663, 661, 648; IR νmax (cm−1) 2962, 2929, 2873,
1776, 1244, 1141.
[1] N.V. Sidwick, The Chemical Elements and their Components, vol.
II, Clarendon Press, Oxford, 1950, p. 1097.
[2] C. Brevard, P. Granger, Handbook of High Resolution Multinuclear
NMR, Wiley, New York, 1981, pp. 106–107.
[3] G.A. Webb, in: P. Laszlo (Ed.), Chemical and Biochemical Appli-
cations, vol. I, Academic Press, New York, 1983, p. 80.
[4] Z. Trontelj, J. Pirnat, L. Ehremberg, in: J.A.S. Smith (Ed.), Advances
in Nuclear Quadrupole Resonance, Heyden, New York, 1974, pp.
71–78.
[5] R.E. Moore, in: P.J. Scheuer (Ed.), Marine Natural Products: Chem-
ical and Biological Perspectives, vol. I, Academic Press, New York,
1978.
[6] T. Irie, M. Susuki, T. Masamune, Tetrahedron Lett. (1965) 1091.
[7] T. Irie, M. Susuki, T. Masamune, Tetrahedron 24 (1968) 4193.
[8] W. Fenical, J.J. Sims, P. Radlick, Tetrahedron Lett. (1973) 313.
[9] W. Fenical, J. Am. Chem. Soc. 96 (1974) 5580.
[10] R. Kazlauskas, P.T. Murphy, R.J. Quinn, R.J. Wells, Tetrahedron
Lett. (1977) 177.
[11] J.J. Sims, S.A. Pettus, R.M. Wing, 1976 Abstr., NATO Conf. Mar.
Nat. Prod.
[12] R. Mathias, P. Weyerstahl, Chem. Berichte 112 (1979) 3041.
[13] E. D´ıaz, H. Barrios, R. Villena, R.A. Toscano, Acta Crystallogr. C
47 (1991) 2720.
Compound 7, m.p. 158–159 ◦C; C16H18O2I2; MW 496
MS FAB m/z MH+ 497, 154 (1 0 0), 136; IR νmax (cm−1
)
3081, 2931, 2871, 1765, 1638, 1237, 1141, 1017.
Compound 8, m.p. 68–70 ◦C; C17H18O2Br2I2; mw 666.
MS m/z 666 [M+], 668 [M+ + 2], 670 [M+ + 4], 391, 359, 337,
307, 263, 205; IR νmax (cm−1) 2954, 2933, 1768, 1303, 1137,
1029.
Compound 9, m.p. 199–200 ◦C; C18H18Br4I2O2; MW
836. MS m/z FAB MH+ 836, 838 [MH+ + 2], 840 [MH+ + 4],
842 [MH+ + 6], 844 [MH+ + 8], 154 (1 0 0), 136; IR νmax
(cm−1) 2960, 2875, 1768, 1137, 1037.
Compound 10, m.p. 148–149 ◦C; C17H18O2Br2Cl2; MW
482. MS m/z 482 [M+], 484 [M+ + 2], 486 [M+ + 4], 488
[M+ + 6]; IR νmax (cm−1) 3080, 2952, 2873, 1776, 1467,
1417, 1327, 1238, 1201, 1143.
[14] E. D´ıaz, H. Barrios, R. Villena, R.A. Toscano, Acta Crystallogr. C
47 (1991) 2723.
[15] I. Salazar, E. D´ıaz, Tetrahedron 35 (1979) 815.
[16] (a) E. D´ıaz, R.A. Toscano, A. Alvarez, J.N. Shoolery, C.K.
Jankowski, Can. J. Chem. 68 (1990) 701;
Compound11, m.p. 243–245 ◦C;C17H18Cl4O2;MW394.
MS m/z 312 [M+ − 82], 314 [M+ − 84], 316 [M+ − 86], 230,
149 (1 0 0), 216 (85), 277 (35), 91 (50).
Compound 12, m.p. 240–242 ◦C; C18H18Cl4Br2O2; MW
564. MS m/z [M+] 564, 566 [M+ + 2], 568 [M+ + 4], 570
[M+ + 6], 572 [M+ + 8], 574 [M+ + 10], 533, 470, 435, 374
(1 0 0), 309 (30), 201 (50), 91 (70).
(b) E. D´ıaz, G.G. Dominguez, A. Mannino, G. Negro´n, C.K.
Jankowski, Magn. Reson. Chem. 23 (1985) 494.
[17] E. D
´ıaz, J.L. Nava, H. Barrios, D. Corona, A. Guzman, M.L. Mucin˜o,
A. Fuentes, Rev. Soc Quim. Mex. 47 (2003) 117.
[18] (a) T. Hiyama, Y. Okude, I.K. Kumura, N. Nozaki, Bull. Chem. Soc.
Jpn. 55 (1982) 561;
(b) E. D´ıaz, A. Ortega, C. Guerrero, A. Guzma´n, B. Lizama, A.
Fuentes, C.K. Jankowski, Nat. Prod. Lett. 12 (1998) 241.
[19] C.K. Jankowski, A.B. Laouz, D. Lesage, E. D´ıaz, Spectroscopy 17
(2003) 735.
[20] (a) J.W. Emsley, J. Feeney, L.H. Sutcliffe, High Resolution in NMR
Spectroscopy, vol. 2, Pergamon Press, 1966;
(b) R.M. Golden, L.C. Stubbs, J. Magn. Reson. 40 (1980) 115.
[21] A. Derome, Modern Techniques for Chemistry Research, Pergamon
Press, New York, 1987, p. 240.
[22] G.E. Martin, A.S. Sektzer, Two Dimensional NMR Methods for
Establishing Molecular Connectivity, VCH Publishers, New York,
1988, p. 219.
6. Calculations
The atomic coordinates used for these calculations were
obtained from the crystallographic data or from the molecular
modelling (Table 2). The homodihedral angles were calcu-
lated by optical superimposition of the first dihedral angle
H Ca Cb Cc on the second dihedral angle Ca Cb Cc X,
which indicates that the carbon Cb is not as important as the
Cc for these estimations.
Two different mode of applications of the electronegativ-
ity as a corrective term to this equation were used to take
into account the difference in the halogen atom nuclear char-
[23] A. Bax, S. Subramanian, J. Magn. Reson. 67 (1986) 565.
[24] A. Bax, J. Freeman, J. Magn. Reson. 44 (1981) 542.
[25] A. Bax, M.F. Summers, J. Am. Chem. Soc. 108 (1986) 2093.

D. Corona et al. / Spectrochimica Acta Part A 62 (2005) 604–613
613
[26] E. D´ıaz, G. Ontiveros, I. Salazar, G. Negro´n, P. Joseph-Nathan, Spec-
trochim. Acta 37A (1981) 569.
[30] (a) B.P. Dailey, J.N. Shoolery, J. Am. Chem. Soc. 77 (1955)
3977;
[27] D. Seyferth, Acc. Chem. Res. 5 (1972) 65.
[28] D. Seyferth, J.Y.P. Mui, J.M. Burlitch, J. Am. Chem. Soc. 89 (1967)
4953.
[29] D. Seyferth, J.M. Burlitch, K.H. James, J. Chem. Soc., Chem. Com-
mun. (1962) 1401.
(b) H. Spiesecke, D.G. Schneider, J. Chem. Phys. 35 (1961)
722;
(c) J.R. Cavanaugh, B.P. Dailey, J. Chem. Phys. 34 (1961)
1099.
[31] M. Karplus, J. Am. Chem. Soc. 85 (1963) 2870.