Pd(II)-catalysed aminocarbonylation as a key step in the total synthesis of C-6 homologues of 1-deoxynojirimycin and 1-deoxy-L-idonojirimycin

Article abstract of DOI:10.1016/S0957-4166(00)00226-3

The first successful Pd(II)-catalysed aminocarbonylation of the highly substituted benzylaminoalkene 5 allows the direct preparation of fused piperidine lactones 3 and 4, which are subsequently converted to the novel C-6 homologue of 1-deoxynojirimycin 1

Full text of DOI:10.1016/S0957-4166(00)00226-3

Tetrahedron: Asymmetry 11 (2000) 2579±2597
Pd(II)-catalysed aminocarbonylation as a key step in the total
synthesis of C-6 homologues of 1-deoxynojirimycin and
1-deoxy-l-idonojirimyciny
Peter Szolcsanyi,a Tibor Gracza,a,* Marian Koman,b Nad'a Pronayova
and Tibor Liptajc
aDepartment of Organic Chemistry, Faculty of Chemical Technology, Slovak University of Technology,
Â
Radlinskeho 9, 812 37 Bratislava, Slovakia
bDepartment of Inorganic Chemistry, Faculty of Chemical Technology, Slovak University of Technology,
Â
Radlinskeho 9, 812 37 Bratislava, Slovakia
cCentral NMR Laboratories, Faculty of Chemical Technology, Slovak University of Technology,
Â
Radlinskeho 9, 812 37 Bratislava, Slovakia
Received 4 May 2000; accepted 5 June 2000
Abstract
The ®rst successful Pd(II)-catalysed aminocarbonylation of the highly substituted benzylaminoalkene 5
allows the direct preparation of fused piperidine lactones 3 and 4, which are subsequently converted to the
novel C-6 homologue of 1-deoxynojirimycin 1 and 1-deoxy-l-idonojirimycin 2. The study of the in¯uence
of various catalytic conditions on the diastereoselectivity and product distribution of the key amino-
carbonylation is presented. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
Azasugar glycosidase inhibitors1 are analogues of monosaccharides in which the ring oxygen is
replaced by nitrogen. This substitution renders the compounds metabolically inert, but does not
prevent their recognition by glycosidases and other carbohydrate-recognising proteins.2 As a
consequence, an in vivo competitive inhibition of glycosidases results in signi®cant alteration of
many crucial biochemical pathways and other important biological functions of living organisms.
* Corresponding author. E-mail: gracza@chelin.chtf.stuba.sk
y
Part of a PhD Thesis by P.S., Slovak University of Technology, Bratislava, 1998.
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00226-3

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2580
Therefore, glycosidase inhibitors have a great potential in the diagnostics and treatment of obesity
and diabetes mellitus,3 cancer,4 viral infections5 including AIDS,6 and hereditary lysosomal storage
diseases.7
In that context there is continued interest in the synthesis and biological evaluation of chiral
non-racemic polyhydroxylated piperidines and their various derivatives as potent glycosidase
inhibitors. Numerous successful synthetic approaches to this class of compounds have appeared
over the last decade. Homoazasugars with CH2-homologation in the side chain may be interesting
compounds for testing structure±activity relationships. Indeed, there is a strong synthetic e€ort
put toward the ecient preparation of di€erent homologues of polyhydroxylated piperidines.8
However, only a few of the already published synthetic strategies deal with the preparation of C-6
homologues of azasugars, iminoheptitols9 and iminooctitols.10 Here we report on the ®rst stereo-
selective synthesis of such new homologues of 1-deoxynojirimycin 1 and 1-deoxy-l-idonojiri-
mycin9f 2 as their hydrochloride salts11 (Fig. 1).
Figure 1.
Our retrosynthetic plan relied on the key Pd(II)-catalysed aminocarbonylation of highly oxygen-
ated and N-benzyl protected aminoalkene 5, which could, in principle, provide both desired dia-
stereoisomeric lactones with l-gulo 3, and l-ido 4 con®gurations, respectively. We envisaged that
in the case of their successful preparation, our targets 1 and 2 would be in hand by ®nal routine
transformations of lactones 3 and 4 (Scheme 1).
Scheme 1.

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2581
2. Results and discussion
2.1. Methodology
Generally, Pd(II)-catalysed amino-/amidocarbonylation of N-protected 1-amino-pent-4-ene-3-ols
proceeds well,12 and selectively provides cis-fused pyrrolidine-g-lactones,13 which have since been
utilised in ecient total syntheses of valuable natural products14 and related compounds.15 Com-
pared to 1-amino-pent-4-ene-3-ols, the homologous 6-amino-hex-2-ene-3-ols are much less reactive
towards intramolecular Pd(II)-catalysed amino-/amidocarbonylation. It should be noted at this
point that reactions leading to the formation of ®ve-membered rings are not necessarily applicable
to the synthesis of six-membered rings. Moreover, the additional issue of the diastereoselectivity
of the formation of corresponding piperidine lactones must be taken into account.
To the best of our knowledge, there are only three papers in the literature, which deal with such
a transformation on similar but less functionalised substrates.12c,13c,16 However, reported yields of
aminocarbonylations have not as yet reached the range of preparative usefulness and no further
synthetic elaboration of the prepared piperidine lactones has been reported. The important
common feature of almost all published amino-/amidocarbonylations (either producing pyrrolidine
or piperidine lactones) until now is the use of electron withdrawing protecting groups (tosyl,
carbamate, amide) of the NH-function of aminoalkenes (Scheme 2).
Scheme 2.
This was claimed to be essential for optimal adjustment of the N-nucleophilicity in order to
favour the cyclisation step over the competing N±Pd complexation.15 However, the properties
of such protecting groups (introduction, deprotection, chemical inertness and stability) were not
suitable for our proposed plan of the total synthesis. Therefore, we decided to explore the
applicability of the benzyl-protecting group in the Pd(II)-catalysed aminocarbonylation to produce
the piperidine lactones. We have formulated two basic aims:
(i) to attempt the Pd(II)-catalysed aminocarbonylation of highly functionalised and N-
benzyl-protected aminoalkene 5 producing the corresponding fused piperidine-lactones
3 and 4; to explore the in¯uence of di€erent catalytic conditions of the key reaction (i.e.
catalyst, reoxidant, additive, solvent, temperature) on the product composition;

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2582
(ii) to ®nd and select the best reaction conditions for the highest possible stereoselective
preparation of both desired lactones 3 and 4; to utilise the prepared compounds in the
total synthesis and to prepare two C-6 homologues of 1-DNJ 1 and l-ido-1-DNJ 2 as
perspective glycosidase inhibitors.
2.2. Synthesis of the key aminoalkenetriol 5
Our preparation of the key substrate for the Pd(II)-catalysed aminocarbonylation, the benzyl-
protected aminoalkenetriol 5, commences from cheap and commercially available methyl-a-d-
glucoside 6 and is analogous to known literature procedures. Thus, benzylidenation of the start-
ing material 6 with benzaldehyde diethylacetal17 a€orded the protected diol 7,18 which was sub-
sequently benzylated under standard conditions19 to give fully protected 8 in 61% yield in two
steps. An acid-mediated deprotection of acetal 8 using 2% sulfuric acid in MeOH gave the diol 9
in almost quantitative yield and a subsequent nucleophilic displacement20 of the primary OH
function under Mukaiyama conditions furnished pure bromoalcohol 10 in 79% yield. The last
transformation of 10 involved a one-pot three-step sequence21 (reductive elimination, ring open-
ing and reductive amination) giving the desired (2S,3S,4R)-1-benzylamino-2,3-di-O-benzyl-hex-5-
ene-2,3,4-triol 5 in 64% yield (Scheme 3). With the key aminoalkenetriol 5 in our hands (30%
overall yield in ®ve steps) the stage was set for the key Pd(II)-catalysed aminocarbonylation.
Scheme 3. Reactions and conditions: (i) PhCH(OEt)2, cat. CSA, CHCl3, re¯ux, 3 h, 81%; (ii) BnBr, NaH, DMF, 0ꢀC,
3 h, 75%; (iii) 2% H2SO4/MeOH, rt, 3 h, 97%; (iv) Ph3P, CBr4, pyridine, 0ꢀC, 3 h!60ꢀC, 30 min, 79%; (v) Zn dust,
BnNH2, NaBH3CN, n-PrOH:H2O (19:1), re¯ux, 2 h, 64%
2.3. Palladium(II)-catalysed aminocarbonylation of alkene 5
First attempts to cyclise aminoalkenetriol 5 under standard carbonylation conditions12c (1 atm.
CO, cat. PdCl2, 3 equiv. CuCl2, 3 equiv. AcONa, AcOH) at room temperature failed and the
starting material was always reisolated. However, elevation of the temperature up to 50ꢀC led to
the complete consumption of aminoalkenetriol 5 while the colour of the reaction mixture changed
from dark green to ochre.{ However, aminocarbonylation of 5 furnished a crude brown oil as a
{
This signi®cant colour change allows a convenient monitoring of the reaction course and actually re¯ects a reduction
of CuCl2, used in excess as reoxidant, to its Cu+ form.

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2583
complex mixture of compounds from which desired lactones with l-gulo 3 and l-ido 4 con®g-
uration were isolated in the ratio of 1:4.8 (65% d.e. by quantitative 13C NMR) and in 46% yield.
Much to our surprise, the major by-product of the reaction turned out to be N-benzyl-2,3-di-O-
benzyl-6-chloro-1,6-dideoxy-l-idonojirimycin 11 in 9% yield (Scheme 4).
Scheme 4. Reactions and conditions: (i) CO (balloon), 0.1 equiv. PdCl2, 3 equiv. CuCl2, 3 equiv. AcONa, glacial
AcOH, 50ꢀC, 6.5 h, FLC, chloride 11 (9%)+mixture of l-gulo lactone 3/l-ido lactone 4 (1/4.8, 46%)
The course of the aminocarbonylation of alkene 5, as well as the product distribution, was
surprising and requires further comment. First of all, this reaction is the ®rst known example of a
Pd(II)-catalysed aminocarbonylation of a highly functionalised and N-benzyl-protected amino-
alkene producing the corresponding fused piperidine lactones. In contrast to the known literature
precedents, our result with the alkene 5 shows that aliphatic amines are also capable of under-
going the intramolecular Pd(II)-catalysed aminocarbonylation giving bicyclic [4.3.0] piperidine
lactones. Secondly, this is the ®rst known precedent of a Pd(II)-catalysed aminocarbonylation of
a protected aminoalkenetriol producing both possible diastereoisomers of piperidine lactones. It
is obvious that in contrast to pyrrolidines, piperidines can by fused to ®ve-membered lactones
either in the cis- or trans-fashion. However, known literature examples report either the forma-
tion of cis-fused lactones12c,16 exclusively (but without any explanation of such an impressive
stereodiscrimination) or this issue was not addressed at all.13c Thirdly, there are no reports detailing
the formation of cyclic chloro derivative by-products such as 11 during aminocarbonylations in
previous papers; however, Tamaru and Yoshida12c,16 observed the formation of acyclic 6-chloro-
hex-4-enylamines as side products. The literature search revealed papers describing the use of
Pd(II)/CuCl2 systems for chlorohydroxylation22 of alkenes. Our suspicion that the formation of
undesired 11 might be a result of abuse of CuCl2 (used in excess as reoxidant) for such PdCl2-
mediated side reactions was proved by applying non-chlorinating23 aminocarbonylation conditions
to alkene 5. We were pleased to ®nd that a catalytic system consisting of PdCl2, p-benzoquinone
(BQ), LiCl, AcONa and THF a€orded the mixture of l-gulo 3 and l-ido 4 lactones in the ratio of
3.7:1 (58% d.e. by quantitative 13C NMR) and in 66% yield with no formation of 11. The de®nitive
con®rmation of our hypothesis came from a simple experiment in which aminoalkenetriol 5 was
subjected to the standard aminocarbonylation conditions but with exclusion of carbon monoxide.
The only product we were able to isolate was the chloroderivative 11 (Scheme 5).
Scheme 5. Reactions and conditions: (i) 0.1 equiv. PdCl2, 3 equiv. CuCl2, 3 equiv. AcONa, glacial AcOH, rt, 48 h,
FLC, chloride 11 (70%)

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2584
Thus, we coincidentally found an interesting type of PdCl2/CuCl2-mediated chloroamino-
cyclisation of alkene 5. To the best of our knowledge, there is only one literature precedent24
describing an analogous transformation. However, the authors employed a much simpler substrate
[PdCl2(MeCN)2, CuCl2, PrCN] and neither the con®guration of the product(s) nor the diastereo-
isomeric excess were given (Scheme 6).
Scheme 6.
The result of the aminocarbonylation of 5 that employed BQ as an alternative reoxidant
brought our attention to another interesting issue.x It was apparent that a change in the nature of
the reoxidant (CuCl2$BQ), solvent (AcOH$THF) and additive (AcONa$LiCl) has a tremendous
in¯uence on the diastereoselectivity and product distribution in the aminocarbonylation of alkene
5. This was a very important and yet unprecedented observation and therefore it was decided to
explore this phenomenon in a more detailed and systematic fashion. Our screening strategy was
based on the variations of all reaction components (catalyst, reoxidant, additive, solvent) and
temperature. The following reagents and conditions were chosen: catalyst: three Pd(II)-salts with
di€erent ligand spheres and `secondary' structuresÐPdCl2 (oligomer), PdCl2(MeCN)2 (monomer)
and Pd(OAc)2 (dimer); reoxidant: two `complementary' Cu(II)-salts and one organic oxidantÐ
CuCl2 (chlorinating), Cu(OAc)2 and BQ (both non-chlorinating); additive: AcONa as a weak base
eliminating 2 equiv. of HCl evolved during the aminocarbonylation (also acts as a bu€er in the
case of AcOH as solvent) and LiCl as the promotor of the redox process; solvent: four solvents
commonly used in Pd-catalysed reactions, which di€er in polarity parameters and proton-donating
propertiesÐAcOH, MeOH, THF and MeCN; temperature: two di€erent temperatures: rt and
50ꢀC. From all of these selected options we set up three suitable combinations of di€ering catalytic
systems for the aminocarbonylation of 5 as follows:
A: 0.1 equiv. PdCl2, 3 equiv. CuCl2, 3 equiv. AcONa;
B: 0.1 equiv. PdCl2, 1 equiv. BQ, 2 equiv. LiCl, 2 equiv. AcONa;
C: 0.1 equiv. Pd(OAc)2, 3 equiv. Cu(OAc)2, 3 equiv. AcONa.
These were applied under an atmospheric pressure of carbon monoxide in all four di€erent
solvents and at indicated temperatures. Our results are summarised in Tables 1±3 (Schemes 7±9).
Our search for optimal reaction conditions for the aminocarbonylation of 5 gave a number of
encouraging results. First of all, we gathered further proof of our hypothesis that CuCl2 might
be responsible for the formation of chloro derivative by-product 11. We were able to isolate
compound 11 in all cases (except that of MeOH) when employing conditions A (Table 1) but in
no case without the presence of CuCl2 (conditions B and C, Tables 2 and 3). In addition, we did
not observe the formation of 11 in the stoichiometric aminocarbonylation of 5 (1 equiv. PdCl2)
with exclusion of CuCl2 (Table 1, entry 7). However, the ratio of lactones 3:4 in this case (1:1.9)
was di€erent to that obtained under the same but catalytic conditions (1:4.8, Table 1, entry 2).
x
Note that the ratio of lactones 3/4 formed by using BQ (3.7:1) is reversed to that obtained by using CuCl2 (1:4.8).

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2585
Table 1
Table 2
Table 3

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2586
Scheme 7.
Scheme 8.
Scheme 9.
This could suggest the existence of two di€erent catalytic species involved: in the case of PdCl2/
CuCl2, the formation of certain types of Pd±Cu heterobimetallic species is reasonable25 in con-
trast to the oligomeric monometallic structure of [PdCl2]n itself in the case of stoichiometric
aminocarbonylation. Secondly, we discovered that other catalytic systems are employable in the
aminocarbonylation of the alkene 5 and produce desired lactones 3 and 4 (Tables 2 and 3).
However, we could not ®nd any reasonable correlation, which would explain the in¯uence of
certain reaction parameters on the diastereoselectivity of aminocarbonylation of 5. Thirdly, we
were delighted to ®nd that we are able to govern the course of the reaction and product
distribution by simple variation of reaction components. Thus, the best d.e. (69%) in favour of
l-gulo lactone 3 was obtained under conditions C in AcOH (entry 2, Table 3); however, due to
the low yield (19%) of the reaction it is much more practical to employ conditions B in THF
(58% d.e., 66% yield, Table 2, entry 3) for the preparation of this product. On the other hand, the
best d.e. (ꢁ90%, the other diastereoisomer was not observable in the 300 MHz NMR spectrum)
in favour of l-ido lactone 4 was obtained under conditions B in MeOH (entries 4 and 5, Table 2);
however, due to the slow reaction, incomplete conversion and therefore low yields (9 and 21%) it

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2587
is much more practical to employ either the same conditions but in AcOH (70% d.e., 32% yield,
entry 2, Table 2) or conditions A in AcOH (65% d.e., 46% yield, entry 2, Table 1) for the pre-
paration of l-ido lactone 4. Thus, at this stage we were able to prepare either desired lactones 3 or
4 in a more or less stereoselective manner by a simple variation of reaction components.
The last issue that should be discussed is the determination of d.e.'s in the previous study as
well as the determination of relative con®gurations of both lactones 3, 4 and chloro derivative 11.
Diastereoisomeric excesses have been determined by integration of corresponding OCH2Ph sig-
nals in quantitative 13C NMR spectra (75.43 MHz, CDCl3, TMS: 73.5 and 73.8 ppm for l-gulo 3;
72.3 and 74.1 ppm for l-ido 4) with suppressed NOE e€ect of the crude reaction mixtures
(just ®ltered through short plug of silica gel to remove the inorganic salts). The relative con®g-
1
uration of l-gulo lactone 3 was established on the basis of H NMR as follows: H-4 proton
appeared to be a triplet (instead of dd) with two coupling constants: J3,4=9.0 and J4,5=9.2 Hz
indicating the trans-position of H-4 and H-5 protons of 3. Additionally, the NOESY spectrum of
l-gulo lactone 3 shows the following space relationships, which also con®rm the proposed con-
®guration (Fig. 2).
Figure 2.
The relative con®guration of l-ido lactone 4 was established on the basis of DIFNOE and
NOESY 1H NMR experiments, which revealed the cis-relationship of H-4 and H-5 protons (Fig. 3).
Also, a lower value of their coupling constant (J4,5=7.3 Hz) in comparison to 3 supports this
observation. However, such a value of the coupling constant of two cis-related protons is rather
high and would suggest a signi®cant distortion of the chair conformation of the piperidine ring as
a result of the lactone fusion in a cis-fashion.
Figure 3.
Finally, the relative con®guration of chloro derivative 11 was established as in the previous two
1
cases using H NMR. The coupling constant J4,5=3.0 Hz suggested the cis-relationship and this
1
statement was supported by DIFNOE H NMR experiment, which con®rmed a cis-relationship
of H-4 and H-5 protons and thus established the con®guration of 11 as l-ido (Fig. 4).

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2588
Figure 4.
2.4. Completion of the total synthesis of 1 and 2
With pure lactones l-gulo 3 and l-ido 4 in our hands the total synthesis was set up for the last
two steps. Reductive opening15 of both lactone rings with LiBH4 led to the diols 12 and 13 in
good yields. Final catalytic debenzylation followed by ion-exchange chromatography puri®cation
a€orded the desired targets 1,5,6-trideoxy-1,5-imino-d-gluco-heptitol 1 and 1,5,6-trideoxy-1,5-
imino-l-ido-heptitol 2 as hydrochlorides (Scheme 10).
Scheme 10. Reactions and conditions: (i) LiBH4, dry THF, 0ꢀC!rt, 3 h, FLC, 12 (64%), 13 (65%); (ii) H2 (balloon),
10% Pd/C, HCl, MeOH, rt, overnight, Dowex1 (H+ form), 1 (90%), 2 (82%)
The values of 1H NMR coupling constants of d-gluco 12 (J4,5=9.0 Hz) and l-ido 13 (J4,5=4.7
Hz) diols were in perfect agreement with their 4,5-trans and 4,5-cis proton relationships, respec-
tively. The trans-relationship of H-4 and H-5 protons of the d-gluco iminoheptitol 1 was
con®rmed by their coupling constant (J4,5=8.2±8.5 Hz), which indicates the desired relative
con®guration. Finally, the absolute con®guration of 1 was determined by single crystal X-ray
analysis26 (Fig. 5).
3. Conclusion
We have performed the ®rst successful Pd(II)-catalysed aminocarbonylation of highly functio-
nalised and N-benzyl-protected aminoalkene 5 a€ording the corresponding piperidine lactones 3

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2589
Figure 5.
and 4. This is an unprecedented case of the employment of an aliphatic amine in such an amino-
carbonylation reaction. In contrast to the known literature precedents we observed the formation
of both possible diastereoisomers depending on the catalytic conditions used. We also observed
the formation of chloro derivative by-product 11 when employing CuCl2 in excess as a reoxidant.
The reason for this was the abuse of CuCl2 for Pd-mediated side reactions as con®rmed by the
experiment with exclusion of CO; the only isolated product was the diastereomerically pure 11.
Thus, we coincidentally found the highly stereoselective chloroaminocyclisation of substituted
aminoalkene 5 leading to a piperidine derivative that could be very useful as a building block in
the synthesis of naturally occurring piperidines and azepines.27
Finally, we conducted a screening study in order to ®nd the optimal conditions for the amino-
carbonylation of 5 in terms of diastereoselectivity and suppression of the formation of side
products. By simple variation of di€erent components (catalyst, reoxidant, additive, solvent,
temperature) we were able to govern the course of the aminocarbonylation as well as the product
distribution in a desirable way and thus prepare both lactones 3 and 4 in a more or less stereo-
selective manner. The ®nal synthetic manipulations led to the ®rst total synthesis of iminoheptitol
1, a new homologue of naturally occurring 1-deoxynojirimycin. The potential glycosidase inhibitory
properties of the iminoheptitols 1 and 2 will be tested in the very near future.
4. Experimental
4.1. General methods
All reagents were used as received without further puri®cation unless otherwise speci®ed. All
solvents were distilled before use: THF and Et2O from Na/benzophenone; toluene from Na;
MeCN from P2O5; DMF from KOH; MeOH from MeONa; CH2Cl2 and CHCl3 from activated 4
A molecular sieves. Petrol ether refers to the fraction of light petroleum ether boiling in the range
40±60ꢀC and hexanes refer to the fraction 60±65ꢀC. Flash column liquid chromatography (FLC)
was performed on silica gel Kieselgel 60 (40±63 mm, 230±400 mesh) and analytical thin-layer
chromatography (TLC) was performed on aluminum plates pre-coated with either 0.2 mm (DC-
Alufolien, Merck) or 0.25 mm silica gel 60 F254 (ALUGRAM1 SIL G/UV254, Macherey±Nagel).
The compounds were visualised by UV ¯uorescence and by dipping the plates in an aqueous

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2590
H2SO4 solution of cerium sulfate/ammonium molybdate followed by charring with a heat-gun.
Melting points were obtained using a Buchi 510 capillary melting point apparatus and a Ko¯er
hot plate and are uncorrected. Optical rotations were measured with a Perkin±Elmer 241 polari-
meter and a POLAR L-mP polarimeter (IBZ Messtechnik) with a water-jacketed 10 000 cm cell at
a wavelength of sodium line D (l=589 nm). Speci®c rotations are given in units of 10^1 deg cm2
g^1 and concentrations are given in g/100 ml. Elemental analyses were performed by the Micro-
analytical Service of University of Stuttgart and by Microanalytical Service of Slovak University
of Technology. Infrared spectra were recorded either on a Philips Analytical PU9800 FTIR
spectrometer or a Perkin±Elmer 1750 FTIR spectrophotometer as KBr discs (KBr) or as thin
1
®lms on KBr plates (®lm). H NMR spectra were recorded on a Varian Gemini 200 (200 MHz)
and a Varian VXR-300 (299.94 MHz) spectrometer. Chemical shifts (ꢀ) are quoted in ppm and
are either referenced to the tetramethylsilane (TMS) as internal standard (ꢀMe=0.00 ppm for
299.94 MHz) or the residual protic solvent (CHCl3, ꢀH=7.26 ppm for 200 MHz; CH3OH,
ꢀH=3.34 ppm for 299.94 MHz) was used as the internal reference. Coupling constants (J) are
recorded in hertz. The following abbreviations were used to characterise signal multiplicities: s
(singlet), d (doublet), t (triplet), m (multiplet), b (broad). Abbreviations with quotation marks
mean that the appearance of the signal is di€erent to that theoretically predicted. The COSY,
NOESY and DIFNOE techniques were used in assignment of proton±proton relationships and
the determination of relative con®guration. 13C NMR spectra were recorded on a Varian Gemini
200 (50.32 MHz) and a Varian VXR-300 (75.43 MHz) spectrometer. The multiplicities of carbons
were assigned from a broadband decoupled analysis used in conjunction with either an APT or a
DEPT program. Chemical shifts are quoted in ppm and are either referenced to the tetra-
methylsilane (TMS) as internal standard (ꢀ=0.00 ppm for 75.43 MHz) or the central resonance
of CDCl3 (ꢀ=77.0 ppm for 50.32 MHz) and CD3OD (ꢀ=49.0 ppm for 75.43 MHz) was used as
the internal reference. The following abbreviations were used to characterise signal multiplicities:
s (singlet), d (doublet), t (triplet), q (quartet). The mark `$' means that corresponding signals can
be interchanged. The numbering of lactones 3 and 4 for characterisation of NMR signals as
depicted in Figs. 2 and 3 is in accord with the numbering of piperidines 1, 2, 12 and 13 for better
comparison of all spectral data. The HETCOR and HMQC techniques were used throughout for
the assignment of the proton±carbon relationships. Low-resolution mass spectra were recorded
on a V.B. Masslab 20-250 (EI), a V.G. Platform II (APCI), and a MS 902 S (EI) mass spectro-
meter, respectively, with only molecular ions being reported.
4.2. (2S,3S,4R)-1-Benzylamino-2,3-di-O-benzyl-hex-5-ene-2,3,4-triol 5
Acid treated zinc dust [(102.3 g, 1.565 mol, 60 equiv.) prepared by gradual washing with 3%
HCl (600 ml), H2O (3Â500 ml), EtOH (200 ml), acetone (200 ml), dry Et2O (200 ml) and ®nally
n
dried in vacuo] was added to the solution of bromoalcohol 10 (11.4 g, 0.26 mol) in PrOH:H2O
(19:1, 460 ml) mixture with vigorous stirring. Then BnNH2 (42 g, 42.7 ml, 0.392 mol, 15 equiv.)
and NaBH3CN (3.28 g, 0.052 mol, 2 equiv.) were added and the resulting mixture was stirred at
re¯ux for 2 h. After cooling to rt the suspension was ®ltered through Celite1 pad and washed with
EtOH (100 ml). The ®ltrate was evaporated in vacuo and the residue was dissolved in Et2O (500 ml)
and treated with 20% aq. HCl (90 ml). After 1 h the mixture was basi®ed with 15% aq. sol. NaOH
and diluted with H2O (100 ml), the organic phase was separated and the water layer was extracted
with CH2Cl2 (3Â200 ml). The combined organic extracts were dried over MgSO4 and evaporated
in vacuo giving the crude product as an oil, which was puri®ed by FLC (silica gel, 50% AcOEt/

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2591
petrol ether) a€ording pure aminoalkenetriol 5 (Rf 0.25) as a pale yellow oil (6.9 g, 64%);
‰ꢁŠ2D3=+27.8 (c 1 in CHCl3); ꢀH (200 MHz; CDCl3) 2.86 (1H, dd, J1a,2=4.0, J1a,1b=12.3, H-1a),
3.00 (1H, dd, J1b,2=1.5, J1a,1b=12.0, H-1b), 3.60±3.69 (2H, m, H-2, H-3), 3.66 (1H, d, J=13.4,
NCH2Ph), 3.86 (1H, d, J=12.9, NCH2Ph), 4.43±4.45 (1H, m, H-4), 4.47 (1H, d, J=12.2,
OCH2Ph), 4.54 (1H, d, J=12.2, OCH2Ph), 4.58 (1H, d, J=11.9, OCH2Ph), 4.68 (1H, d, J=11.8,
OCH2Ph), 5.20 (1H, d, J5,6E=10.4, H-6E), 5.43 (1H, d, J5,6Z=17.2, H-6Z), 6.05 (1H, ddd,
J4,5=5.0, J5,6E=10.5, J5,6Z=17.1, H-5); ꢀC (50.32 MHz, CDCl3) 46.3 (t, C-1), 53.7 (t, NCH2Ph),
69.3 (d, C-4), 72.3, 74.0 (2d, 2OCH2Ph), 77.4, 81.1 (2d, C-2$C-3), 115.0 (t, C-6), 127.6, 127.9,
128.1, 128.5, 128.6, 128.7, 128.8 (all d, all CH of Ph), 138.4, 138.5, 139.0 (3s, 3Cipso of Ph), 139.4
(d, C-5); ꢂmax (®lm)/cm^1 3312 (w, OH); m/z (APCI+) 418 (M+); found: C, 78.08; H, 7.58; N, 3.68;
C27H31O3N requires: C, 77.67; H, 7.48; N, 3.35%.
4.3. General procedure for the aminocarbonylation of an alkene 5
To a dry, one-necked round-bottomed ¯ask equipped with a side-arm stopcock and containing
a magnetic stirring bar, a mixture of reagents depending on the catalytic conditions used [i.e. (A)
PdCl2 (0.1 equiv.), CuCl2 (3 equiv.) and AcONa (3 equiv.); (B) PdCl2 (0.1 equiv.), BQ (1 equiv.),
LiCl (2 equiv.) and AcONa (2 equiv.); (C) Pd(OAc)2 (0.1 equiv.), Cu(OAc)2 (3 equiv.) and
AcONa (3 equiv.)] and ®tted with a CO balloon was added the aminoalkenetriol 5 (1 equiv.)
dissolved in the appropriate solvent (AcOH, THF, MeOH). [The setup for reactions running in
MeCN was di€erent: PdCl2 was placed in a ¯ask and stirred in dry acetonitrile overnight while
the brown suspension turned into yellow slurry of PdCl2(MeCN)2. Then CuCl2, AcONa and the
substrate in dry MeCN were added.] The apparatus was then purged with CO and the resulting
mixture was stirred at the indicated temperature (rt or 50ꢀC) and for the indicated period of time
while the colour of the reaction changed. The complex reaction mixture was ®ltered on a Celite1
pad, washed several times with the appropriate solvent and the ®ltrate was evaporated in vacuo.
[In the case of AcOH last traces of the residual solvent were removed by co-evaporation with
toluene.] The oily residue was subjected to puri®cation by FLC on silica gel (10% AcOEt/hex-
anes) yielding chloro derivative 11 as a pale yellow oil and/or mixture of two diastereoisomeric
lactones 3 and 4 (l-gulo+l-ido) as a pale yellow oil.
4.3.1. Conditions A
AcOH: PdCl2 (43 mg, 0.2397 mmol), CuCl2 (967 mg, 7.19 mmol), AcONa (590 mg, 7.19
mmol), substrate 5 (1 g, 2.397 mmol) in glacial AcOH (30 ml); 50ꢀC, 6.5 h; colour of the reaction
mixture changed from green to ochre; chloride 11 100 mg (9%)+mixture of l-gulo 3 and l-ido 4
lactones 488 mg (46%) in a ratio 1:4.8 (65% d.e.).
THF: PdCl2 (5 mg, 0.03 mmol), CuCl2 (97 mg, 0.721 mmol), AcONa (59 mg, 0.719 mmol),
substrate 5 (100 mg, 0.24 mmol) in dry THF (4 ml); rt, 7 h; colour of the reaction mixture chan-
ged from dark green to light green; chloride 11 15 mg (14%)+mixture of l-gulo 3 and l-ido 4
lactones 30 mg (28%) in a ratio 1:1.1 (5% d.e.).
MeOH: PdCl2 (4 mg, 0.024 mmol), CuCl2 (97 mg, 0.721 mmol), AcONa (59 mg, 0.719 mmol),
substrate 5 (100 mg, 0.24 mmol) in dry MeOH (4 ml); rt, 10 h; colour of the reaction mixture
remained green; mixture of l-gulo 3 and l-ido 4 lactones 38 mg (36%) in a ratio 1:2.2 (37% d.e.).
MeCN: PdCl2 (4 mg, 0.024 mmol) in dry MeCN (0.5 ml) (brown!yellow); CuCl2 (97 mg,
0.721 mmol), AcONa (59 mg, 0.719 mmol) (yellow!gray green); substrate 5 (100 mg, 0.24
mmol) in dry MeCN (2.5 ml); rt, 8 h; colour of the reaction mixture changed from dark green to

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2592
light green; chloride 11 5 mg (5%)+mixture of l-gulo 3 and l-ido 4 lactones 25 mg (24%) in a
ratio 1:2.0 (34% d.e.).
4.3.2. Conditions B
AcOH: PdCl2 (5 mg, 0.028 mmol), p-benzoquinone (26 mg, 0.24 mmol), LiCl (21 mg, 0.495
mmol), AcONa (40 mg, 0.488 mmol), substrate 5 (100 mg, 0.24 mmol) in glacial AcOH (3 ml);
50ꢀC, 47 h; colour of the reaction mixture changed from brown to black; mixture of l-gulo 3 and
l-ido 4 lactones 34 mg (32%) in a ratio 1:5.6 (70% d.e.).
THF: PdCl2 (5 mg, 0.028 mmol), p-benzoquinone (26 mg, 0.24 mmol), LiCl (21 mg, 0.495
mmol), AcONa (40 mg, 0.488 mmol), substrate 5 (100 mg, 0.24 mmol) in dry THF (5 ml); rt, 17
h; colour of the reaction mixture changed from red±brown to brown±black; mixture of l-gulo 3
and l-ido 4 lactones 70 mg (66%) in a ratio 3.7:1 (58% d.e.).
MeOH: PdCl2 (5 mg, 0.028 mmol), p-benzoquinone (26 mg, 0.24 mmol), LiCl (21 mg, 0.495
mmol), AcONa (40 mg, 0.488 mmol), substrate 5 (100 mg, 0.24 mmol) in dry MeOH (4 ml); rt, 3
days; colour of the reaction mixture changed from red±brown to black; incomplete conversion;
mixture of l-gulo 3 and l-ido 4 lactones 10 mg (9%) in a ratio <1:>19 (ꢁ90% d.e.).
MeCN: PdCl2 (5 mg, 0.028 mmol) in dry MeCN (0.5 ml) (brown!yellow); p-benzoquinone
(26 mg, 0.24 mmol), LiCl (21 mg, 0.495 mmol), AcONa (40 mg, 0.488 mmol), substrate 5 (100
mg, 0.24 mmol) in dry MeCN (2.5 ml); 50ꢀC, 24 h; colour of the reaction mixture changed from
brown to black; mixture of l-gulo 3 and l-ido 4 lactones 46 mg (43%) in a ratio 1.9:1 (31%
d.e.).
4.3.3. Conditions C
AcOH: Pd(OAc)2 (9 mg, 38 mmol), Cu(OAc)2 (206 mg, 1.136 mmol), AcONa (93 mg, 1.136
mmol), substrate 5 (158 mg, 0.3787 mmol) in glacial AcOH (5 ml); 50ꢀC, 23 h; colour of the
reaction mixture changed from blue±green to black; mixture of l-gulo 3 and l-ido 4 lactones 21
mg (19%) in a ratio 5.4:1 (69% d.e.).
THF: Pd(OAc)2 (6 mg, 24 mmol), Cu(OAc)2 (131 mg, 0.719 mmol), AcONa (59 mg, 0.719
mmol), substrate 5 (100 mg, 0.24 mmol) in dry THF (5 ml); 50ꢀC, 48 h; mixture of l-gulo 3 and
l-ido 4 lactones 31 mg (29%) in a ratio 1:1.4 (17% d.e.).
MeCN: Pd(OAc)2 (6 mg, 24 mmol), Cu(OAc)2 (131 mg, 0.719 mmol), AcONa (59 mg, 0.719
mmol), substrate 5 (100 mg, 0.24 mmol) in dry MeCN (5 ml); rt, 48 h; mixture of l-gulo 3 and
l-ido 4 lactones 28 mg (26%) in a ratio 1:1.3 (13% d.e.).
4.4. Stoichiometric aminocarbonylation of an alkene 5 using PdCl2
A dry, 25 ml one-necked round-bottomed ¯ask equipped with a side-arm stopcock, containing
a magnetic stirring bar, PdCl2 (43 mg, 0.24 mmol, 1 equiv.) and AcONa (59 mg, 0.719 mmol, 3
equiv.) was ®tted with a CO balloon and aminoalkenetriol 5 (100 mg, 0.24 mmol, 1 equiv.)
dissolved in glacial AcOH (3 ml) was added. The apparatus was purged with CO and the resulting
green mixture was stirred at 50ꢀC for 7 h while the colour of the reaction turned black. The
mixture was ®ltered on a Celite1 pad, washed several times with AcOH and the ®ltrate was
evaporated in vacuo. The last traces of the solvent were removed by co-evaporation with toluene
and the residue was ®ltered through short plug of silica gel yielding the crude mixture of two
diastereoisomeric lactones 3 and 4 (l-gulo+l-ido) as an oil (34 mg, 32%) in a ratio 1:1.9 (32%
d.e.).

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2593
All crude mixtures of lactones from aminocarbonylations were combined and l-gulo 3 and l-ido
4 lactones were separated by FLC on silica gel (gradient elution with hexanes:AcOEt=100:0!
95:1!9:1!4:1) yielding pure compounds.
4.5. 3,7-Anhydro-2,7-dideoxy-N-benzyl-5,6-di-O-benzyl-3,7-imino-l-gulo-1,4-heptonolactone 3
Pale yellow oil; Rf (0.28, 25% AcOEt/hexanes); ‰ꢁŠ2D4=^64.9 (c 1.23 in CH2Cl2); ꢀH (299.94
MHz, CDCl3) 2.07 (1H, `t', J1n,2=10.6, J1n,1x=11.1, H-1n), 2.41 (1H,`t', J5,6n=2.4, J6x,6n=14.2,
H-6n), 2.59 (1H, dd, J5,6x=6.2, J6x,6n=14.4, H-6x), 2.56±2.63 (1H, m, H-5), 3.11 (1H, dd,
J1x,2=4.3, J1n,1x=11.1, H-1e), 3.35 (1H, d, J=13.5, NCH2Ph), 3.60±3.68 (2H, m, J1x,2=4.7,
J3,4=8.8, H-2, H-3), 3.70 (1H, d, J=13.1, NCH2Ph), 3.97 (1H,`t', J3,4=9.0, J4,5=9.2, H-4), 4.57
(1H, d, J=11.5, OCH2Ph), 4.73 (1H, d, J=11.4, OCH2Ph), 4.74 (1H, d, J=11.4, OCH2Ph), 4.91
(1H, d, J=11.5, OCH2Ph), 7.25±7.39 (15H, m, 15 CH of Ph); ꢀC (75.43 MHz, CDCl3) 35.5 (t,
C-6), 57.2 (t, C-1), 60.0 (t, NCH2Ph), 64.6 (d, C-5), 73.5 (t, OCH2Ph), 73.8 (t, OCH2Ph), 77.9 (d,
C-2), 82.2 (d, C-3), 84.3 (d, C-4), 127.6, 127.7, 127.8, 127.9, 128.4, 128.5, 128.8 (all d, all CH of
Ph), 136.8, 138.0, 138.1 (all s, all Cipso of Ph), 173.9 (s, C-7); ꢂmax (®lm)/cm^1 1794 (s, CˆO); m/z
(EI) 443 (M+^1); found: C, 75.64; H, 6.69; N, 3.10; C28H29O4N requires: C, 75.82; H, 6.59; N,
3.16%.
4.6. 3,7-Anhydro-2,7-dideoxy-N-benzyl-5,6-di-O-benzyl-3,7-imino-l-ido-1,4-heptonolactone 4
Pale yellow oil; Rf (0.34, 25% AcOEt/hexanes); ‰ꢁŠ3D0=^6.05 (c 0.74 in CH2Cl2); ꢀH (299.94
MHz, CDCl3) 2.34 (1H, dd, J5,6x=7.4, J6n,6x=17.0, H-6x), 2.53 (1H, dd, J1n,2=7.8, J1n,1x=12.5,
H-1n), 2.61 (1H, dd, J5,6n=10.1, J6n,6x=17.1, H-6n), 2.81 (1H, dd, J1x,2=3.5, J1n,1x=12.8, H-1x),
3.51 (1H, d, J=13.5, NCH2Ph), 3.56±3.62 (2H, m, H-2, H-3), 3.63±3.73 (1H, m, H-5), 3.68 (1H,
d, J=13.7, NCH2Ph), 4.52 (1H, `t', J3,4J4,5=7.3, H-4), 4.54 [1H, d, J=12.0, (C-2)OCH2Ph],
4.60 [1H, d, J=11.7, (C-2)OCH2Ph], 4.74 [1H, d, J=11.3, (C-3)OCH2Ph], 4.82 [1H, d, J=11.3,
(C-3)OCH2Ph], 7.20±7.34 (15H, m, 15 CH of Ph); ꢀC (75.43 MHz, CDCl3) 27.7 (t, C-6), 48.6 (t,
C-1), 58.1 (d, C-5), 59.2 (t, NCH2), 72.3 [t, (C-2)OCH2], 74.1 [t, (C-3)OCH2], 76.4 (d, C-3), 81.2
(d, C-4), 81.3 (d, C-2), 126.7, 127.7, 127.9, 128.4, 128.5, 128.6 (all d, all CH of Ph), 137.1, 138.0,
138.1 (3s, 3Cipso of Ph), 174.7 (s, C-7); ꢂmax (®lm)/cm^1 1779 (s, CˆO); m/z (EI) 444 (M+); found:
C, 75.64; H, 6.69; N, 3.10; C28H29O4N requires: C, 75.82; H, 6.59; N, 3.16%.
4.7. N-Benzyl-2,3-di-O-benzyl-6-chloro-1,5,6-trideoxy-1,5-imino-l-iditol 11
Pale yellow oil; Rf (0.7, 25% AcOEt/hexanes); ‰ꢁŠ2D5=+35.1 (c 0.7 in CH2Cl2); ꢀH (299.94 MHz,
CDCl3) 2.63 (1H, dd, J1e,2=2.6, J1a,1e=12.8, H-1e), 2.83 (1H, dd, J1a,2=5.1, J1a,1e=12.8, H-1a),
3.12 (1H, `dt', J4,5=3.0, J5,6a=6.1, J5,6b=6.2, H-5), 3.35 (1H, d, J4,OH=9.0, OH), 3.53 (1H, d,
J=14.2, NCH2Ph), 3.56 (1H, m, H-2), 3.66 (1H, `t', J2,3J3,4 4.8, H-3), 3.85 (2H, `d', J=6.7, H-
6), 3.99 (1H, m, H-4), 4.08 (1H, d, J=13.8, NCH2Ph), 4.31 (1H, d, J=11.8, OCH2Ph), 4.41 (1H,
d, J=11.8, OCH2Ph), 4.60 (1H, d, J=11.8, OCH2Ph), 4.67 (1H, d, J=11.8, OCH2Ph), 7.17±7.34
(15H, m, 15 CH of Ph); ꢀC (75.43 MHz; CDCl3) 42.2 (t, C-6), 49.5 (t, C-1), 58.0 (t, NCH2Ph), 63.7
(d, C-5), 69.4 (d, C-4), 71.0 (t, OCH2Ph), 72.9 (t, OCH2Ph), 75.6 (d, C-2), 76.9 (d, C-3), 127.2,
127.6, 127.7, 127.8, 127.9, 128.1, 128.2, 128.4, 128.5, 128.6 (all d, all CH of Ph), 137.6, 138.1,
139.2 (all s, all Cipso of Ph); ꢂmax (®lm)/cm^1 1102 (s); m/z (EI) 451 (M+^1); found: C, 71.74; H,
6.73; N, 3.14, Cl, 7.83; C27H30O3NCl requires: C, 71.75; H, 6.69; N, 3.10, Cl, 7.84%.

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2594
4.8. Chloroaminocyclisation of an alkene 5 using PdCl2/CuCl2
The aminoalkenetriol 5 (50 mg, 0.12 mmol) dissolved in glacial AcOH (1.5 ml) was added to a
mixture of PdCl2 (4 mg, 0.24 mmol, 0.2 equiv.), CuCl2 (48 mg, 0.36 mmol, 3 equiv.) and AcONa
(29 mg, 0.36 mg, 3 equiv.) and stirred at rt for 48 h. The green suspension was ®ltered on a Celite1
pad, washed with AcOH and evaporated in vacuo. The last traces of the residual solvent were
removed by co-evaporation with toluene and the crude residue was puri®ed by FLC (10%
AcOEt/hexanes) yielding chloro derivative 11 as a pale yellow oil (37 mg, 70%). The physical and
spectroscopical data were identical to those of chloro derivative 11 previously obtained from
aminocarbonylation of 5.
4.9. N-Benzyl-2,3-di-O-benzyl-1,5,6-trideoxy-1,5-imino-d-gluco-heptitol 12
The l-gulo lactone 3 (77 mg, 0.174 mol) was dissolved in dry THF (5 ml), cooled to 0ꢀC and
LiBH4 (23 mg, 1.044 mmol, 6 equiv.) was added under Ar with vigorous stirring. After 3 h at rt,
the mixture was carefully neutralised with aq. HCl (pHꢂ7) and diluted with water (30 ml). The
mixture was then extracted with AcOEt (3Â30 ml) and dried over MgSO4. Evaporation in vacuo
furnished the crude solid, which was puri®ed by FLC (silica gel, 50% AcOEt/hexanes) yielding
d-gluco diol 12 (Rf 0.15) as a white solid (50 mg, 64%); m.p. 97±98ꢀC; ‰ꢁŠD31=^19.4 (c 0.29 in
CH2Cl2); ꢀH (299.94 MHz, CDCl3) 1.99±2.08 (2H, m, J1a,1b=11.9, H-1a, H-6a), 2.13±2.28 (1H, m,
H-6b), 2.51±2.63 (1H, m, J5,6aJ5,6b=4.5, J4,5=9.0, H-5), 3.03 (1H, dd, J1b,2=4.4, J1a,1b=12.1,
H-1b), 3.10±3.40 (2H, b, 2OH), 3.36 (1H, `t', J2,3J3,4=8.5, H-3), 3.35 (1H, d, J=13.7,
NCH2Ph), 3.60 (1H, `dt', J1b,2=4.3, J2,3=8.6, H-2), 3.66 (1H, `t', J3,4=8.6, J4,5=9.0, H-4), 3.74±
3.93 (2H, m, H-7), 4.13 (1H, d, J=13.3, NCH2Ph), 4.45 (1H, d, J=11.5, OCH2Ph), 4.52 (1H, d,
J=11.5, OCH2Ph), 4.70 (1H, d, J=11.5, OCH2Ph), 4.98 (1H, d, J=11.4, OCH2Ph), 7.18±7.40
(15H, m, 15 CH of Ph); ꢀC (75.43 MHz; CDCl3) 29.4 (t, C-6), 53.1 (t, C-1), 56.6 (t, NCH2Ph), 60.2
(t, C-7), 64.1 (d, C-5), 71.2 (d, C-4), 72.2 (t, OCH2Ph), 74.8 (t, OCH2Ph), 77.2 (d, C-2), 85.2 (d,
C-3), 127.4, 127.7, 127.8, 127.9, 128.4, 128.5, 128.6, 128.9 (all d, all CH of Ph), 138.0, 138.6 (2s,
2Cipso of Ph); ꢂmax (KBr)/cm^1 3403 (bs, OH); m/z (EI) 417 (M+^CH2OH); found: C, 74.86; H,
7.48; N, 3.02; C28H33O4N requires: C, 75.12; H, 7.44; N, 3.13%.
4.10. N-Benzyl-2,3-di-O-benzyl-1,5,6-trideoxy-1,5-imino-l-ido-heptitol 13
The l-ido lactone 4 (510 mg, 1.151 mmol) was dissolved in dry THF (15 ml), cooled to 0ꢀC and
LiBH4 (100 mg, 4.602 mmol, 4 equiv.) was added under Ar with vigorous stirring. After 3 h at rt,
the mixture was carefully neutralised with aq. HCl (pHꢂ7) and diluted with water (30 ml). The
mixture was then extracted with AcOEt (3Â30 ml) and dried over MgSO4. Evaporation in vacuo
furnished a crude oil, which was puri®ed by FLC (silica gel, 50% AcOEt/hexanes) yielding l-ido
diol 13 (Rf 0.22) as a colourless oil (331 mg, 65%); ‰ꢁŠ3D1=^1.32 (c 0.34 in CH2Cl2); ꢀH (299.94
MHz; CDCl3) 1.82±1.95 (2H, m, H-6), 2.74 (2H, `d', J=6.8, H-1), 3.05±3.15 (1H, m, J4,5=4.7,
H-5), 3.12±3.78 (6H, m, H-2, H-3, H-7, 2 OH), 3.60 (1H, d, J=13.3, NCH2Ph), 3.82 (1H, d,
J=13.2, NCH2Ph), 3.86 (1H, dd, J4,5=4.7, J3,4=8.4, H-4), 4.50 (2H, `s', OCH2Ph), 4.67 (1H, d,
J=11.5, OCH2Ph), 4.92 (1H, d, J=11.3, OCH2Ph), 7.23±7.40 (15H, m, 15 CH of Ph); ꢀC (75.43
MHz, CDCl3) 29.7 (t, C-6), 48.0 (t, C-1), 58.1 (t, NCH2Ph), 61.2 (d, C-5), 62.1 (t, C-7), 69.7 (d,
C-4), 71.9 (t, OCH2Ph), 74.5 (t, OCH2Ph), 76.6, 81.2 (2 d, C-2$C-3), 127.4, 127.7, 127.8, 127.9,
128.4, 128.5, 128.7 (all d, all CH of Ph), 137.9, 138.3, 138.4 (all s, all Cipso of Ph); ꢂmax (®lm)/cm^1

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2595
3370 (bs, OH); m/z (EI) 447 (M+^1); found: C, 74.92; H, 7.42; N, 3.13; C28H33O4N requires: C,
75.12; H, 7.44; N, 3.13%.
Alternatively, the LiBH4 reduction was carried out with a mixture of lactones 3 and 4. Thus,
the mixture of l-gulo 3 and l-ido 4 lactones (172 mg, 0.388 mmol) was dissolved in dry THF (10
ml), cooled to 0ꢀC and LiBH4 (34 mg, 1.552 mmol, 4 equiv.) was added under Ar with vigorous
stirring. After 3 h at rt, the mixture was carefully neutralised with aq. HCl (pH ꢂ7), diluted with
AcOEt (50 ml), washed with water (30 ml) and dried over MgSO4. Evaporation in vacuo furnished
a semicrystalline oil (164 mg), which was puri®ed by FLC (silica gel, gradient elution with hexanes:
AcOEt: 9:1!4:1!3:1!2:1) yielding two fractions: (1) fraction: l-ido diol 13 as a colourless oil
(75 mg, 43%); and (2) fraction: d-gluco diol 12 as a white solid (25 mg, 14%). The physical and
spectroscopic data were identical to those of previously obtained.
4.11. 1,5,6-Trideoxy-1,5-imino-d-gluco-heptitol hydrochloride 1
The d-gluco diol 12 (70 mg, 0.157 mmol) was dissolved in EtOH (8 ml) containing 35% HCl
(seven drops) and 10% Pd/C (44 mg) was added. The ¯ask was purged with hydrogen by repeated
pumping-®lling and the mixture was stirred under the H2 atmosphere overnight. The catalyst was
®ltered o€, washed with EtOH and the ®ltrate was evaporated in vacuo. The residue was
dissolved in deionised H2O (2 ml) and put on the top of a column of Dowex1 50WX8-400 ion-
exchange resin [1.5 g, H2O (3 ml), washed sequentially with H2O (15 ml)!MeOH (15 ml)!H2O
(15 ml) prior to use]. The column was washed with water to the neutrality (ca. 10 ml) and the
product was eluted with 2.5% NH4OH solution. Removal of volatiles in vacuo and redissolving
the residue in MeOH, addition of two drops of 35% HCl, ®ltration and evaporation a€orded the
hydrochloride salt of d-gluco homologue 1 as a colourless semicrystalline oil (30 mg, 90%), which
can be crystallised from MeOH/tBuOMe; m.p. 177±179ꢀC; ‰ꢁŠD26=+30 (c 0.55 in MeOH); ꢀH
(299.94 MHz, CD3OD) 1.86 (1H, `ddt', J6a,7b=4.6, J6a,7a=8.1, J6a,6b=12.8, H-6a), 2.27 (1H,
`ddt', J6b,7a=4.1, J6b,7b=5.9, J6a,6b=12.4, H-6b), 2.89 (1H, `t', J1a,1eJ1a,2=12.0, H-1a), 3.20
(1H, `dt', J4,5=8.5, J5,6a=9.5, H-5), 3.39 (1H, dd, J1e,2=4.7, J1a,1e=12.9, H-1e), 3.40 (1H, `t',
J4,5=8.2, J3,4=10.2, H-4), 3.44 (1H, `t', J2,3=8.9, J3,4=10.2, H-3), 3.73 (1H, ddd, J1e,2=5.0,
J2,3=8.8, J1a,2=11.5, H-2), 3.81 (1H, ddd, J6b,7a=4.4, J6a,7a=8.0, J7a,7b=12.3, H-7a), 3.91 (1H,
`dt', J6a,7b=4.7, J6b,7b=6.1, H-7b); ꢀC (75.43 MHz, CD3OD) 32.2 (t, C-6), 48.1 (t, C-1), 60.2 (t, C-
7), 60.8 (d, C-5), 68.5 (d, C-2), 72.5 (d, C-3), 78.3 (d, C-4); ꢂmax (®lm)/cm^1 3353 (bs, OH); m/z
(EI) 177 (M+^HCl); found: C, 39.21; H, 7.58; N, 6.38, Cl, 16.50, C7H16O4NCl requires: C, 39.35;
H, 7.55; N, 6.56, Cl, 16.59%.
4.12. 1,5,6-Trideoxy-1,5-imino-l-ido-heptitol hydrochloride 2
The l-ido diol 13 (112 mg, 0.25 mmol) was dissolved in EtOH (10 ml) containing 35% HCl (10
drops) and 10% Pd/C (70 mg) was added. The ¯ask was purged with hydrogen by repeated
pumping-®lling and the mixture was stirred under an H2 atmosphere overnight. The catalyst was
®ltered o€, washed with EtOH and ®ltrate evaporated in vacuo. The residue (58 mg) was
dissolved in deionised H2O (2 ml) and put on the top of a column of Dowex1 50WX8-400 ion-
exchange resin [4.6 g, H2O (5 ml), washed sequentially with H2O (20 ml)!MeOH (30 ml)!H2O
(20 ml) prior to use]. The column was washed with water to the neutrality (ca. 20 ml) and the
product was eluted with 2% NH4OH solution. Removal of volatiles in vacuo and redissolving the
residue in MeOH, addition of two drops of 35% HCl, ®ltration and evaporation a€orded a the

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2596
hydrochloride salt of l-ido homologue 2 as a colourless oil (44 mg, 82%); ‰ꢁŠ3D2=+20.2 (c 0.42 in
MeOH); {lit.:9f [ꢁ]=+13.85 (c 0.8 in H2O) for free base}; ꢀH (299.94 MHz, CD3OD) 1.95±2.05
(2H, m, J6,7=5.3, J5,6=6.7, H-6), 3.25 (1H, d, J1a,1b=12.8, H-1a), 3.46 (1H, d, J1a,1b=13.1,
H-1b), 3.60-3.71 (1H, `t', J5,6=6.4, H-5), 3.72±3.87 (2H, m, J6,7=5.2, H-7), 3.87±3.92 (1H, `s',
H-4), 3.93±4.20 (2H, m, H-2, H-3); ꢀC (75.43 MHz, CD3OD) 32.2 (t, C-6), 47.3 (t, C-1), 55.0 (d,
C-5), 59.0 (t, C-7), 67.9, 68.1 (2 d, C-2$C-3), 70.4 (d, C-4); ꢂmax (®lm)/cm^1 3800±3000 (bs, OH);
m/z (EI) 177 (M+^HCl); found: C, 39.56; H, 7.57; N, 6.52, Cl, 16.23, C7H16O4NCl requires: C,
39.35; H, 7.55; N, 6.56, Cl, 16.59%.
Acknowledgements
We are grateful to Professor V. Jager (Stuttgart) for helpful ideas, discussions and support.
This research was supported by VW Stiftung (Hannover) and The Slovak Grant Agency. We
thank TauChem Ltd (Bratislava) for providing us with carbon monoxide and Dr. Darren J.
Dixon (Cambridge) for proof reading and comments.
References
1. Stutz, A. E. Iminosugars as Glycosidase Inhibitors; Wiley-VCH: Weinheim, 1998.
2. (a) Winchester, B.; Fleet, G. W. J. Glycobiology 1992, 2, 199; (b) Ganem, B. Acc. Chem. Res. 1996, 29, 340.
3. (a) Taylor, R. H.; Barker, H. M.; Bowey, E. A.; Can®eld, J. E. Gut 1986, 27, 1471; (b) Yoshikuni, Y. Trends
Glycosci. Glycotech. 1991, 3, 184.
4. Olden, K.; Breton, P.; Grzegorzewski, K.; Yasuda, B.; Gause, B. L.; Oredipe, O. A.; Newton, S. A.; White, S. L.
Pharmacol. Ther. 1991, 50, 285.
5. Elbein, A. D.; Legler, G.; Tlusty, A.; McDowell, W.; Schwarz, R. Arch. Biochem. Biophys. 1984, 235, 579.
6. (a) Ratner, L. AIDS Res. Hum. Retroviruses 1992, 8, 165; (b) van der Broek, L. A. G. M.; Vermaas, D. J.;
Heskamp, B. M.; van Boeckel, C. A. A.; Tan, M. C. A. A.; Bolscher, J. G. M.; Ploegh, H. L.; van Kemenade, F. J.;
de Goede, R. E. Y.; Miedema, F. Recl. Trav. Chim. Pays-Bas 1993, 112, 82; (c) Asano, N.; Kizu, H.; Oseki, K.;
Tomioka, E.; Matsiu, K.; Okamoto, M.; Baba, M. J. Med. Chem. 1995, 38, 2349.
7. (a) Platt, F. M.; Neises, G. R.; Reinkensmeier, G.; Townsend, M. J.; Perry, V. H.; Proia, R. L.; Winchester, B.;
Dwek, R. A.; Butters, T. D. Science 1997, 276, 428; (b) Kolter, T. Angew. Chem., Int. Ed. Engl. 1997, 36, 1955.
8. (a) Saavedra, O. M.; Martin, O. R. J. Org. Chem. 1996, 61, 6987; (b) Chakraborty, T. K.; Jayaprakash, S.
Tetrahedron Lett. 1997, 38, 8899; (c) Shilvock, J. P.; Nash, R. J.; Lloyd, J. D.; Winters, A. L.; Asano, N.; Fleet,
G. W. J. Tetrahedron: Asymmetry 1998, 9, 3505; (d) Jotterand, N.; Vogel, P. J. Org. Chem. 1999, 64, 8973.
9. (a) Rassu, G.; Pinna, L.; Spanu, P.; Culeddu, N.; Casiraghi, G.; Gasparri Fava, G.; Belicchi Ferrari, M.; Pelosi, G.
Tetrahedron 1992, 48, 727; (b) Kilonda, A.; Compernolle, F.; Toppet, S.; Hoornaert, G. J. Tetrahedron Lett. 1994,
35, 9047; (c) Lundt, I.; Madsen, R. Synthesis 1995, 787; (d) Herdeis, C.; Schi€er, T. Tetrahedron 1996, 52, 14745;
(e) Compernolle, F.; Joly, G.; Peeters, K.; Toppet, S.; Hoornaert, G. J.; Kilonda, A.; Babady, B. Tetrahedron
1997, 53, 12739; (f) Desai, V. N.; Saha, N. N.; Dhavale, D. D. Chem. Commun. 1999, 1719.
10. (a) Chen, Y.; Vogel, P. J. Org. Chem. 1994, 59, 2487; (b) Picasso, S.; Chen, Y.; Vogel, P. Carbohydr. Lett. 1994, 1,
1; (c) Baudat, A.; Picasso, S.; Vogel, P. Carbohydr. Res. 1996, 281, 277; (d) Shilvock, J. P.; Hsia, K. Y.; Nash, R. J.;
Lloyd, J. D.; Winters, A. L.; Asano, N.; Fleet, G. W. J. Tetrahedron: Asymmetry 1998, 9, 4157.
11. For the preliminary report, see: Szolcsanyi, P.; Gracza, T.; Koman, M.; Pronayova, N.; Liptaj, T. Chem. Commun.
2000, 471.
12. (a) Tamaru, Y.; Kobayashi, T.; Kawamura, S.; Ochiai, H.; Yoshida, Z. Tetrahedron Lett. 1985, 26, 4479; (b)
Tamaru, Y.; Yoshida, Z. J. Organomet. Chem. 1987, 334, 313; (c) Tamaru, Y.; Hojo, M.; Yoshida, Z. J. Org.
Chem. 1988, 53, 5731.
13. (a) Jager, V.; Hummer, W. Angew. Chem., Int. Ed. Engl. 1990, 29, 1171; (b) Hasenohrl, T. PhD Thesis, University
of Stuttgart, 1995; (c) Jager, V.; Gracza, T.; Dubois, E.; Hasenohrl, T.; Hummer, W.; Kautz, U.; Kirschbaum, B.;

Â
P. Szolcsanyi et al. / Tetrahedron: Asymmetry 11 (2000) 2579±2597
2597
Lieberknecht, A.; Remen, L.; Shaw, D.; Stahl, U.; Stephan, O. In Pd(II)-Catalyzed Carbonylation of Unsaturated
Polyols and Aminopolyols; Helmchen, G.; Dibo, J.; Flubacher, D.; Wiese, B., Eds. Organic Synthesis via
Organometallics OSM 5. Vieweg: Braunschweig, 1997; pp. 331±360.
14. Oh, Ch.; Kim, K.; Ham, W. Tetrahedron Lett. 1998, 39, 2133.
15. Hummer, W.; Dubois, E.; Gracza, T.; Jager, V. Synthesis 1997, 634.
16. Tamaru, Y.; Kimura, M. Synlett 1997, 749.
17. Davies, S. G.; Goodfellow, C. L. J. Chem. Soc., Perkin Trans 1 1990, 393.
18. Ferro, V.; Mocerino, M.; Stick, R. V.; Tilbrook, D. M. G. Aust. J. Chem. 1988, 41, 813.
19. Richtmyer, N. K. Methods Carbohydr. Chem. 1962, 1, 107.
20. Anisuzzamann, N. K.; Whistler, R. L. Carbohydr. Res. 1978, 61, 511.
21. (a) Bernotas, R. C.; Pezzone, M. A.; Ganem, B. Carbohydr. Res. 1987, 167, 305; (b) Bernotas, R. C. Tetrahedron
Lett. 1990, 31, 469.
22. (a) Lai, J.; Wang, F.; Guo, G.; Dai, L. J. Org. Chem. 1993, 58, 6944; (b) Quisari, A.; Hamed, O.; Henry, P. M.
J. Org. Chem. 1998, 63, 2790.
23. Hegedus, L. S.; McKearin, J. M. J. Am. Chem. Soc. 1982, 104, 2444.
24. Wada, M.; Akiba, K. Heterocycles 1987, 26, 929.
25. Zargarian, D.; Alper, H. Organometallics 1991, 10, 2914.
26. Koman, M.; Szolcsanyi, P.; Gracza, T. Acta Crystallogr. 2000, c56, e138.
27. Morie, T.; Kato, S. Heterocycles 1998, 48, 427.