Synthesis and reactivity of nonstabilized diazo sugars

Article abstract of DOI:10.1016/j.carres.2006.11.003

1,6-Anhydro-4-deoxy-4-diazo-2,3-O-isopropylidene-β-d-lyxo-hexopyranose (4) is a stable crystalline compound readily accessible by an improved synthetic procedure. It has been used as a model for evaluating the reactivity of the diazo group, when not stabilized by an adjacent carbonyl function, in a rigid chiral matrix. A range of carbene-type, electrophile-promoted, and 1,3-dipolar reactions were evaluated, leading to 4,4′-alkene dimers, 4-deoxy-3-enose and related derivatives, 4,4-dihalo compounds, 4-spirocyclopropane derivatives, 4-spiropyrazole structures, and by skeletal rearrangement, branched-chain anhydropentose structures having a bicyclo[2.2.2] skeleton.

Full text of DOI:10.1016/j.carres.2006.11.003

Carbohydrate Research 342 (2007) 31–43
Synthesis and reactivity of nonstabilized diazo sugars
Michael S. Alexander and Derek Horton*
Department of Chemistry, American University, 4400 Massachusetts Avenue, NW, Washington, DC 20016, USA
Received 28 September 2006; received in revised form 31 October 2006; accepted 1 November 2006
Available online 10 November 2006
Abstract—1,6-Anhydro-4-deoxy-4-diazo-2,3-O-isopropylidene-b-D-lyxo-hexopyranose (4) is a stable crystalline compound readily
accessible by an improved synthetic procedure. It has been used as a model for evaluating the reactivity of the diazo group, when
not stabilized by an adjacent carbonyl function, in a rigid chiral matrix. A range of carbene-type, electrophile-promoted, and
1,3-dipolar reactions were evaluated, leading to 4,40-alkene dimers, 4-deoxy-3-enose and related derivatives, 4,4-dihalo compounds,
4-spirocyclopropane derivatives, 4-spiropyrazole structures, and by skeletal rearrangement, branched-chain anhydropentose struc-
tures having a bicyclo[2.2.2] skeleton.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Keywords: Diazo sugars; gem-Dihalo sugars; Dimeric sugars; Pyrazole; Spirocyclopropane; Branched-chain sugars
1. Introduction
the action of cupric oxide.11 Such stabilized diazo deriv-
atives are not decomposed in acetic acid solution,12
although displacement of nitrogen occurs with stronger
acids.10
Diazomethane and diazoethane have long been used
for the alkylation of acidic and enolic hydroxyl groups
in sugars1,2 and for the esterification of aldonic acids.3
A classic synthesis of 2-ketoses involves reaction of
acetylated aldonyl chlorides with two moles of dia-
zomethane to generate terminal diazoketones and
thence the ketose derivatives.4–7 The canonical forms
of the diazo group (Scheme 1) indicate a high negative
polarity on the carbon atom adjacent to nitrogen, and
consequently derivatives having a carbonyl or ester
group adjacent to the diazo function are greatly stabi-
lized8 by comparison with those derivatives lacking
such stabilization.
Typical reactions of the diazo group include loss of N2
to form transient carbenes13 or carbenoid species in
either the singlet or triplet state,14–16 which may react
intramolecularly or intermolecularly in a variety of
modes.17–27 Singlet carbenes, generally formed thermally
or photochemically, tend to react indiscriminately,28–32
whereas triplet (diradical) carbenes33,34 show greater
selectivity.35,36 Carbenoids, formed catalytically by the
action of copper salts on diazo compounds,37–40 are of
lower energy, typically reacting with unsaturated func-
tions to form cyclopropanes,41,42 and may display car-
bocation-like character.43 Carbenes and carbenoids
have been reviewed in detail.44–47
Thus terminal diazo-2-ketose derivatives may be
reduced to 1-deoxyketoses,9 converted by HCl or HBr
into 1-deoxy-1-haloketoses,10 and converted into
dimeric C@C linked structures with loss of nitrogen by
The diazo function can also act as a 1,3-dipole, react-
ing with p systems to form pyrazoles. Furthermore, it
..
N
+
-
-
+
-
+
:
C
N
N
..
:
:
N
:
N
: C
N
C
Scheme 1.
*
Corresponding author. Tel.: +1 202 895 1767; fax: +1 202 895 1752; e-mail: carbchm@american.edu
Present address: Bioanalytical Systems, Inc., McMinnville, OR 97128, USA.
0008-6215/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2006.11.003

32
M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
may be considered formally as a deprotonated diazo-
nium ion, so that in a protic environment it is suscepti-
ble to protonation at carbon with rapid loss of N2 to
generate a ‘hot’ carbocation. The latter can be attacked
by a nucleophile, or undergo deprotonation to form an
alkene, or be stabilized by carbon-skeleton rearrange-
ment before reacting with a nucleophile.48,49
Our previous studies on the reactivity of 2-deoxy-
2-diazoaldonic acid ester derivatives50demonstrated
their conversion upon thermolysis or photolysis into
unsaturated or deoxygenated structures, presumably
via carbene-type intermediates, as well as 1,3-dipolar
cycloaddition of phenylacetylene to generate pyrazole
derivatives. The action of strong base leads to chain deg-
radation with loss of the ester group and formation of a
chain-terminal acetylenic sugar, again presumably via a
carbene species.51 Subsequent work52,53 has shown the
conversion of 2-deoxy-2-diazoaldonic esters into cyclo-
propanes by carbenoid-type insertion into alkenes, their
intramolecular reaction with alcohols to form cyclic
ethers,54 and a range of insertion reactions involving
diazoester substituents.55–57
There is little in the literature on diazo sugars where
the diazo functionality is not stabilized by an adjacent
ketone or ester group. Earlier work from this labora-
tory58 applied the Bamford–Stevens reaction on several
2,4,5-trichlorobenzenesulfonylhydrazones of protected
sugars, affording nonstabilized diazo derivatives having
the diazo group at the 2-, 3-, and 4-positions of a hexose
chain. The products were highly reactive, of generally
limited stability, and difficult to prepare in quantity.
Nonstabilized 1-diazo sugar derivatives have been em-
ployed in 1,3-dipolar addition reactions with alkynes
as a route to C-nucleoside analogues,59,56 and recent
work61 has described other applications of diazo sugars
in the generation of heterocyclic structures.
The present work62–64 reports an improved synthesis
of the 4-diazo derivative, 1,6-anhydro-4-deoxy-4-diazo-
2,3-O-isopropylidene-b-D-lyxo-hexopyranose (4), a well-
crystallized compound accessible on a larger scale
without recourse to chromatography and amenable to
storage for several days. This compound was used as
a model to evaluate the reactivity of the nonstabilized
diazo function, embedded in a rigid chiral matrix, in a
variety of representative reactions of potential utility
in synthetic transformations of sugars.
2. Results and discussion
2.1. Preparation of 1,6-anhydro-4-deoxy-4-diazo-2,3-O-
isopropylidene-b-D-lyxo-hexopyranose (4)
The starting compound for this work, 1,6-anhydro-2,3-
O-isopropylidene-b-D-mannopyranose (1) was obtained
by pyrolysis of ivory-nut mannan and acetonation of
the resultant crude 1,6-anhydro-b-D-mannopyranose
by a revision of the procedure of Hudson and co-work-
ers.65 Oxidation of 1 by ruthenium tetraoxide by a min-
or modification of an earlier procedure66 afforded the
corresponding 4-ketone, 1,6-anhydro-2,3-O-isopropyl-
idene-b-D-lyxo-hexopyranos-4-ulose (2) in 85% yield.
In our earlier work,58 this ketone was converted into
its 2,4,5-trichlorobenzenesulfonylhydrazone, the latter
being deprotonated by sodium hydride in hexane,
followed by pyrolysis of the sodium salt under high
vacuum in a sublimation apparatus to give 1,6-an-
hydro-4-deoxy-4-diazo-2,3-O-isopropylidene-b-D-lyxo-
O
O
O
O
O
O
O
pyrolyze
Me2CO
O
N H
O
O
O
RuO
2
4
Pb(OAc)
N
4
O
O
4
O
O
D-Mannan
O
N
N
O
HO
H N
2
C
C
C
C
Me
Me
Me
2
Me
2
2
2
2
3
4
1
CuI
O
3
O
O
O
O
O
O
O
MeOH, Pt
O
O
O
O
O
+
O
O
MeO
Me C
2
O
O
Me C
C
2
C
Me
Me
2
2
7
6
5
Scheme 2.

M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
33
hexopyranose (4) as yellow needles, collected on a cold
finger. While this synthesis is adequate for the prepara-
tion of small amounts of material, it has shortcomings:
the overall yield is low, and scaling up the reaction to
multigram quantities is difficult. Accordingly, an alter-
native synthesis was devised (Scheme 2).
products, with a major fraction having Rf 0.95 that was
isolated by chromatography and recrystallized from 2-
propanol as short colorless needles in 62% yield. Its mass
spectrum showed a parent ion at m/z 368, clearly indicat-
ing that the product was a dimer resulting from the cou-
pling of two molecules of 4 with concurrent loss of two
molecules of nitrogen. Its IR spectrum gave no evidence
of a double bond, which is not surprising for a highly
symmetrical tetrasubstituted alkene, but ozonolysis at
ꢀ78 ꢁC in dichloromethane–pyridine gave a single, ana-
lytically pure product having physical constants identical
to those of the glycos-4-ulose derivative 2.
From this evidence, this product was formulated as
the unsaturated dimer, 1,6;10,60-dianhydro-4,40-dide-
oxy-2,3;20,30-di-O-isopropylidene-bi(b-D-lyxo-hexopyr-
anos-4-ylidene) (5). In support of this assignment, the
1H NMR spectrum showed the presence of only six
apparent proton signals, indicating the highly symmetri-
cal form of the molecule. It is noteworthy that H-5 and
H-6endo show essentially zero coupling; inspection of
molecular models show that the dihedral angle between
these protons is approximately 90ꢁ (Scheme 3), and this
is in accord with observations on many related com-
pounds having this same bicyclo[3.2.1] skeleton.68–70
The dimer 5 is strongly levorotatory. The (Z) config-
uration about the double bond is depicted, but this point
was not established by independent criteria. Attempts to
reduce the alkene 5 by catalytic hydrogenation were
unsuccessful, and the starting material was recovered
quantitatively.
The 4-ketone 2 was allowed to react with a large
excess of hydrazine (to inhibit azine formation) in
boiling methanol to give the hydrazone 3 in 95% yield.
After recrystallization from 2-propanol, the hydrazone
in dichloromethane solution (containing tetramethyl-
guanidine) at ꢀ78 ꢁC was brought into reaction with
1 M equiv of lead tetraacetate with rapid stirring for
40 min, whereupon the resultant intensely orange-col-
ored solution afforded, after processing, 1,6-anhydro-
4-deoxy-4-diazo-2,3-O-isopropylidene-b-D-lyxo-hexo-
pyranose (4) as orange needles in 74% yield. The net
yield by this route is more than double that of the pre-
vious procedure,58 and the synthesis can be readily
scaled up. The observed specific rotation of 4 in ether
at the sodium D line is close to zero, but its ORD spec-
trum shows a positive Cotton effect with a maximum at
488 nm and zero at 442 nm, and a positive circular
dichroism curve centered at 455 nm.58 The compound
could be stored at room temperature in a vacuum desic-
cator for several days, but it subsequently decomposed;
consequently, for reference purposes the experimental
section records X-ray powder diffraction data for this
and other compounds in this work that are of limited
stability.
Lead tetraacetate oxidizes the hydrazone in a two-step
dehydrogenative process,67 and tetramethylguanidine
acts as an efficient trap for the acetic acid formed in the
reaction, which would otherwise decompose the product.
The organic base is removed quantitatively from the mix-
ture by simple addition of dry ice to precipitate the insol-
uble carbonate salt for removal by filtration.
A simple change of solvent from acetonitrile to
iodomethane markedly altered the course of the fore-
going reaction. The same two products were observed by
TLC, but the minor, fast-moving component was now
the predominant product. It was isolated by column
chromatography and recrystallized from ether–hexane
in 48% yield as clear prisms. The IR spectrum showed
a strong band at 1700 cmꢀ1, and the mass spectrum
showed a parent ion at m/z 184. The material rapidly
decolorized bromine in carbon tetrachloride. Based on
these considerations, this product is formulated as the
enol ether 1,6-anhydro-4-deoxy-2,3-O-isopropylidene-
b-D-threo-hex-3-enopyranose (6). Its 1H NMR spectrum
again showed, as with the dimer 5, an essentially zero
2.2. Catalytic decomposition of diazo derivative 4 in inert
solvents
The catalytic decomposition of diazo compounds with
metal salts generates reactive intermediates whose prop-
erties are closely dependent on the salt used and the sol-
vent in which the reaction is performed. The use of
copper salts generally produces a carbenoid having elec-
trophilic character, whose mode of the reaction often
involves addition to alkenes to produce cyclopropanes.
When this possibility is not available, the routes of the
reaction may involve formation of a dimeric alkene, or
intramolecular collapse to generate an alkene; azine for-
mation from the diazo precursor is another possibility.
A rapidly stirred solution of diazo derivative 4 in ace-
tonitrile was treated with cuprous iodide, whereupon
there was immediate reaction with evolution of gas and
darkening of the solution. TLC analysis showed two
H-6endo
O
H-6exo
H-5
H-6endo
C-4
O-5
H-1
O
O
H-5
H-6exo
O
O-6
H-3
H-2
Me2
C
Scheme 3. Dihedral angle between H-5 and H-6endo in the bicy-
clo[3.2.1] skeleton.

34
M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
coupling between H-5 and H-6endo, along with several
long-range couplings in this strained-ring system, as ob-
served in related systems.68–70 Formation of an alterna-
tive 4,5-unsaturated compound was not observed nor
expected, as its structure would be in violation of Bredt’s
rule.71,72
initially formed may add to the alkene to form a cyclo-
propane; if the diazo precursor is nonterminal, and the
resulting product is a spirocyclopropane. In the sugar
field, such a product offers a potential route, via opening
of the three-membered ring, to branched-chain sugars.
Photolysis of a dilute solution of 4 in neat vinyl ace-
tate with the Pyrex-filtered light of a mercury arc led
to an intractable mixture of products containing a large
amount of polymerized vinyl acetate. Cooling the same
solution to ꢀ78 ꢁC and addition of cuprous bromide
under rapid stirring led to disappearance of the yellow
color in 2–3 min. Removal of the cuprous bromide
and evaporation gave crystals that, after recrystalliza-
tion from methanol, afforded short needles in 43% yield;
this product was formulated (Scheme 4) as 2,3-O-isopro-
pylidenespiro[1,6-anhydro-4-deoxy-b-D-lyxo-hexopyr-
anose-4,10-cyclopropan]-20-y1 acetate (8). Its mass
spectrum showed a parent ion at m/z 270 and strong
carbonyl absorption in the infrared. The 1H NMR spec-
trum showed an acetyl-group resonance and a high-field
(d 1.90) two-proton pattern, the part of an essentially
A2X spin system with the cyclopropyl methine proton
at low field (d 6.80) on account of deshielding by the
acetoxy group.
Attempted catalytic hydrogenation of 6 with platinum
in methanol under 60 lb in.ꢀ2 (414 kPa) of hydrogen
gave a crystalline product in 68% yield, but this did
not arise from simple hydrogenation of the double
bond. Instead, mass-spectral analysis showed a parent
ion having m/z 216, indicating net incorporation of a
molecule of methanol into 6. Treatment of a methanolic
solution of 6 with a trace of hydrochloric acid gave the
same compound, but in lower yield. The 1H NMR spec-
trum showed the H-1 and H-2 signals as simple dou-
blets, indicating absence of a proton at C-3, along
with a high-field two-proton doublet indicative of a
methylene group at C-4, and allowing formulation of
the product as 1,6-anhydro-4-deoxy-2,3-O-isopropyl-
idene-3-methoxy-b-D-threo-hexopyranos-3-ulose (7). The
complex multiplet at d 4.1–4.0 arises from H-5 being di-
rectly coupled to four adjacent protons. Evidently the
reaction follows the course of simple addition of an
alcohol (methanol) to an enol ether (6) under the elec-
trophilic catalysis of platinum or a proton source. The
configuration, R or S, at C-3 was not established.
Attempts to hydrolyze 7 led to decomposition and the
formation of a tarry product from which no discrete
products could be identified.
With two new asymmetric centers in the molecule,
four possible stereoisomers could, in principle, be
formed, but only a single product was isolated. Conceiv-
ably steric effects might favor introduction of the acet-
oxy group on the endo side of the pyranose ring to
give the (4R,20R) isomer, but specific evidence for this
stereochemical assignment was not obtained.
2.3. Catalytic decomposition of diazo derivative 4 in the
presence of an alkene
2.4. Reaction of diazo derivative 4 with halogens
When a diazo compound is catalytically decomposed by
copper salts in the presence of an alkene, the carbenoid
Diazo ketones are known to react photochemically with
iodine and bromine to produce gem-dihalo com-
O
O
O
O
O
O
O
O
O
I2
O
O
CH2=CHOAc
I
O
N
N
C
I
C
C
AcO
Me2
Me2
Me2
8
Br2
4
9
O
O
O
O
O
O
+
Br
Br
O
O
C
Br
C
Me2
Me2
11
10
Scheme 4.

M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
35
pounds,73 and an analogous reaction of 4 was investi-
gated. It was found that 4 reacted very rapidly with
iodine in dichloromethane, even in the dark, indicating
that the halogen molecule itself initiated the reaction.
The crystalline product isolated was identified (Scheme
4) as the 4,4-diiodo derivative 9 from its mass spectrum,
which showed a molecular ion at m/z 438, correspond-
ing to loss of N2 and incorporation of I2, along with
its 1H NMR spectrum. The compound is not stable,
and it decomposed in 24 h at room temperature.
intermediate carbene species, cannot be ruled out. A
free-radical pathway is implausible, given that light is
not needed for the reaction to proceed, and addition
of hydroquinone (a radical trap) to the reaction medium
did not diminish the yield of dibromide 10.
A minor side-product (<5%) accompanying 10 was
identified as 1,6-anhydro-4-bromo-4-deoxy-2,3-O-iso-
propylidene-b-D-talopyranose (11); its mass spectrum
indicated it to be a monobromo derivative formed by
net addition of HBr to and loss of N2 from the diazo
precursor, and its NMR spectrum was consistent with
the talo configuration with equatorial bromine at C-4.
As with the dibromo derivative 10, H-6endo in 11 reso-
nates at low field, attributable to deshielding by bro-
mine, but the H-3 and H-5 signals lie upfield by
ꢁ0.8 ppm, attributable to the absence of the additional
axial bromine at C-4. Formation of 11 might be ratio-
nalized on the basis of a minor pathway via triplet car-
bene at C-4, abstraction of an atom of bromine, and
addition of hydrogen to the less-hindered side of the
molecule.
A similar reaction was observed between 4 and bro-
mine, and a crystalline 4,4-dibromo product 10 was ob-
tained (Scheme 4), as evidenced by a mass-spectral
1
parent ion at m/z 344 and a H NMR spectrum similar
to that of 9, but better resolved. The H-5,6, and 60 sig-
nals generally appear as an ABX type of multiplet in re-
lated compounds possessing the same bicyclic structure,
but taking the 4-keto compound 2 as a reference for
chemical shifts, it is noteworthy that in the dibromide
10, H-6endo is shifted downfield by about 0.8 ppm,
whereas H-6exo is little affected. This may be attributed
to deshielding because of the close proximity of the exo
bromine atom in 10 to H-6endo.
2.5. 1,3-Dipolar addition of diazo derivative 4
This reaction most probably involves a mechanistic
pathway analogous to the addition of halogen to alk-
enes, with attack by Br2 at C-4 via Br+ and formation
of a transient diazonium ion that in a concerted manner
loses N2 and accepts Brꢀ. Had a C-4 carbocation been
formed, there would have been a rearrangement of the
molecule (see Section 2.6). Given the low ion-solvating
power of the medium, an alternative mechanism via an
The diazo group is highly polarized and can act as a
1,3-dipole toward unsaturated centers. For such a
process, an electron-deficient alkene or alkyne should
be more reactive, as positive polarity on the unsaturated
center should be more attractive toward the nucleophilic
diazo carbon atom. As already observed, an electron-
rich alkene such as vinyl acetate is inert toward
CO2Me
C
O
O
O
O
O
C
N2H4
O
CO2Me
O
O
O
O
O
N
N
O
N
N
N
N
C
C
C
Me2
Me2
Me2
MeO2C
H2NNHCO
CO2Me
CONHNH2
13
12
4
CH2=CHCN
O
O
H
N
O
O
O
O
N
O
N
O
N
C
C
Me2
Me2
NC
NC
14
Scheme 5.

36
M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
compound 4 as a dipolarophile. The reactivity of 4
toward an electron-poor alkyne (dimethyl acetylene-
dicarboxylate) and an alkene (acrylonitrile) was evalu-
ated (Scheme 5).
The 1H NMR spectrum of 14 showed a high-field AB
pattern for the methylene group on the pyrazole ring
plus a broadened singlet for the N–H proton, but other-
wise there was very close resemblance of the spectrum to
that of the 4,4-dibromo derivative 10, emphasizing the
underlying similarity of structure.
An excess of dimethyl 2-butyndioate (dimethyl acetyl-
enedicarboxylate) in dichloromethane was added to a
solution of diazo compound 4 at room temperature.
Over a period of 10 min, the yellow color of the mixture
faded with no evolution of gas, and TLC indicated that
a single product had been formed. Following removal of
volatiles there resulted a syrup that crystallized, and was
recrystallized from 2-propanol to give long needles of a
product identified as 2,3-O-isopropylidenespiro[1,6-
anhydro-4-deoxy-b-D-lyxo-hexopyranose-4,30-pyrazole]-
40,50-di(methyl carboxylate) (12) in 76% yield. The IR
spectrum of the product showed carbonyl absorptions,
2.6. Reactions with protic substrates
Our earlier observation50 that an a-diazo ester in 2-pro-
panol underwent net reduction upon photolysis
prompted a similar investigation on the nonstabilized
diazo sugar derivative 4. However, when a solution of
4 was prepared, and prior to irradiation, it was observed
that the intense yellow color of the solution faded slowly
during ꢁ8 h at room temperature. When the color had
completely disappeared, the solvent was evaporated off
to give a thick syrup that crystallized spontaneously dur-
ing several days. Recrystallization from 2-propanol gave
fine white crystals, mp 103–104 ꢁC, whose mass spec-
trum showed a parent ion of m/z 244, consistent with
loss from 4 of a molecule of nitrogen and incorporation
of a molecule of 2-propanol. This product was inert to-
ward bromine in carbon tetrachloride, and its IR spec-
trum showed no evidence of a double bond. Its 1H
NMR spectrum showed resonances of not one but two
protons in the ‘anomeric region,’ along with a high-field
one-proton multiplet near d 2.2. These results suggested
that the molecule of 4 had not only lost nitrogen and
incorporated a molecule of 2-propanol, but that it had
also undergone internal reorganization (Scheme 6).
When the diazo derivative 4 was dissolved in metha-
nol, a similar reaction took place, although considerably
more rapidly. A crystalline product was formed, mp 92–
93.5 ꢁC, whose mass spectrum showed a parent ion at
1
and its H NMR spectrum was very well resolved and
fully supportive of the assigned structure, with doublets
for H-1 and H-3, a doublet of doublets for H-2, and the
H-5 resonance showing coupling to the C-6 methylene
protons. Notably the H-3 and H-5 signals were more de-
shielded than the anomeric proton, probably because of
the effect of the p electron clouds on the heterocycle at
C-4. Approach of the alkyne from the ‘topside’ of the
diazo group would be subject to severe steric hindrance,
and the (R) configuration assigned at C-4 is based on ap-
proach of the reactant from the unhindered lower face.
Treatment of a methanolic solution of 12 with an ex-
cess of anhydrous hydrazine afforded the corresponding
dihydrazide 13. Attempts to effect thermally a ring-clo-
sure reaction between the hydrazide groups were not
successful; decomposition with evolution of gas was
observed.
Addition of diazo derivative 4 to neat acrylonitrile
failed to yield a spirocyclopropyl derivative analogous
to compound 8; the solution boiled spontaneously and
4 was decomposed. However, when a solution of 4 in
dichloromethane was added to a solution of acrylonitrile
in the same solvent, a controlled reaction ensued with no
evolution of gas. Removal of volatiles gave a crystalline
mass containing (TLC) two products, the major one
of which was separated by fractional crystallization
from 2-propanol and was assigned the structure
(4S)-20,40-dihydro-2,3-O-isopropylidenespiro[1,6-anhy-
dro-4-deoxy-b-D-lyxo-hexopyranose-4,30-pyrazole]-50-
carbonitrile (14). Evidence for this structure is based on
observed IR absorptions for NH and a conjugated CN
group, along with a mass spectrum showing a parent
ion at m/z 265, indicative of the net addition of acrylo-
nitrile to 4, the odd number being consistent with an odd
number of nitrogen atoms. It would appear that initial
1,3-dipolar addition of acrylonitrile to the lower face
of the diazo derivative 4 affords a 3(H)-dihydropyrazole,
and the five-membered ring subsequently tautomerizes
to the 1(H) form, driven by the conjugation stabilization
in the C@N isomer.
O
O
ROH
O
H+
O
O
O
O
O
+
N2
C
C
Me2
Me2
4
Me2CHOH
MeOH
O
O
OMe
OiPr
O
O
O
O
O
O
C
C
Me2
Me2
16
15
Scheme 6.

M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
37
m/z 216, indicating loss from 4 of a molecule of nitrogen
and incorporation of a molecule of methanol. Repeti-
tion of the procedure, but with addition of a catalytic
amount of mineral acid, made the reaction almost
instantaneous, with visible evolution of gas and forma-
tion of the same product. The 1H NMR spectrum of this
product was better resolved than that of the isopropyl
analogue, showing two separate one-proton doublets
in the anomeric region, and all proton resonances could
be readily assigned with the aid of spin-decoupling
experiments.
The combination of two protons resonating in the
anomeric region along with a one-proton multiplet reso-
nating at high (d ꢁ2.2) field is consistent with formation
from 4 of an initial C-4 carbocation that subsequently is
stabilized by a 1,2-alkyl shift to give the branched-chain
anhydro sugar derivative (5S)-5-methyl 1,41-anhydro-4-
deoxy-4-C-hydroxymethyl-2,3-O-isopropylidene-b-
The stereochemical outcome at C-5 may be rational-
ized as possibly the result of a concerted process, with
synchronous loss of N2 from the protonated diazo par-
ent, shift of C-6 to C-4, and equatorial attack of the alk-
oxy group at C-5.
2.7. Nomenclature
Compounds 15 and 16 present challenges in naming by
standard carbohydrate nomenclature,74 but the
branched-chain sugar names given here have the advan-
tage of retaining recognizable stereochemistry in relation
to their precursors. Alternatively they can be named by
standard organic nomenclature75 as (1R,2S,6S,7R,9R)-
4,4-dimethyl-9-methoxy-3,5,8,11-tetraoxa[5.2.3.02,6]-
undecane (16) and its 9-(2-propoxy) analogue 15. The
names assigned to the spiro derivatives 8, 12, and 14
use the recommendations74 in 2-Carb-35.3 in conjunc-
tion with general principles of organic nomenclature.75
L-ribo-pentodialdo-1,5-pyranoside-1,6-pyranose
(16).
The driving force for this rearrangement may be attri-
buted both to the relief of ring strain and formation of
a more-stable carbocation at C-5, which subsequently
captures a molecule of the solvent alcohol. The product
of the preceding experiment may thus be assigned as
the isopropyl congener (15) of compound 16.
The bicyclo[2.2.2] ring system in 15 and 16 confers a
degree of symmetry to the molecule. The large J2,3 value
(8.0 Hz) results from the near-eclipsing of H-2 and H-3.
The H-5 signal falls in the ‘anomeric region’ because, as
with H-1, it is attached to an acetal carbon atom. The
H-4 resonance is at high field because it is attached to
a carbon atom unsubstituted by heteroatoms, and its
multiplicity results from it being coupled to four other
protons.
2.8. Summary conclusions
The reactions described here demonstrate, using the
nonstabilized 4-diazo sugar derivative 4, preparative
procedures for making spiro cyclopropanes from elec-
tron-rich alkenes, spiro pyrazoles from electron-poor
alkenes, gem-dihalo and C@C dimeric structures, and
products of skeletal rearrangement upon protonation.
These reactions have potential in the application to a
variety of structural targets in the sugar field. The im-
proved preparative procedure described here for the 4-
diazo derivative 4 should similarly be applicable with
the 2-diazo and 3-diazo hexose derivatives reported ear-
lier,58 extending the general scope of these reactions in
the sugar field.
Assignment of the configuration at C-5, the new
asymmetric center in 15, may also be made from inspec-
1
tion of the H NMR spectrum. The pyranose ring in 15
is held rigidly in a boat conformation by the anhydro
bridge. The H-5 resonance is not a simple doublet,
but rather a doublet of doublets with spacings of 1.1
and 3.3 Hz. Irradiation of H-4 collapses the H-5 reso-
3. Experimental
3.1. General methods
nance to a narrow doublet, while irradiation of
Evaporations were performed under diminished pres-
sure with a Buchi ‘Rotavapor’ rotary evaporator at
¨
0
H-41,41 collapses the H-5 signal to a wider doublet,
indicating that H-5 is coupled not only to H-4 but also
to one of the methylene protons at C-41. Inspection of
molecular models indicates that, if H-5 is axially ori-
ented, it is correctly aligned in a ‘W’ configuration with
one of the C-41 protons to exhibit 1,3-coupling and gen-
erates the observed narrow spacing. (The same is true
for the relation between H-3 and the other C-41 proton,
and leads to an additional small splitting of the H-3 sig-
nal in addition to its coupling with H-2 and H-4.) Were
H-5 to be oriented equatorially there would be no 1,3-
‘W’ relation with one of the protons at C-41, and so
the axial orientation of H-5 may be assigned to the
two products 15 and 16.
40 ꢁC or less unless otherwise stated. Melting points
were determined on a Thomas–Hoover ‘Unimelt’ melt-
ing point apparatus and are uncorrected. TLC was per-
formed with 0.25-mm layer precoated aluminum plates
(E. Merck, Darmstadt, Germany), or with glass plates
coated with a 0.25-mm layer of silica gel (Silica gel G,
E. Merck, Darmstadt, Germany), activated by heating
to 105 ꢁC before use, using (NH4)2SO4 as an indicator.
Unless otherwise stated, the developing solvent for both
thin-layer and column chromatography was 3:1
CH2Cl2–diethyl ether (v/v). Column chromatography
was performed with Silica gel 7734 (E. Merck) as adsor-
bent. The ratio of silica gel to compound was 100:1 (w/w).

38
M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
IR spectra were recorded with a Perkin–Elmer Model
137 ‘Infracord’ infrared spectrophotometer; samples
were examined as pellets in KBr. Specific rotations were
measured in a 1-dm tube with a Perkin–Elmer Model
141 photoelectric polarimeter. 1H NMR spectra were re-
corded at 100 MHz with a Varian HA-100 spectrometer
equipped with accessories for spin decoupling; coupling
constants are all first order, and chemical shifts are
relative to Me4Si (d = 0.00). Unless otherwise stated,
spectra were measured at 30 ꢁC in CDCl3 solution. Mass
spectra were obtained with an AEI-MS-9 double-focus-
ing, high-resolution mass spectrometer at an ioniza-
tion potential of 70 eV and an accelerating potential
of 8 kV. X-ray powder diffraction data gave interplanar
resulted in complete oxidation of the dioxide to RuO4 in
10 min. The yellow, flocculent mixture was poured into
a 1-L separatory funnel and extracted (3 · 150 mL) with
CCl4. The combined extracts were added directly to a
rapidly stirred solution of 1 (10 g, 50 mmol) in 250 mL
of CH2Cl2. The solution immediately turned black and
opaque. After 10 min, 2-propanol (5 mL) was added to
decompose any excess RuO4. Filtration resulted in a
clear solution with nearly quantitative recovery of
RuO2. Evaporation of the solvent gave the glycos-4-
ulose 2 as a fine, white powder: yield 8.4 g (85%); mp
81–82.5 ꢁC (lit.66 mp 82.5–83 ꢁC). The product was used
in the next step without further purification.
˚
spacings in Angstroms for CuKa radiation (camera
3.4. 1,6-Anhydro-2,3-O-isopropylidene-b-D-lyxo-hexo-
pyranos-4-ulose hydrazone (3)
diameter = 114.59 mm). Relative intensities were esti-
mated visually: m, moderate; s; strong; w, weak; v,
very. The strongest lines are numbered (1, strongest),
and double numbers indicate approximately equal
intensities.
A solution of ketone 2 (20 g, 0.1 mol) in MeOH
(200 mL) was boiled under reflux for 1 h with 95%
hydrazine (50 mL, 1 mol), excess hydrazine being used
to avoid azine formation. TLC indicated complete dis-
appearance of the starting material at this time. Solvent
and unreacted hydrazine were evaporated off to give a
thick syrup that crystallized upon addition of diethyl
ether. Trituration with ether and drying in vacuo gave
the crude hydrazone in essentially quantitative yield.
Recrystallization from 2-propanol gave the pure hydra-
zone 3 (16.7 g, 78%) in two crops. Purification of the
hydrazone was accompanied by some decomposition;
3.2. Preparation of 1,6-anhydro-2,3-O-isopropylidene-b-
D-mannopyranose (1)
Crude D-mannan (ground ivory nut, Pfanstiehl, 500 g)
was pyrolyzed in a modification of the procedure of
Hudson and co-workers65 in 100-g portions under aspi-
rator vacuum with a large bushy flame from a Meker
burner in a Pyrex flask, with the degree of heating main-
tained below the softening point of the glass. The
brown, opaque pyrolyzate was filtered through a 1200
column of activated charcoal in a 300 chromatography
column, and the clear, colorless filtrate was evaporated
under diminished pressure at 50 ꢁC to a thin, light-yel-
low syrup. Acetone (1 L) was gradually added to the syr-
up with stirring, and the mixture was kept for 4 h. The
acetone solution was decanted from a solid, opaque resi-
due into a 2-L flask. Anhyd CuSO4 (250 g) and 2,2-
dimethoxypropane (100 mL) were added, and the
mixture was gently boiled under reflux for 6 h. At the
end of that time, the CuSO4 was removed by filtration,
and evaporation of solvent resulted in a solid mass of
1. Recrystallization from 2-propanol gave the pure
product as thick needles with a yield of 25 g (5% by
weight of D-mannan); mp 159–160 ꢁC (lit.65 mp 161–
162 ꢁC).
21
mp 124–125.5 ꢁC; ½aꢂD ꢀ49 (c 1.0, CHCl3); Rf 0.42; IR
1
mmax 3460 (NH), 1620 (C@N), 1380 cmꢀ1 (CMe2); H
NMR d 5.47 (d, 1H, J1,2 2.5 Hz, H-1), 4.92 (d, 1H,
J5,6exo 6.2 Hz, H-5), 4.20 (dd, 1H, H-2), 3.72–3.92 (m,
2H, H-6,60), 1.46 and 1.31 (s, 3H, CMe2); X-ray powder
diffraction data: 7.82 vw, 7.09 m, 5.62 vs (1), 5.23 vs (2),
4.54 s, 3.97 s, 3.81 s (3), 3.45 w, 3.34 w, 3.24 w, 3.07 m,
2.96 w, 2.87 vw, 2.74 w, 2.63 w. Anal. Calcd for
C9H14N2O4: C, 50.47; H, 6.5; N, 13.08. Found: C,
50.50; H, 6.77; N, 13.05.
3.5. Preparation of 1,6-anhydro-4-deoxy-4-diazo-2,3-O-
isopropylidene-b-D-lyxo-hexopyranose58 (4)
A solution of hydrazone 3 (10 g, 47 mmol) in CH2Cl2
(250 mL) was cooled in a dry ice–2-propanol bath to
ꢀ78 ꢁC. Tetramethylguanidine (25 mL) was added60 to
keep the solution basic. With rapid stirring, 23 g
(1.05 equiv) of Pb(OAc)4 was added neat. The solution
immediately turned bright yellow. Every few minutes, a
drop of the solution was placed on moistened filter paper
to test for unreacted Pb(OAc)4, a positive reaction being
indicated by formation of a brown spot of PbO2. After
40 min, the reaction was negative, and the clear yellow
solution was poured into a separatory funnel containing
300 mL of cold, 10% NaOH. After thorough shaking,
the lower, yellow organic layer was drained from the
3.3. Preparation of 1,6-anhydro-2,3-O-isopropylidene-b-
D-lyxo-hexopyranos-4-ulose (2)
The glycos-4-ulose 2 was prepared by a minor modifica-
tion of the method of Horton and Jewell.66 Ruthenium
dioxide hydrate (5 g, soluble form, 50% Ru, Englehard
Industries) was suspended in 100 mL of water, and the
mixture was cooled in an ice bath. Gradual addition
of 5% aq NaOCl (Clorox) to the rapidly stirred mixture

M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
39
cloudy, white, aqueous layer through a cone of MgSO4
into a 1-L erlenmeyer flask, and small chunks of dry
ice were added during 30 min to precipitate excess tetra-
methylguanidine as its insoluble carbonate salt. Filtra-
tion of the solution and evaporation at room
temperature gave 4 as fine, light-yellow crystals: yield
7.3 g (74%); mp 108 ꢁC (dec). Physical constants were
in agreement with those recorded earlier58 for this com-
pound prepared by the Bamford–Stevens method. The
material could be crystallized from ether or an ether–hex-
ane mixture to afford large, yellow-orange needles, but at
the cost of considerable loss to decomposition.
tion from the copper salt and evaporation gave a thick
syrup that crystallized after addition of a small amount
of 2-propanol and scratching. The 3-alkene 6 was sepa-
rated from a small amount of the dimer 5 by column
chromatography to give clear prisms: yield 0.27 g
22
(42%); mp 50–52 ꢁC; ½aꢂD ꢀ60 (c, 1.0, CHCl3); IR
(KBr) mmax 1700 (enol ether), 1370 cmꢀ1 (CMe2); m/z
184 (M+), 169 (M+ꢀCH3); 1H NMR (acetone-d6) d
5.67 (d, 1H, J1,2 2.2 Hz, H-1), 4.83 (dd, 1H, J4,5 3.2,
J5,6 3.1 Hz, H-5), 4.74 (dd, 1H, J2,4 1.8 Hz, H-4), 4.62
0
(m, 1H, H-2), 3.86 (d, 1H, J6;6 7:6 Hz, H-6endo), 3.74
(dd, H-6exo), 1.42 (s, 6H, CMe2); X-ray powder diffrac-
tion data: 8.38 m, 7.05 vw, 6.37 vs (1), 5.15 s, 4.75 vs (2),
4.57 s (3), 4.41 m, 4.18 m, 3.97 m, 3.70 vw, 3.55 w, 3.42
m, 3.23 m, 2.35 m. Anal. Calcd for C9H12O4: C, 58.74;
H, 6.54, Found: C, 58.83; H, 6.35.
3.6. 1,6;10,60-Dianhydro-4,40-dideoxy-2,3;20,30-di-O-iso-
propylidene-bi(b-D-lyxo-hexopyranos-4-ylidene) (5)
A solution of diazo derivative 7 (0.8 g, 3.8 mmol) in
MeCN (150 mL) was treated with CuI (100 mg) with
rapid stirring at room temperature. The mixture darkened
immediately with rapid evolution of gas and then turned
light green. The MeCN was evaporated, CH2Cl2
(100 mL) was added, and the flask was shaken to dis-
solve the products. The CuI was removed by filtration,
and the CH2Cl2 was evaporated off to give a thick syrup,
shown by TLC to contain two components having Rf
0.55 and 0.90. The major, slower-moving product 5
was isolated by column chromatography as short, fine
The material immediately decolorized Br2 in CCl4.
3.8. 1,6-Anhydro-4-deoxy-2,3-O-isopropylidene-3-meth-
oxy-b-D-threo-hexopyranos-3-ulose (7)
The enol ether 6 (200 mg, 1.1 mmo1) was dissolved in
MeOH (200 mL), PtO2 (100 mg) was added, and the
mixture was shaken at 25 ꢁC under 60 lb in.ꢀ2 pressure
of H2 for 16 h. At the end of that time, TLC indicated
that the starting material had disappeared. Filtration
to remove Pt and evaporation of the MeOH gave a syr-
up that gradually crystallized. Recrystallization from 2-
propanol gave the product as clear crystals: yield 147 mg
20
needles: yield 0.44 g (64%); mp 215–217 ꢁC; ½aꢂD ꢀ256
(c 1.0, CHCl3); IR(KBr) mmax 1380 cmꢀ1; m/z 368
(M+), 353 (MꢀCH3); 1H NMR d 5.40 (d, 1H, J1,2
2.2 Hz, H-1), 5.19 (m, 1H, J5,6exo 5.3, J5,6endo 1.6, J3,5
0.8 Hz, H-5), 4.97 (dd, 1H, H-3), 4.13 (dd, 1H, J2,3
23
(68%); mp 120–122 ꢁC; ½aꢂD ꢀ44 (c 0.5, CHCl3); IR
(KBr) mmax 1370 cmꢀ1 (CMe2); m/z 216 (M+), 201
(M+ꢀCH3); 1H NMR (C6D6): d 5.31 (d, 1H, J1,2
3.0 Hz, H-1), 4.02 (d, 1H, m, 1H, H-2 and H-5), 3.67
0
6.2 Hz, H-2), 4.03 (q, J6;6 7:9 Hz, H-6), 3.84 (q, 1H,
H-60), 1.54 and 1.33 (s, 3H, CMe2); X-ray powder dif-
fraction data: 10.90 vw, 8.79 w, 7.70 m (3), 7.17 vw,
5.23 s (2), 4.75 s (1), 4.38 w, 4.04 vw, 3.64 vw. Anal.
Calcd for C18H24O8: C, 58.69; H, 6.52. Found: C,
58.88; H, 6.32.
Ozonolysis of the dimer 5 (100 mg, 0.2 mmol) in
30 mL of a 2:1 (v:v) CH2Cl2–pyridine mixture at
ꢀ78 ꢁC gave a single analytically pure product, identical
to the glycos-4-ulose 2; yield 88 mg (81%).
(dd, 1H, J5,6endo 0.7, J6;6 6:7 Hz, H-6endo), 3.38 (dd,
0
1H, J5,6exo 6.2 Hz, H-6exo), 3.00 (s, 3H, OMe), 1.81
(d, 2H, J4,5 3.7 Hz, H-4, 40), 1.60 and 1.48 (s, 3H,
CMe2); X-ray powder diffraction data: 7.28 s (2), 6.06
w, 5.64 m, 5.38 vs (1), 5.03 m, 4.75 m, 4.35 s (3), 3.96
w, 3.78 w, 3.51 vw, 3.32 w, 3.02 w, 2.96 w, 2.62 m. Anal.
Calcd for C10H16O5: C, 55.56; H, 7.40. Found: C, 55.83;
H, 7.17.
The dimer 5 (200 mg) was dissolved in MeOH
(150 mL) containing 100 mg of platinum oxide, and
the mixture was shaken under 60 lb in.ꢀ2 (414 kPa) of
H2 for 24 h. Filtration from the catalyst and evapora-
tion of the solution resulted in quantitative recovery of
5.
3.9. 2,3-O-Isopropylidenespiro[1,6-anhydro-4-deoxy-b-D-
lyxo-hexopyranose-4,10-cyclopropan]-20-y1 acetate (8)
Irradiation of a solution of diazo derivative 4 (400 mg,
1.9 mmol) in neat vinyl acetate with Pyrex-filtered light
of a medium-pressure mercury arc gave an intractable
mixture of several products, including polymerized vinyl
acetate. However, when an identical, rapidly stirred
solution was treated at room temperature with CuBr
(150 mg), a simpler mixture resulted; the reaction pro-
ceeding rapidly with evolution of gas. Removal of CuBr
by filtration and evaporation of the filtrate gave a thick
syrup, column chromatography of which afforded the
spirocyclopropane derivative 8 in modest yield (66 mg,
3.7. 1,6-Anhydro-4-deoxy-2,3-O-isopropylidene-b-D-
threo-hex-3-enopyranose (6)
A solution of diazo derivative 4 (0.6 g, 2.8 mmol) in MeI
(100 mL) was treated at room temperature with cuprous
iodide (200 mg) with rapid stirring. The solution imme-
diately lost its yellow color with evolution of gas. Filtra-

40
M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
13%), together with significant quantities of other
products.
25 ꢁC turned yellow and had the odor of iodine. It
decomposed in CH2Cl2 solution during 8–12 h, and
the solution turned violet.
In a more-satisfactory procedure, an identical vinyl
acetate solution of 4 was first cooled to ꢀ78 ꢁC in a
dry ice–acetone bath. Addition of the same amount of
CuBr with rapid stirring brought about a slow reaction,
which required about 10 min for completion. The mix-
ture was filtered, and the filtrate was evaporated to a
thick syrup. Addition of a small amount of 2-propanol
and by scratching the inside of the flask for a few min-
utes brought about the crystallization of 8. Recrystalli-
zation from 2-propanol gave pure 8 as fine needles:
3.11. 1,6-Anhydro-4,4-dibromo-4-deoxy-2,3-O-isopropyl-
idene-b-D-lyxo-hexopyranose (10)
A solution of bromine (ꢁ0.4 M) in CH2Cl2 was slowly
added with rapid stirring to a solution of 4 (0.5 g,
2.4 mmol) in CH2Cl2 (150 mL) at room temperature.
The red color of the bromine was immediately dis-
charged. Addition of Br2 was stopped when a red color
persisted in the solution, and solvent and excess of bro-
mine were evaporated to give a thin, yellow syrup that
spontaneously crystallized over several hours. Recrystal-
lization from 2-propanol gave the pure product 10 as
20
yield 214 mg (42%); mp 89–91 ꢁC; ½aꢂD ꢀ101 (c 1.0,
CHCl3); IR (KBr) mmax 1740 (OAc), 1370 cmꢀ1
(CMe2); m/z 270 (M+), 255 (M+ꢀCH3); 1H NMR d
6.81 (app t, 1H, JA,X 7.0 Hz, cyclopropyl CH), 5.55 (d,
1H, J1,2 3.0 Hz, H-1), 4.50 (dd, 1H, J2,3 5.9 Hz, H-2),
4.11 (d, 1H, H-3), 4.05–3.80 (m, 3H, H-5,6,60), 2.17 (s,
3H, COCH3), 1.88 (m, 2H, AB portion of ABX system,
cyclopropyl CH2), 1.55 and 1.30 (s, 3H, CMe2); X-ray
powder diffraction data: 10.21 vw, 6.44 w, 5.95 m (3),
4.42 m (2), 3.62 m (1), 3.27 vw, 2.96 vw, 2.86 w. Anal.
Calcd for C13H18O6: C, 57.78; H, 6.67. Found: C,
57.51; H, 6.89.
21
fine needles: yield 0.52 g (63%); mp 102.5–104 ꢁC; ½aꢂD
ꢀ53 (c 1.0, CHCl3); IR (KBr) mmax 1370 cmꢀ1 (CMe2);
m/z 344 (M+), 329 (327, 329, 331 in 1:2:2:1 ratio,
M+ꢀCH3); 1H NMR d 5.44 (dd, 1H, J1,2 2.5, J1,3
1.0 Hz, H-1), 4.33 (dd, 1H, J2,3 5.0 Hz, H-2), 4.80
(ddd, 1H, J3,5 1.5 Hz, H-3), 4.75 (ddd, 1H, J5,6exo 6.0,
0
J5,6endo 1.0 Hz, H-5), 4.49 (dd, 1H, J6;6 8:0 Hz, H-
6exo), 3.75 (dd, 1H, H-6endo), 1.60 and 1.40 (s, 3H,
CMe2); X-ray powder diffraction data: 9.08 vs, 6.01 vs
(3), 5.67 w, 2.46 w, 4.77 vs, 4.17 vs (2), 3.95 w, 3.55 vs
(1), 2.73 s. Anal. Calcd for C9H12Br2O4: C, 31.40; H,
3.49; Br, 46.51. Found: C, 31.45; H, 3.32; Br, 46.60.
The compound was unstable at room temperature;
pure material left in a tightly capped bottle for several h
acquired a strong odor of AcOH, and it decomposed
over a period of a few weeks at 0 ꢁC.
3.10. 1,6-Anhydro-4-deoxy-4,4-diiodo-2,3-O-isopropylid-
ene-b-D-lyxo-hexopyranose (9)
3.12. 1,6-Anhydro-4-bromo-4-deoxy-2,3-O-isopropylid-
ene-b-D-talopyranose (11)
A solution of iodine (ꢁ0.2 M) in CH2Cl2 was slowly
added with rapid stirring to a solution of diazo derivative
4 (0.5 g, 2.4 mmol) in CH2Cl2 (150 mL) at room temper-
ature. Initially the violet color of iodine was immediately
discharged. Addition was terminated when the violet col-
or persisted. The excess of I2 was removed by treatment
with activated charcoal. Filtration of the mixture and
evaporation of the filtrate gave a yellow syrup that crys-
tallized upon the addition of a small amount of 2-propa-
nol and by scratching the inside of the flask.
Recrystallization from 2-propanol gave fine, colorless
TLC examination of the mother liquors of the reaction
that led to 10 revealed a slower-moving spot (Rf 0.30),
and this component was separated by column chroma-
tography to afford flat prisms (55 mg, less than 5%) of
20
a product identified as 11: mp 91–92 ꢁC; ½aꢂD ꢀ20 (c
1.0, CHCl3); m/z 264, 266 (M+), 249, 251 (M+ꢀCH3),
1
126 (M+ꢀBrꢀAcO); H NMR (C6D6) d 5.23 (dd, 1H,
0
J1,2 2.5, J1,3 0.7 Hz, H-1), 4.40 (dd, 1H, J6;6 7:8, J5,6endo
1.1 Hz, H-6endo), 4.04 (m, 1H, J2,3 5.3, J3,4 5.1 Hz, H-3),
3.94 (m, 1H, J4,5 5.3, J5,6exo 5.4 Hz, H-5), 3.74 (m, 1H,
J4,6 1.2 Hz, H-4), 3.60 (dd, 1H, H-2), 3.42 (m, 1H, H-
6exo), 1.57 and 1.15 (s, 3H, CMe2); X-ray powder dif-
fraction data: 10.6 m, 8.79 s, 7.37 vs (3), 5.75 vs (1),
5.09 w, 4.78 vs (2), 4.60 m, 4.22 m, 3.94, 3.59 s, 2.75
m, 2.65 vw, 2.55 w, 2.01 vw. Anal. Calcd for C9H13BrO4:
C, 40.75; H, 4.91; Br, 30.15. Found: C, 40.78; H, 4.94;
Br, 30.41.
22
crystals of 9: yield 0.56 g (53%); mp 203 ꢁC (dec); ½aꢂD
ꢀ46 (c 0.2, CHCl3); IR (KBr) 1380 cmꢀ1 (CMe2); m/z
1
438 (M+ꢀCH3), 311 (M+ꢀI); H NMR d 5.54 (dd, 1H,
J1,2 2.4, J1,3 1.0 Hz, H-1), 4.83 (m, 1H, J2,3 4.2, J3,5
1.8 Hz, H-3), 4.77 (m, 1H, J5,6endo 1.0, J5,6exo 5.4 Hz,
0
H-5), 4.50 (dd, J6;6 8:0 Hz, H-6endo), 4.42 (dd, 1H, H-
2), 3.54 (dd, 1H, H-6exo), 1.54 and 1.33 (s, 3H, CMe2);
X-ray powder diffraction data: 6.80 vs (2), 5.80 m, 5.55
m, 5.34 vs (1), 4.98 w, 4.73 m, 4.50 s (3), 4.33 m, 4.07
w, 3.70 m. Anal. Calcd for C9H12I2O4: C, 24.61; H,
2.74; I, 57.91. Found: C, 24.84; H, 2.84; I, 58.03.
3.13. 2,3-O-Isopropylidenespiro[1,6-anhydro-4-deoxy-b-
D-lyxo-hexopyranose-4,30-pyrazole]-40,50-di(methyl car-
boxylate) (12)
Compound 9 was unstable at room temperature for
extended periods: a closed sample kept for 24 h at
Dimethyl acetylenedicarboxylate (1.0 g, 7 mmol) in
CH2Cl2 (10 mL) was added to a stirred solution of 4

M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
41
0
(0.8 g, 3.8 mmol) in the same solvent (100 mL) at room
temperature. Over a 10-min period, the yellow color of
the diazo compound faded with no visible evolution of
gas, and TLC indicated that a single product had
formed. The solvent was evaporated off, and the excess
of dimethoxycarbonylacetylene was removed by re-
peated evaporation with 2-propanol, giving a thick,
slightly colored syrup that crystallized spontaneously
upon being kept for several days; subsequent prepara-
tions crystallized immediately upon removal of the sol-
vent. Recrystallization from 2-propanol gave the
product as long, silky needles: yield 1.02 g (76%); mp
H-3), 4.22 (app d, 1H, J6;6 8:0 Hz, H-6endo), 4.08 (dd,
1H, H-2), 3.65 (dd, 1H, H-6exo), 3.14 (dd, 2H, J
17.2 Hz, CH2 of pyrazole ring), 1.49 and 1.26 (s, 3H,
CMe2); X-ray powder diffraction data: 9.20 vs (2), 7.25
vs (1), 6.08 in, 5.75 s, 4.90 vs (3), 4.45 m, 4.35 w, 4.16
s, 3.82 m, 3.65 m, 3.47 w, 3.12 m, 3.01 vw, 2.76 m. Anal.
Calcd for C12H15N3O4: C, 56.25; H, 5.86; N, 16.40.
Found: C, 56.09; H, 5.83; N, 16.22.
The mother liquors contained the minor component,
which was not characterized, along with additional 14.
3.15. (5S)-5-Isopropyl 1,41-anhydro-4-deoxy-4-C-
hydroxymethyl-2,3-O-isopropylidene-b-L-ribo-pento-
dialdo-1,6-pyranose-1,5-pyranoside [(1R,2S,6S,7R,9R)-
4,4-dimethyl-9-(2-propoxy)-3,5,8,11-tetraoxa[5.2.3.02,6]-
undecane] (15)
21
158.5–159 ꢁC; ½aꢂD ꢀ70 (c 1.0, CHCl3); IR (KBr) mmax
1750 and 1730 (CO2Me), 1370 cmꢀ1 (CMe2); m/z 354
1
(M+), 339 (M+ꢀCH3), 323 (M+ꢀCH3O); H NMR d
6.30 (d, 1H, J2,3 5.8 Hz, H-3), 6.03 (dd, 1H, H-5), 5.66
(d, 1H, J1,2 3.0 Hz, H-1), 4.39 (dd, 1H, H-2), 4.26–4.08
(m, 2H, H-6,60), 3.92 and 3.85 (s, 3H, OMe), 1.58 and
1.48 (s, 3H, CMe2); X-ray powder diffraction data:
12.10 m, 7.92 vs (1), 7.49 s, 6.12 m, 5.75 w, 4.79 vs (2),
4.41 s, 4.30 m, 4.12 m, 3.99 s (3), 3.72 m, 3.60 m, 3.45
s, 3.28 m, 3.08 m, 2.93 s, 2.84 w, 2.37 w, 2.23 w, 2.05
vw. Anal. Calcd for C15H18N2O8: C, 50.85; H, 5.08;
N, 7.91. Found: C, 50.58; H, 5.17; N, 7.61.
To a solution of 0.7 g (3.3 mmol) of compound 4 in an-
hyd 2-propanol (50 mL) was added one drop of acidic 2-
propanol (prepared by adding 1 drop of concentrated
HCl to 20 mL of 2-propanol). An immediate reaction
occurred with evolution of gas and instantaneous dis-
charge of the yellow color of the diazo compound.
The mixture was immediately made neutral by the addi-
tion of anhyd Na2CO3 (2.0 g), filtered, and the filtrate
was evaporated to a thick syrup that slowly crystallized.
Recrystallization from 2-propanol gave 15 as fine, white
crystals, yield 0.35 g (41%); mp 103–104 ꢁC; [a]D ꢀ46 (c
0.5, CHCl3); IR (KBr) mmax 1370 cmꢀ1 (CMe2): m/z 244
3.13.1. Hydrazide derivative of 12. A solution of 12
(0.6 g, 1.7 mmol) in MeOH (50 mL) was stirred for 1 h
at room temperature with 95% hydrazine (0.3 mL,
10 mmol). Solvent and excess hydrazine were evapo-
rated off to give a thick syrup that crystallized spontane-
ously. Recrystallization from 2-propanol gave the
product, 2,3-O-isopropylidenespiro[1,6-anhydro-4-deoxy-
b-D-lyxo-hexopyranose-4,30-pyrazole]-40,50-di(carboxhydr-
(M+), 229 (M+ꢀCH3); H NMR d 5.03 (m, 2H, J1,2
1
2.5 Hz, H-1,5), 4.43 (dd, 1H, J2,3 8.2, J3,4 3.0 Hz, H-3),
0
4.18 (dd, 1H, H-2), 4.03 (m, 2H, H-41,41 ), 4.03 (septet,
1H, CH of i-Pr), 2.16 (m, 1H, H-4), 1.62 and 1.43 (s, 3H,
CMe2), 1.32 and 1.21 (two s, 3H each, J 6.0 Hz, non-
equivalent Me of i-Pr); X-ray powder diffraction data:
17.31 m (2), 8.62 vw, 7.89 w, 6.20 vw, 5.74 w, 5.45 vw,
4.81 m (3), 4.20 s (1). Anal. Calcd for C12H20O5: C,
59.02; H, 8.20. Found: C, 59.03; H, 8.12.
22
azide) (13): yield 0.52 g (87%); mp 222 ꢁC (dec); ½aꢂD
323 (M+ꢀNHNH2).
ꢀ51 (c 1.0, CHCl3); m/z 354 (M+), 339 (M+ꢀCH3),
3.14. 20,40-Dihydro-2,3-O-isopropylidenespiro[1,6-anhy-
dro-4-deoxy-b-D-lyxo-hexopyranose-4,30-pyrazole]-50-
carbonitirile (14)
3.16. (5S)-5-Methyl 1,41-anhydro-4-deoxy-4-C-hydroxy-
methyl-2,3-O-isopropylidene-b-L-ribo-pentodialdo-1,6-
pyranose-1,5-pyranoside [(1R,2S,6S,7R,9R)-4,4-
dimethyl-9-methoxy-3,5,8,11-tetraoxa[5.2.3.02,6]unde-
cane] (16)
To a stirred solution of 4 (0.6 g, 2.8 mmol) in CH2Cl2
(50 mL) was added a solution of acrylonitrile (1.0 mL,
2.0 mmol, in 10 mL of CH2Cl2) at room temperature.
The yellow color of the diazo compound faded over a
period of 8 min, and evaporation of solvent and unre-
acted acrylonitrile gave a white solid, shown by TLC
to contain two components, Rf 0.70 and 0.58. The faster
migrating, major component was separated from the
minor one by fractional crystallization from 2-propanol
and was isolated as fine, short needles: yield 0.42 g
To a solution of 0.6 g (2.8 mmol) of 4 in MeOH (50 mL)
was added one drop of acidic MeOH (prepared by add-
ing one drop of concentrated HCl to 20 mL of MeOH),
The reaction was immediate, with evolution of gas and
loss of the yellow color. The solution was made neutral
by addition of anhyd Na2CO3 (2.0 g). Filtration and
evaporation of solvent gave a thick syrup that crystal-
lized slowly. Recrystallization from 2-propanol gave 16
as fine, white crystals: yield 0.30 g (50%); mp 92–
23
(58%); mp 231 ꢁC (dec); ½aꢂD ꢀ406 (c 1.0, CHCl3); IR
(KBr) mmax 3330 (NH), 2210 (CN, conj.), 1370 cmꢀ1
1
(CMe2); m/z 265 (M+), 251 (M+ꢀCH3); H NMR (ace-
20
tone-d6): d 5.22 (d, 1H, J1,2 2.8 Hz, H-1), 4.43 (m, 1H,
J5,6exo 5.2 Hz, H-5), 4.36 (dd, 1H, J2,3 5.8, J3,5 1.4 Hz,
93.5 ꢁC; ½aꢂD +18 (c 1.0, CHCl3); IR (KBr) mmax
1380 cmꢀ1 (CMe2); m/z 216 (M+), 201 (M+ꢀCH3); H
1

42
M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
23. Frey, H. M. Proc. R. Soc. London, Ser. A 1959, 250, 409–
421.
24. Frey, H. M. Proc. R. Soc. London, Ser. A 1959, 251, 575–
587.
25. Bethell, D.; Whittaker, D.; Callister, J. D. J. Chem. Soc.
1965, 2466–2476.
26. Overberger, C. G.; Anselme, J. P. J. Org. Chem. 1964, 29,
1188–1190.
NMR d 5.04 (d, 1H, J1,2 2.5 Hz, H-1), 4.81 (dd, 1H, J4,5
3.2, J14;5 1:1 Hz, H-5), 4.40 (dd, 1H, J2,3 8.0, J3,4 3.6 Hz,
0
H-3), 4.18 (dd, 1H, H-2), 4.10 (m, 2H, H-41,41 ), 3.53 (s,
3H, OMe), 2.21 (m, 1H, H-4), 1.60 and 1.41 (s, 3H,
CMe2); X-ray powder diffraction data: 15.72 w, 8.41 m
(3), 7.79 m, 5.95 w, 5.36 s (2), 4.20 m, 3.77 s (1), 3.54
vw, 2.78 w. Anal. Calcd for C10H16O5: C, 55.56; H,
7.41. Found: C, 55.75; H, 7.28.
27. Bethell, D.; Brown, K. C. J. Chem. Soc., Chem. Commun.
1967, 24, 1266–1267.
28. Doering, W. von E.; Buttery, R. G.; Laughlin, R. G.;
Chaudhuri, N. J. Am. Chem. Soc. 1956, 78, 3224.
29. Richardson, D. B.; Simmons, M. C.; Dvoretzky, I. J. Am.
Chem. Soc. 1960, 82, 5001–5002.
Acknowledgements
30. Richardson, D. B.; Simmons, M. C.; Dvoretzky, I. J. Am.
Chem. Soc. 1961, 83, 1934–1937.
31. Simons, J. W.; Mazac, C. J.; Taylor, G. W. J. Phys. Chem.
1968, 72, 749–751.
32. Richardson, D. B.; Durrett, L. R.; Martin, J. M., Jr.;
Putnam, W. E.; Slaymaker, S. C.; Dvoretzky, I. J. Am.
Chem. Soc. 1965, 87, 2763–2765.
This work was supported in part by NIH grant GM-
11976. The authors acknowledge helpful discussions
with Dr. A. D. McNaught concerning nomenclature.
33. Kopecky, K. R.; Hammond, G. S.; Leermakers, P. A. J.
Am. Chem. Soc. 1961, 83, 2397–2398.
References
34. Kopecky, K. R.; Hammond, G. S.; Leermakers, P. A. J.
Am. Chem. Soc. 1962, 84, 1015–1019.
35. Skell, P. S.; Woodworth, R. C. J. Am. Chem. Soc. 1956,
78, 4496–4997, and 6427.
36. Woodworth, R. C.; Skell, P. S. J. Am. Chem. Soc. 1959,
81, 3383–3386.
37. Kirmse, W. Ref. 13, p 6.
1. Reichstein, T.; Grussner, A. Helv. Chim. Acta 1934, 17,
311–328.
2. Walton, E.; Rodin, J. O.; Stammer, C. H.; Holly, F. W.;
Folkers, K. J. Am. Chem. Soc. 1958, 80, 5168–5173.
3. Wolfrom, M. L.; Wood, H. B., Jr. J. Am. Chem. Soc. 1955,
77, 3096–3098.
38. Loose, A. J. Prakt. Chem. 1909, 79, 505–510.
39. Ebel, F.; Brunner, R.; Mangelli, P. Helv. Chim. Acta 1929,
12, 19–26.
40. Seyferth, D. Chem. Rev. 1955, 55, 1155–1177.
41. Skell, P. S.; Etter, R. M. Chem. Ind. (London) 1958, 624–
625.
42. Mueller, E.; Kessler, H.; Zeeh, B. Fortschr. Chem. Forsch.
1966, 7, 128–171.
43. Kirmse, W. Ref. 13, p 87.
44. Hine, J. Divalent Carbon; Ronald Press: New York, 1964.
45. Harrison, J. F. in Ref. 13, Chapter 5, p 159.
46. Kobrich, G. Angew. Chem., Int. Ed. Engl. 1967, 6, 41–
52.
47. Doyle, M. P.; McKervey, M. A.; Ye, T. Modern Catalytic
Methods for Organic Synthesis with Diazo Compounds:
from Cyclopropanes to Ylides; Wiley–Interscience: New
York, 1998.
4. Wolfrom, M. L.; Berkebile, J. M.; Thompson, A. J. Am.
Chem. Soc. 1949, 71, 2360–2362.
5. Wolfrom, M. L.; Thompson, A. Methods Carbohydr.
Chem. 1962, 1, 118–120.
6. Webber, J. M. Adv. Carbohydr. Chem. 1962, 17, 15–63.
7. Wolfrom, M. L.; Kashimura, N.; Horton, D. Carbohydr.
Res. 1974, 36, 211–213.
8. Regitz, M. In The Chemistry of Diazonium and Diazo
Groups; Patai, S., Ed.; Wiley: New York, 1978; Chapter
17, pp 751–820; Sorriso, S. In The Chemistry of Diazonium
and Diazo Groups; Patai, S., Ed.; Wiley: New York, 1978;
Chapter 3, pp 95–135.
9. Wolfrom, M. L.; Olin, S. M.; Evans, E. F. J. Am. Chem.
Soc. 1944, 66, 204–206.
10. Wolfrom, M. L.; Brown, R. L. J. Am. Chem. Soc. 1943,
65, 1516–1521.
11. Zhdanov, Y. A.; Kornilov, V. I.; Bogdanova, G. V.
Carbohydr. Res. 1966, 3, 139–140.
12. Levene, P. A.; LaForge, F. B. J. Biol. Chem. 1915, 21,
345–350.
48. Friedman, L. In Carbonium Ions; Olah, G. A., Schleyer,
P., von, R., Eds.; Wiley: New York, 1970; Vol. II, Chapter
16.
49. Roberts, J. D.; Watanabe, W. J. Am. Chem. Soc. 1950, 72,
4869–4879.
13. Kirmse, W. Carbene Chemistry; Academic Press: New
York, 1971; p 5.
50. Horton, D.; Philips, K. D. Carbohydr. Res. 1972, 22, 151–
162.
14. Frey, H. M. J. Chem. Soc., Chem. Commun. 1965, 260–
261.
51. Gelas-Mialhe, Y.; Horton, D. J. Org. Chem. 1978, 43,
2307–2310.
15. Elliot, C. S.; Frey, H. M. Trans. Faraday Soc. 1968, 64,
2352–2368.
52. Gallos, J. K.; Massen, Z. S.; Koftis, T. V.; Dellios, C. C.
Tetrahedron Lett. 2001, 42, 7489–7491.
53. Gallos, J. K.; Koftis, T. V.; Massen, Z. S.; Dellios, C. C.;
Mourtzinos, I. T.; Coutouli-Argyropoulou, E.; Koumbis,
A. E. Tetrahedron 2002, 58, 8043–8053.
16. Skell, P. S.; Klebe, J. J. Am. Chem. Soc. 1960, 82, 247–248.
17. Kirmse, W. Ref. 13, p 209.
18. Kirmse, W.; von Scholz, H. D.; Arnold, H. Justus Liebigs
Ann. Chem. 1968, 711, 22–30.
19. Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1959, 81,
5512–5513.
´
54. Sarabia, F.; Chammaa, S.; Lopez Herrera, F. J. Tetra-
hedron 2001, 57, 10271–10279.
55. Desai, V. N.; Saha, N. N.; Dhavale, D. D. J. Chem. Soc.,
Perkin Trans. 1 2000, 147–151.
20. Frey, H. M. Proc. Chem. Soc. London 1954, 318.
21. Whitten, G. Z.; Rabinovitch, B. S. J. Phys. Chem. 1965,
69, 4348–4357.
56. Berndt, D. F.; Norris, P. N. Tetrahedron Lett. 2002, 43,
3961–3962.
22. Ring, D. F.; Rabinovitch, B. S. Can. J. Chem. 1968, 46,
2435–2450.

M. S. Alexander, D. Horton / Carbohydrate Research 342 (2007) 31–43
43
57. Brandenhorst, H. M.; Kemmink, J.; Liskamp, R. M. J.;
Pieters, R. J. Tetrahedron Lett. 2002, 43, 9601–9603.
58. Horton, D.; Just, E. K. J. Chem. Soc., Chem. Commun.
1969, 1116–1117.
59. Acton, E. M.; Fujiwara, A. N.; Goodman, L.; Henry, D.
W. Carbohydr. Res. 1974, 33, 135–151.
60. Sauer, D. R.; Schneller, S. W.; Gabrielsen, B. Carbohydr.
Res. 1993, 241, 71–79.
61. Sacui, I.; Vicarel, M.; Risi, R.; Zeller, M.; Norris, P. Abstr.
Papers Am. Chem. Soc. Meet. 2006, 231, CARB-029.
62. Alexander, M. S.; Horton, D. Abstr. Papers Am. Chem.
Soc. Meet. 1977, 174, CARB-12.
68. Horton, D.; Jewell, J. S. Carbohydr. Res. 1967, 5, 148–160.
69. Horton, D.; Jewell, J. S.; Just, E. K.; Wander, J. D.
Carbohydr. Res. 1971, 18, 49–56.
70. Horton, D.; Just, E. K. Carbohydr. Res. 1971, 18, 81–
94.
71. Bredt, J.; Thouet, H.; Schnitz, J. Liebigs Ann. 1924, 437,
1–13.
72. Horton, D.; Jewell, J. S. Carbohydr. Res. 1966, 3, 255–257.
73. Taylor, T. W. J.; Forscey, L. A. J. Chem. Soc. 1930, 2272–
2277.
74. IUPAC. Nomenclature of Carbohydrates (Recommenda-
tions, 1996). Pure Appl. Chem. 1996, 68, 1919; Adv.
Carbohydr. Chem. Biochem. 1997, 52, 43–177; <http://
www.chem.qmul.ac.uk/iupac/2carb/> (accessed 1 Novem-
ber 2006).
63. Horton, D. Jpn. J. Antibiot 1979, 32, S145–S162.
64. Alexander, M. S.; M.S. Thesis, Ohio State University,
1978.
65. Knauf, A. E.; Hann, R. M.; Hudson, C. S. J. Am. Chem.
Soc. 1941, 63, 1447–1451.
66. Horton, D.; Jewell, J. S. Carbohydr. Res. 1966, 2, 251–260.
67. Gale, D. M.; Middleton, W. J.; Krespan, C. G. J. Am.
Chem. Soc. 1965, 87, 657.
75. IUPAC Recommendations on Organic and Biochemical
Nomenclature, Symbols and Terminology, <http://
www.chem.qmul.ac.uk/iupac/spiro/> (accessed 1 Novem-
ber 2006) and <http://www.chem.qmul.ac.uk/iupac/fused-
ring/> (accessed 1 November 2006).