close

Вход

Забыли?

вход по аккаунту

?

The Stories of Santonin and Santonic Acid.

код для вставкиСкачать
Essays
Early Days of Structure Determination
The Stories of Santonin and Santonic Acid**
Ludmila Birladeanu*
Keywords:
history of chemistry · natural products · reaction
mechanisms · santonin · structure elucidation
S
ince the middle of the 19th century, a
major goal of organic chemistry has been
the determination of the structure of
natural products. Fascination with naturally occurring products dates back to
time immemorial, stimulated by exquisite colors, seductive smells, and often
very remarkable physiological effects.
Before 1859 the concept of structure
did not exist. Only with the invention of
the structural theory by Butlerov, Couper, Kekul&, and van't Hoff could structure determination become a conceivable goal. The late Robert Burns Woodward extolled the creation of structural
theory in these words, “The discovery
that the structure and form of molecules
provide the fundamental basis for the
differentiation of all forms of matter was
one of man's greatest discoveries”.[1]
Today's chemists can scarcely imagine how structures of great complexity
were determined before the availability
of the powerful physical tools of IR and
NMR spectroscopy, mass spectrometry,
and the ultimate weapon, X-ray crystallography. But two earlier stages in the
practice of structure determination can
be distinguished, and are exemplified in
this history of santonin and its provocative derivative, santonic acid.
In the first period, between around
1870 to 1900, characterization depended
on melting points, boiling points, densities, refractive indices, and optical
[*] Dr. L. Birladeanu
Department of Chemistry and Chemical
Biology
Harvard University
Cambridge, MA 02138 (USA)
Fax: (+ 1) 617-495-1792
E-mail: lily@chemistry.harvard.edu
[**] I thank Professor W. von E. Doering for his
suggestions and corrections and Dr. Edmund Keliher for his assistance in preparing the manuscript.
1202
activity. Elemental analysis provided
empirical formulas. Depression of freezing point, elevation of boiling point
provided molecular weights. Major
structural insights came exclusively
from chemical transformations and degradations into more simple, already
recognized structures. Inferences from
such transformations were based on
analogy, empirical approaches, and intuition. Chance—according to Pasteur,
an important ingredient in scientific
advances when its significance is correctly perceived—played its role.
In the second period, as theories of
the mechanisms of reactions were rapidly developing, inferences about structure derived from chemical transformations could be subjected to a new type of
critical evaluation. At the turn of the last
century, organic chemistry began its
rapid change from an empirical, essentially taxonomic science into one dramatically enhanced by theories of the
mechanism by which reactions occurred.
True pioneers, Arthur Lapworth and Sir
Robert Robinson, mark the beginning,
but many years would pass before the
powers of these new insights would be
widely incorporated into the thinking of
natural-product chemists. During both
periods, structural hypotheses received
their final verification through rational
synthesis, and the establishment of the
identity of the synthetic and the natural
product.
The story of santonin and its transformation product, santonic acid, captures the working of these earlier periods. It not only reveals how the limitations of the thinking in the first period
could totally block a successful resolution of structure, but how the new
concepts of mechanism could bring light
into an otherwise impenetrable darkness.
2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4211-1202 $ 20.00+.50/0
Space limitation precludes a comprehensive discussion of both stories,
but since a brief history of the structure
determination of santonin has been
mentioned before,[2] whereas that of
santonic acid has never been reviewed,
the former is limited to its essentials,
while the latter constitutes the main
focus of this essay.
Santonin
Santonin was first isolated by Kahler
in 1830 from Artemisia Santonica, a
member of a family of plants known
for centuries as “wormseed”.[3] The very
name attests to their remarkable efficacy in ridding both human beings and
animals of debilitating round worms
(Nematodes).
Early knowledge of its chemistry
came almost entirely from the investigations the Italian chemist Cannizzaro
and his school carried out in the decades
between 1870 and 1920. Elemental analysis for carbon and hydrogen (oxygen by
difference) established the empirical
formula C15H18O3. The difference of 14
between its 18 hydrogen atoms and the
32 hydrogen atoms required by a saturated hydrocarbon C15H32 divided by
two, implied seven degrees of unsaturation.
The behavior of santonin on treatment with dilute base or acid, as well as
towards reagents regarded as being able
to indicate the presence of a carbonyl
group led the Italian researchers to the
conclusion that santonin is a g-ketolactone. This inference accounted for all
three oxygen atoms and three degrees of
unsaturation (two carbonyl groups and
the ring of the lactone). Reduction with
hydrogen iodide and red phosphorus
gave a monobasic acid, santonous acid,
Angew. Chem. Int. Ed. 2003, 42, No. 11
Angewandte
Chemie
which on distillation with barium hydroxide at 300 8C gave a phenol,
C12H12O, recognized as 1,4-dimethyl-2naphthol. Heating at still higher temperature led to the identification of propene
and propionic acid. A three-carbon fragment, two methyl groups, and the C10
naphthalene now accounted for all
15 carbon atoms in santonin. Two fused
six-membered rings accounted for two
more elements of unsaturation, bringing
the total to five. The two remaining of
the seven degrees of unsaturation required by the empirical formula were
assigned by default to two carbon–
carbon double bonds.
From the recognition of the 1,4dimethyl-2-naphthol and the easy transformation of santonin on treatment with
acid into the isomeric phenol, desmotroposantonin, it was concluded that the
two methyl groups and the keto group
were located in the same ring. The
isolation of propene and propionic acid
placed a methyl group in the ring of the
lactone. These facts led to the proposal
of formulas 1 or 1 a for santonin, with
CH3
CH3
1
O 2
CH3
O
CH3
1a
O
O
3
4
O
5
CH3
1
the correct location of the lactone oxygen atom and of the isopropyl side chain
remaining to be determined.[4] Formula
1 was given preference over 1 a, but
quite arbitrarily. Desmotroposantonin
and santonous acid were assigned structures 2 and 3, respectively.
CH3
CH3
CH3
HO
HO
CO
O
CH3
2
CH3
3
The acceptance of structure 1 was far
from universal. Particularly questionable was the grouping CH2 CO , since
the santonin molecule could not be
made to participate in such characteristic reactions as, for example, aldol
condensation with an aldehyde. The
problem remained unsolved. Today it
Angew. Chem. Int. Ed. 2003, 42, 1202 – 1208
is hard to understand how structure 1
could have been thought to represent a
stable compound like santonin, without
realizing that it was the keto form of a
phenol, desmotroposantonin, which
they had in their hands, and that the
existence of such keto forms in equilibrium with phenols had never been
observed. But then, tautomerism, although discovered decades earlier, was
still not a familiar concept. What we
now take for granted took decades to
become universally accepted.
An even bigger blow to formula 1
came several years later. when Angeli
and Marino isolated a heptanetetracarboxylic acid, C11H16O8, from the mixture
obtained on oxidation of santonin with
permanganate.[5] This acid behaved like
a disubstituted malonic acid and thus
contained a quaternary carbon. But
formulas 1 or 1 a did not contain such a
carbon atom! Unfortunately, the numerous structures subsequently proposed to
accommodate the quaternary carbon
were no better and so 1 continued to
be accepted by default.
In a 1924 review Angeli
stressed again the significance of
O
the existence of a quaternary
carbon in the isolated heptaneCH3
tetracarboxylic acid.[6] The chemists who had been involved for
decades in work on the structure
of santonin were fully aware that
they had not succeeded in solving
the problem.[7]
The problem lay dormant for many
years until a solution was finally provided by Clemo and his co-workers in
1929.[8] Armed with a greatly increased
awareness and understanding of the
keto–enol equilibrium,[9] and rearrangements of carbon skeletons,
often induced by acidic condiCH3
tions,[10] Clemo was able to
COOH
state that “on general theoretical grounds, structure 1
should be an unstable form of
the stable phenolic isomeride
desmotroposantonin” and to
propose for santonin a new structure,
4, based on the assumption that the two
methyl groups in santonin were not
situated in the same ring. By shifting
the methyl group from position 4 to
position 5 he obtained a structure containing a quaternary carbon, which on
treatment with acid could undergo a
2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
CH3
CH3
O
CO
O
CH3
4
Isoprene Units
Selinene
dienone–phenol rearrangement to the
phenol desmotroposantonin. In addition, structure 4 could be constructed
on the principle of Ruzicka's isoprene
rule, showing a close relationship to the
terpene selinene.
The correctness of the assumptions
made by Clemo et al. were proven by
the synthesis of racemic santonous
acid,[11] and racemic desmotroposantonin,[12] both readily obtainable from
natural santonin. Santonous acid, obtained by a six-step synthesis, was identical in all respects with the natural one.
In addition, by this synthesis the b position for the acidic side chain was now
firmly established.
Based on the structure of santonous
acid, two formulas could now be suggested for desmotroposantonin, depending on whether the lactonic oxygen
was assigned to position 7 (I = 2) or 9
(II). Clemo et al. resolved the problem
CH3
CH3
CH3
9
HO
O CO
HO
CH3
CO
7
O
I
CH3
CH3
II
by synthesizing a compound of structure
II which turned out to be identical to a
sample of natural racemic desmotroposantonin.
Consistent with structure II for desmotroposantonin, the structure of santonin had to be rewritten as 4 a. This
structure became now universally accepted.
CH3
O CO
CH3
O
CH3
Santonin (4a)
1433-7851/03/4211-1203 $ 20.00+.50/0
1203
Essays
But there still remained a dark
shadow lingering over the santonin
problem. Although Clemo's formulation permitted a decisive clarification
of the results obtained both on oxidative
degradation and on treatment with acid,
it was unable to account for the behavior
of santonin on treatment with strong
base. Whereas in dilute sodium hydroxide santonin gave the isolable salt of
santoninic acid, from which it was easily
regenerated on acidification, treatment
with more concentrated sodium hydroxide gave an isomeric acid, santonic acid,
with totally different properties. “While
there was no longer any question of the
validity of formula 4 a, the situation
remained somewhat unsatisfactory in
so far as all attempts at a rational
formulation of santonic acid and the
extensive series of substances derived
from it by degradation have failed”.[21]
Not surprisingly, the elucidation of the
structure of santonic acid became the
most intriguing aspect of the chemistry
of santonin.
The Structure of Santonic Acid
One of the first transformation products of santonin was discovered by
Hvoslev in 1863,[13] and rediscovered
some ten years later by Cannizzaro and
Sestini[14] in apparent unawareness of
the previous work (means for searching
the “literature” did not become available until many years later). As shown
above, Hvoslev had found that the
action of concentrated sodium hydroxide on santonin gave rise to an acid,
isomeric with santoninic acid, named
santonic acid. The ease and simplicity of
its formation from santonin prompted
many to investigate its chemistry in the
hope of arriving at a structure of santonin. The Cannizzaro school, mainly at
the hands of Francesconi, devoted almost fifty years to the effort.
So fruitless were these efforts that
santonic acid played no role whatsoever
in the elucidation of santonin, but instead became itself an even thornier
problem. One might have expected the
structure of santonic acid to become
clear once the correct structure of santonin had been elucidated by Clemo
et al. But, “No”! The problem remained
as intractable as ever. The state of affairs
1204
in 1934 was reviewed in detail by
Wedekind and Engel.[15] Here only the
few well established facts relating to
santonic acid are presented.
From the empirical formula,
C15H20O4, santonic acid was seen to
contain six elements of unsaturation
and to be isomeric with santoninic acid,
the product of the mild alkaline hydrolysis of santonin. Titration and analysis of
its salts established santonic acid as
monobasic, and thereby accounted for
two oxygen atoms and one degree of
unsaturation. Unlike santoninic acid,
which rapidly lactonized to santonin,
santonic acid did not revert to santonin
on acidification but remained unchanged.[16] Treatment with hydroxylamine gave a monoxime under normal
conditions and a dioxime under more
drastic conditions.[17] The remaining two
oxygen atoms, therefore, were present
as ketones of differing reactivity, and
accounted for two more degrees of
unsaturation. Unlike santonin and santoninic acid, santonic acid did not add
bromine or undergo catalytic hydrogenation. It was oxidized only very slowly
by permanganate, and was not reduced
by hydrogen iodide and phosphorus, a
treatment which had converted santoninic acid into santonous acid. It was
concluded that santonic acid did not
contain carbon–carbon double bonds,
and must therefore be saturated and
tricyclic, if the final three degrees of
unsaturation were to be accommodated.
The relation of the more reactive,
less hindered carbonyl group and the
carboxylic acid was revealed by the
behavior of a reduction product, a
secondary alcohol, dihydrosantonic
acid. Among the products formed on
treatment with acetic anhydride was an
acetyl lactone, which could be hydrolyzed to a hydroxy acid and reclosed to
the lactone with the ease normally
associated with a g-lactone.[18] This behavior placed the carbonyl group of
santonic acid no further than two carbon
atoms away from the carboxylic acid
group.
Vigorous oxidation by potassium
permanganate transformed santonic
acid into santoric acid, C13H18O8, a
saturated, monocyclic tetrabasic acid.
This compound reacted with acetic anhydride to afford a monoanhydride and
two isomeric dianhydrides. Fusion of
santoric acid with sodium hydroxide at
300 8C gave (along with acetic acid,
carbon dioxide, and hydrogen) santoronic acid, C10H16O6, a saturated, consequently acyclic tricarboxylic acid. On
further heating with potassium hydroxide at still higher temperature santoronic acid was transformed into a ketone,
santorone, C8H14O, by loss of two more
carbon atoms.[19] These products were of
unknown structure at the time and made
no contribution to the proposed skeleton of santonic acid. Only the barest
partial structure could be written
(Scheme 1). More detailed structures
C
C10H19
O
C O (OH)
C
C C OH
O
Scheme 1. Imprecise partial structure of
santonic acid.
were proposed, but, being based on
erroneous structures for santonin, were
themselves of no value.
After the establishment of the structure of santonin by Clemo et al., Wedekind (and others) attempted further
degradations of santonic acid, but the
reaction products remained obstinately
unidentifiable.[20] The structure of santonic acid continued to remain elusive.
The classical approach of the first period
which involved reassembling a structural hypothesis from various bits of information gleaned from smaller components produced by degradation, and
from studies of the response to chemical
reagents, was not working in the case of
santonic acid. As we saw, even in the
case of santonin, Clemo had to place
critical reliance on the knowledge of the
mechanisms of the Wagner–Meerwein
rearrangement and of base-catalyzed
tautomerization, thus providing an early
example of the impact of the emerging
mechanistic way of thinking about
chemical transformations on structure
determination.
In the case of santonic acid, the time
had not yet come when mechanistic
analysis alone would lead to the complete elucidation of structure. The problem lay dormant for more than a decade
until R. B. Woodward took up the
Angew. Chem. Int. Ed. 2003, 42, 1202 – 1208
Angewandte
Chemie
challenge in the early 1940s.[21] In contemplating the changes that had been
brought about by hydroxide ion in the
transformation of santonin to santonic
acid, he was struck by several observations that seemed particularly in need of
rationalization. In the overall change,
the two carbon–carbon double bonds of
santonin had disappeared, one to reappear as a carbon–oxygen double bond at
the cost of the incipient hydroxy group
of the lactone ring, the second to appear
as a third ring. He was further impressed
that the heptanetetracarboxylic acid
isolated by Angeli and Marino from
the alkaline permanganate oxidation of
santonin still contained a quaternary
carbon. This observation implied that
the methyl group at the 5-position in
santonin had not rearranged to the 4position as it had under the influence of
acids. Woodward was convinced that the
transformation by alkali was unique to
santonin, unlikely to be the result of a
simple process, and could be resolved
“by a rational consideration of the
effects to be anticipated in the action
of strong bases on the santonin molecule”. His approach was based on his
life-long faith that chemical transformations could be understood through insight into their mechanism.
By the time Woodward tackled the
santonic acid problem, giving it his
almost undivided attention in the early
months of 1943,[22] the mechanisms of a
large number of base-catalyzed reactions had been elucidated, most notably
through the ground-breaking achievements of Arthur Lapworth.[23] Systematically and exhaustively, Woodward
analyzed the consequences of addition
of hydroxide ion to each of the several
electrophilic carbon atoms in santonin
and the consequences of removal of
acidic hydrogen atoms, no matter how
weakly they might appear to be activated.
The first response of santonin to
hydroxide ions had been the addition to
the carbonyl group of the lactone ring to
give the sodium salt of santoninic acid.
The problem then was to predict the
further changes that a hydroxide ion
might initiate in this acid. For example, a
Michael-type addition to the electrophilic b carbon atom of either of the
a,b unsaturated carbonyl systems could
be followed by a retroaldol cleavage to
Angew. Chem. Int. Ed. 2003, 42, 1202 – 1208
an aldehyde and methyl ketone, and
these in turn could lead to further
transformations. But such journeys
ended only in blind alleyways and certainly not in a tricyclic system.
Turning to the consequences of abstraction of acidic hydrogen atoms,
O
H3C
H
CH3
O
O
Santonin
10
CH3 HO
9
CH
H
1
H
CH3
(1)
8
OH
H3C
O
4
H3C
H3C
HO
O
H3C
hydroxy groups in santoninic acid were
identical to those in 3-hydroxy-6-ketocholestene. He assumed that the same
sequence of events could take place in
santoninic acid and would lead to the
diketone III [Eq. (1)]. The attractiveness of this hypothesis was irresistible:
COOH
Santoninic Acid
O
H C 11
COOH
H3C
H
III
Woodward focused his attention on the
hydrogen atom attached to the C-9
carbon atom bearing the hydroxy group.
At first glance, this hydrogen atom
might not have appeared to be particularly acidic—unless one were familiar
with Lapworth's demonstration that a
g hydrogen atom in an a,b-unsaturated
carbonyl system is as acidic and easily
removable by base as the a hydrogen
atom of a carbonyl compound. An
appropriate example, known to Woodward, had been discovered independently by Butenandt and Schramm,[24a]
and Heilbron et al.[24b] It involved conversion of 3-hydroxy-6-ketocholestene4 with hydroxide ions into 3,6-diketocholestane (Scheme 2).
the unsaturation in one carbon–carbon
double bond transformed the hydroxy
group of santoninic acid into the obligatory second carbonyl group, and
thereby provided an intermediate containing four acidic hydrogen atoms,
three of them a to carbonyl groups, the
fourth only weakly activated by a carboxylate anion (see III).
The task then was to discover how
the carbon–carbon double bond remaining in III might give rise to the tricyclic
system present in santonic acid. Any of
the four carbanions hypothetically derivable from III by extraction of hydrogen atoms, C-1, C-10, C-8, and C-11,
could in principle engage the a,b unsaturated ketone in an intramolecular Michael condensation at the
b C-4 atom to generate a
H3C
H3C
third ring by formation of
OH–
a new carbon–carbon
γ
α
–
bond. Intramolecular adHO
HO
β
ditions
to
carbonyl
O
O
3-Hydroxy-6-oxocholest-4-ene
groups were not considα,β
ered on the grounds that
they would destroy a carH3C
H3C
bonyl group and leave
the carbon–carbon douO
HO
ble bond intact.
O
O
The results of the four
3-Hydroxy-6-oxocholest-3-ene
3,6-Dioxocholestane
β,γ
possible intramolecular
Michael condensations
Scheme 2. Rearrangement of 3-hydroxy-6-oxocholest-4-ene
are the structures a–d.
induced by the abstraction of a proton.
Two of the hypothetical
structures, a and b, would
In this instance the b,g isomer is also contain strained four-and three-memthe enol of a g diketone and, conse- bered rings (~ 26 kcal mol 1), respectivequently, is transformed essentially irre- ly, and thus would not be favored at
versibly
into
3,6-diketocholestane. equilibrium in a reversible Michael reWoodward recognized that the relative action. Woodward recognized that the
positions of the double bond, keto, and carbanion at C-8 leading to c was five
1205
Essays
–
–
O
H3C
H3C
CH3
O
H3C
CH
a COOH
CHCH3
O
H3C
CH
COOH
b
carbon atoms removed from the C-4 of
the double bond, a potentially ideal
relation for an intramolecular Michael
reaction. However, it was not clear if
these two carbon atoms were close
enough in space to allow the reaction
to occur. The same doubt applied to the
fourth possibility, d. Unlike the previous
examples of intramolecular Michael additions, first observed by Guthzeit,[25]
that could be easily visualized in two
dimensions, a planar representation of c
obscured its viability. Only when Woodward had rethought the required bond
making in three dimensions did the
difficulty disappear. In the three-dimensional model, the distance between C-8
and C-4 was no longer prohibitive, but
ideal for generation of the new bond.
The resulting structure IV (or c in two
dimensions) became a credible and
attractive structure for santonic acid.
The structure IV contained the same,
only slightly strained bicyclo[2.2.1]heptane system found in camphor. It was
not expected to suffer further changes in
alkali, and could be reasonably considered to be the mechanistically logical
end product of the treatment of santonin
with strong base.
But did the mechanistically attractive, hypothetical structure IV actually
represent santonic acid? The experimental facts presented above were certainly satisfied by structure IV: It represented a saturated, tricyclic, monobasic acid having a more sterically hindered ketone group (at C-2) and a less
hindered ketone group (at C-9). More
convincing evidence for structure IV
came from one of the degradation
experiments carried out by Francesconi.
Permanganate oxidation of santonic
acid had given a-santoric acid, a saturated, monocyclic tetrabasic acid,
C13O18O8. In terms of IV, this degradation would have occurred by oxidative
cleavage along the indicated dotted lines
in Scheme 3 by way of the enol tautomers of the ketones, and would be
expected to give a compound of structure V.
1206
–
O
–
O
stabilized carbanions. With this mechanistic hypothesis in mind, he envisioned
the following sequence of events
(Scheme 4): during alkali fusion loss of
a proton from the carbon atom a to a
carboxylate anion would give VI; breaking of a carbon–carbon bond by a
reverse Michael reaction[26] indicated
by either arrow a or b, both giving the
same end result; cleaving the resulting
a,b unsaturated compound VII by the
two-step sequence of addition of a
hydroxide ion and retro-aldol-like condensation
of
the
b hydroxycarboxylate anion to generate VIII and
O
H3C
H3C
CH3
CH3
O
O
CH
H3C c COOH
C
H3C
d COOH
On treatment with acetic anhydride,
a-santoric acid gave, among other products, a monoanhydride from which it
was regenerated by hydrolysis. This
anhydride was converted into an isomeric, thermally stable anhydride on
further heating. Hydrolysis of this second anhydride gave b-santoric acid, an
O
H
4 H3C
6
7
5
H3C
H
8 9
CH
CH2COOH
CH3
H
CH3
10
COOH
COOH
KMnO4
O
COOH
CH
H3C
COOH
IV
COOH
CH2COOH
H3C
H
COOH
CH
H3C COOH
α-Santoric Acid
V
Scheme 3. Suggestions for the formation of a-santoric acid from the hypothetical structure of
santonic acid IV.
–
–
–
COO
H3C
COO
H3C
isomer of a-santoric acid. H3C COO –
–
–
CH
–
CH CHCOO
CHCOO
COO –
This complicated behavior
a
–
HC
–
COO
COO
b
–
COO
–
was also explicable in terms
COO
COO
VIII CH
VI CH
VII
–
CH
of structure V by hypotheCOO
CH3
CH3
CH3
sizing epimerization of the
hydrogen atom in bold face
COOH
H3C
CH
in Scheme 3 during the
COOH
transformation to b-santorCH
COOH
IX CH
ic acid.
CH3
A more critical test was
provided by Francesconi's
Scheme 4. Woodwards explanation for the conversion of
finding that fusion of either santoric acid into santoronic acid.
a- or b-santoric acids with
alkali generated acetic acid,
carbon dioxide, and santoronic acid, a an acetate ion; (a precedent for such a
tribasic, saturated, optically inactive change was known in the conversion, on
acid of empirical formula C10H16O3, alkali fusion, of oleic acid to palmitic
m.p. 127 8C. Was the structure V pro- acid).[27] As the dicarboxylate of a maposed for santoric acid compatible with lonic acid, VIII could reasonably be
this degradation product? At the time expected to undergo decarboxylation at
no experience was available in the the high temperature of the fusion to
literature on the basis of which the give, after neutralization, the saturated,
course of the high-temperature alkali acyclic tribasic heptane-2,3,6-tricarboxfusion of saturated, cyclic polyacids ylic acid, C10H16O6, of structure IX.
could be predicted. Woodward hypothe- According to Woodward's mechanistic
sized that at the high temperature of analysis, this should be the correct
alkali fusion, loss of a very weakly acidic expression for Francesconi's santoronic
hydrogen a to a carboxylate ion could acid. But it could also be the structure of
generate a dicarbanion, that would be- the heptane tricarboxylic acid of m.p.
have similarly to conventional better- 88 8C that Angeli and Marino had ob-
Angew. Chem. Int. Ed. 2003, 42, 1202 – 1208
Angewandte
Chemie
tained by permanganate oxidation of
santonin, or of the acid of m.p. 137 8C
that had been synthesized by Ruzicka to
establish the position of the angular
methyl group in santonin.[28] On the
resolution of this dilemma depended
the credibility of the edifice of Woodward's analysis.
Woodward, Brutschy, and Baer resolved the dilemma by repeating Ruzicka's synthesis of the racemic acid of
m.p. 137 8C and finding that on fusion
with alkali, it was indeed isomerized to
the same acid of m.p. 127 8C that Francesconi had obtained from the fusion of
santoric acid, that Angeli and Marino
had obtained on fusion of the acid of
m.p. 88 8C, and that Ruzicka also had
obtained from the fusion of his synthetic
acid.[28] On repeating the work of Angeli
and Marino, Woodward et al. confirmed
that the optically active acid of m.p.
88 8C gave the racemic acid of m.p.
127 8C on alkali fusion. With this last
obstacle removed, structure IV for santonic acid turned out to be, in Woodward's words, “the most satisfactory
vehicle for the rationalization of the
hitherto baffling chemistry of santonic
acid”.
It is impossible to overestimate the
beauty, elegance, and impact of Woodward's mechanistic scheme. Not only did
it solve the elusive structure of santonic
acid, but it illustrated the power of the
mechanistic approach, and the importance of thinking in terms of threedimensional models. The market for
Dreiding models and later 3-D computer models was assured!
Incidentally,
this
achievement
served as a basis for a new approach to
the construction of a complicated ring
system. An example is the historically
important synthesis of longifolene by
Corey et al.[29] Corey had conceived a
homodecalin system, to serve as the
precursor of an intramolecular Michael
addition (Scheme 5). “Although this
O
O
change would seem unusual and perhaps
unreasonable, some confidence in its
validity seemed indicated from the remarkably smooth transformation of santonin to santonic acid under the influence of alkali”.
In the third and current phase of
structure determination, physical methods have displaced both the first, classical phase, and the second phase in which
the power of careful mechanistic analysis of chemical transformations played
such an important role. NMR spectroscopy and X-ray single-crystal analysis as
the ultimate weapon have made structure determination a routine process
requiring little or no “chemical intuition”. One may well ask whether there
will be stories like that of santonin and
santonic acid to tell in the future.
Probably not.
Santonin is an easily accessible,
crystalline natural product, the structure
of which can (and has been) determined
by X-ray analysis in a matter of days,
rather than the decades of chemical
investigation. Santonic acid is not a
natural product, but rather the product
of the chemical transformation of santonin on vigorous boiling with concentrated alkali, one of the many reactions
carried out in the course of the classical
search for transformations leading to
useful structural information. Would
this experiment be done today? Probably not. One could argue, that the more
rapid progress resulting from the availability of modern tools is infinitely more
important than a missed chemically
provocative experiment. I will let Sir
Robert Robinson give the answer:
“….there is a price to pay and that is
the short-circuiting of chemical investigation, because it resulted in the loss of
general and important chemical information, and substitution by rather mechanical means of the complex intuitive
puzzle-solving processes”.[30] No better
testimony to the validity of this prognosis is needed than Woodward's santonic acid paper.[31]
O
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
CH2
CH3
CH3
O
[5]
[25]
Five steps
C
[3]
[4]
CH3
Longifolene
Scheme 5. Corey's proposal for the synthesis
of Longifolene under following the conversion
of santonin into santonic acid.
Angew. Chem. Int. Ed. 2003, 42, 1202 – 1208
[1] “Art and Science in the Synthesis of
Organic Compounds: Retrospect and
Prospect”: R. B. Woodward in Pointers
and Pathways in Research, CIBA of
India, Bombay, 1954.
[2] J. B. Hendrickson, The Molecules of
Nature (Ed.: R. Breslow), Benjamin
[26]
[27]
[28]
Reading, Massachusetts, 1973, pp. 111 –
112.
M. Kahler, Arch. Pharm. 1830, 34, 318.
S. Cannizzaro, P. Gucci, Rend. Accad.
Naz. Lincei Ser. 5 1892, 1, 149.
A. Angeli, L. Marino, Rend. Accad. Naz.
Lincei Ser. 5 1907, 16, 385.
A. Angeli, Rend. Accad. Naz. Lincei Ser.
6 1924, 33, 10.
The essential papers pertinent to this
work can be found in the bibliography to
ref. [21].
G. R. Clemo, R. D. Haworth, E. J. Walton, J. Chem. Soc. 1929, 2368 – 2387;
G. R. Clemo, R. D. Haworth, E. J. Walton, J. Chem. Soc. 1930, 1110 – 1115.
C. K. Ingold, Structure and Mechanism
in Organic Chemistry, Cornell University Press, Ithaca, 1953, pp. 794 – 802.
L. Birladeanu, J. Chem. Educ. 2000, 77,
858 – 863.
S. Cannizzaro, J. Carnelutti, Ber. Dtsch.
Chem. Ges. 1880, 13, 1516 – 1517.
A. Andreocci, P. Bertolo, Ber. Dtsch.
Chem. Ges. 1898, 31, 3131 – 3133.
H. Hvoslev, F,rhandlingar vod Skandinaviska Naturforsharem,tat 1863, 304.
S. Cannizzaro, F. Sestini, Gazz. Chim.
Ital. 1873, 3, 241 – 251.
E. Wedekind, O. Engel, J. Prakt. Chem.
1934, 139, 115 – 140.
S. Cannizzaro, Gazz. Chim. Ital. 1876, 6,
355; F. Sestini, Gazz. Chim. Ital. 1876, 6,
148.
L. Francesconi, Gazz. Chim. Ital. 1899,
29ii, 181.
S. Cannizzaro, F. Sestini, Gazz. Chim.
Ital. 1876, 6, 345.
L. Francesconi, Gazz. Chim. Ital. 1892,
22i, 181; L. Francesconi, Gazz. Chim.
Ital. 1893, 23ii, 452; L. Francesconi,
Gazz. Chim. Ital. 1899, 29ii, 224; L.
Francesconi, Rend. Accad. Naz. Lincei
Ser. 5 1896, 5ii, 214.
K. Tettweiler, O. Engel, E. Wedekind,
Liebigs Ann. Chem. 1932,492, 105.
R. B. Woodward, F. G. Brutschy, H.
Baer, J. Am. Chem. Soc. 1948, 70,
4216 – 4221.
W. von E. Doering, private communication.
M. Saltzman, J. Chem. Educ. 1972, 49,
750 – 753.
a) A. Butenandt, G. Schramm, Ber.
Dtsch. Chem. Ges. 1936, 69, 2289 –
2299; b) I. M. Heilbron, E. R. H. Jones,
F. S. Spring, J. Chem. Soc. 1937, 801 –
805.
M. Guthzeit, Ber. Dtsch. Chem. Ges.
1898, 31, 2753 – 2758; M. Guthzeit, Ber.
Dtsch. Chem. Ges. 1901, 34, 675 – 680.
C. K. Ingold, W. J. Powell, J. Chem. Soc.
1921, 119, 1976 – 1982.
F. G. Edmed, J. Chem. Soc. 1898, 627 –
634.
L. Ruzicka, L. Steiner, Helv. Chim. Acta
1934, 17, 614 – 621.
1207
Essays
[29] E. J. Corey, M. Ohno, R. B. Mitra, P. A.
Vatakencherry, J. Am. Chem. Soc. 1964,
86, 478 – 485.
[30] Robert Robinson, as quoted by in “Sir
Robert Robinson: A Contemporary
Historical Assessment and a Personal
Memoir”, A. J. Birch, J. Proc. R. Soc. N.
S. W. 1976, 109, 151 – 160.
[31] It must be mentioned that Clemo's work
resulted in the determination of the
gross structure of santonin, but had
nothing to say about its detailed stereo-
1208
chemistry. It took quite a few years to
establish the absolute configuration of
the natural ( )-a-santonin, one of the 16
theoretically possible stereoisomers.
This topic is not pursued here. The
interested reader may find a complete
list of references in J. W. Huffman, W. T.
Pennington, D. W. Bearden, J. Nat.
Prod. 1992, 55, 1087 – 1092). The absolute configuration of santonic acid was
firmly established in 1999 by single
crystal X-ray analysis; A. P. J. Brunskill,
H. W. Thompson, R. A. Lalancette, Acta
Crystallogr. Sect. C 1999, 55, 566 – 568.
The two configurations are:
H3C H2
C
CH3
H
CH2
CH3
H
O
CH3
C 11
CH
H
CH3
C
O
(–)-α-Santonin
O
O
11
H
O
COOH
CH3
Santonic Acid
Angew. Chem. Int. Ed. 2003, 42, 1202 – 1208
Документ
Категория
Без категории
Просмотров
2
Размер файла
405 Кб
Теги
acid, stories, santoni, santonici
1/--страниц
Пожаловаться на содержимое документа