Catalytic, asymmetric synthesis of phosphonic γ-(hydroxyalkyl) butenolides with contiguous quaternary and tertiary stereogenic centers

Article abstract of DOI:10.1002/chem.201304331

A procedure that enables high yielding access to phosphonic γ-(hydroxyalkyl)butenolides with excellent regio-, diastereo- and enantiocontrol is reported. The simultaneous construction of up to two adjacent quaternary stereogenic centers by a catalytic asymmetric vinylogous Mukaiyama aldol reaction unites biologically and medicinally relevant entities, namely α-hydroxy phosphonates and γ-(hydroxyalkyl)butenolides. This is achieved by utilizing a readily available chiral copper-sulfoximine catalyst showing a broad functional group tolerance for both the electrophilic and nucleophilic reactants. A discussion about potential factors affecting the observed level of enantioselectivity, which stems from the enantiopure sulfoximine ligand, is also included. Best of two worlds: A chiral copper-sulfoximine catalyst unites biologically relevant α-hydroxy phosphonates and γ-(hydroxyalkyl) butenolides into one structural motif with regio-, diastereo- and enantioselectivity up to the limits of detection (see scheme; Tf=trifluoromethanesulfonyl, TMS=trimethylsilyl). Copyright

Full text of DOI:10.1002/chem.201304331

DOI: 10.1002/chem.201304331
Full Paper
&
Synthetic Methods
Catalytic, Asymmetric Synthesis of Phosphonic
g-(Hydroxyalkyl)butenolides with Contiguous Quaternary and
Tertiary Stereogenic Centers
Marcus Frings,*[a] Isabelle Thomꢀ,[a] Ingo Schiffers,[a] Fangfang Pan,[b] and Carsten Bolm*[a]
Abstract: A procedure that enables high yielding access
to phosphonic g-(hydroxyalkyl)butenolides with excellent
regio-, diastereo- and enantiocontrol is reported. The simul-
taneous construction of up to two adjacent quaternary ste-
reogenic centers by a catalytic asymmetric vinylogous Mu-
kaiyama aldol reaction unites biologically and medicinally
relevant entities, namely a-hydroxy phosphonates and
g-(hydroxyalkyl)butenolides. This is achieved by utilizing
a readily available chiral copper-sulfoximine catalyst showing
a broad functional group tolerance for both the electrophilic
and nucleophilic reactants. A discussion about potential fac-
tors affecting the observed level of enantioselectivity, which
stems from the enantiopure sulfoximine ligand, is also in-
cluded.
Introduction
The ability to utilize phosphorus is one of the essential prereq-
uisites for any known form of planetary life, and for a long
time it was an accepted doctrine that all earthly organophos-
phorus molecules from natural sources consist of a fully oxi-
dized phosphorus atom such as in phosphate esters. A turning
point was marked by the discovery of the first phosphonic
acid in a living organism.[1] Since this groundbreaking finding,
plenty of related substances, in which the respective phospho-
rus atoms reside in lower oxidation states, have been located
in microorganisms and higher life forms.[2] While large parts of
their biosyntheses have not been completely understood yet,
and many of the associated metabolic processes still appear
enigmatic,[3] biogenic phosphonic acids became attractive tar-
gets for man-made syntheses.[4]
Scheme 1. Chiral a-hydroxy phosphonates (circled) and g-(hydroxyalkyl)-
butenolides (boxed) are connected in a single molecule 1 synthetically ac-
cessed from a-keto phosphonates 2 and silyloxy furan 3a.
nodeficiency virus[6] (HIV) over herbicidal characteristics[7] to an-
tioxidant properties.[8] As 1-hydroxy-1,1-bisphosphonates they
embody a precious class of compounds for the therapeutic
treatment of osteoporosis and similar bone diseases.[9,10] Be-
sides, a-hydroxy phosphonates interfere with various metabol-
ic pathways and inhibit a wide range of fundamentally impor-
tant enzymes such as renin,[11] CD-45 tyrosine phosphatase,[12]
tyrosine-specific protein kinases,[13] farnesyl transferase,[14] myo-
inositol monophosphatase,[15] 5-enolpyruvoylshikimate-3-phos-
phate synthase,[16] or acetylcholinesterase.[17]
A particularly interesting subset within the numerous isolat-
ed natural organophosphorus compounds is based on
a-hydroxy phosphonic acids or their derivatives.[5] Synthetic a-
hydroxy phosphonates with their general structure depicted in
Scheme 1 (circled) emerged as a sine qua non for modern bio-
medical and pharmaceutical research because of the spectacu-
lar array of diversified biological activities. For example, their
physiological profiles range from activity against human immu-
Two intrinsic properties help them to unfold their impressive
biological potential: On the one hand, a-hydroxy phosphonic
acids constitute close analogs of natural a-hydroxy carboxylic
acids with the crucial difference that the geometry of the acid
moiety is tetrahedral instead of trigonal planar—on the other
hand, phosphonates are able to mimic metabolically regulating
phosphate esters. This allows them to serve as bioisosteric and
isopolar replacements.[18] Apart from these biochemically ori-
ented applications, they also represent very useful precursors
in the repertoire of organic chemists because a-hydroxy phos-
phonates can further be transformed into many a-functional-
ized derivatives.[19]
[a] M. Frings, Dr. I. Thomꢀ, Dr. I. Schiffers, Prof. Dr. C. Bolm
Institute of Organic Chemistry, RWTH Aachen University,
Landoltweg 1, 52074 Aachen (Germany)
Fax: (+49)241-8092391
E-mail: marcus.frings@oc.rwth-aachen.de
carsten.bolm@oc.rwth-aachen.de
[b] Dr. F. Pan
Institute of Inorganic Chemistry, RWTH Aachen University
Landoltweg 1, 52074 Aachen (Germany)
Other prominent structural motifs in the biota are buteno-
lides and the allied saturated butyrolactones, respectively.[20]
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304331.
Chem. Eur. J. 2014, 20, 1691 – 1700
1691
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Quite commonly, the g-position of the lactone core is accom-
panied by substituted alkyl fragments with secondary or terti-
ary alcohols (Scheme 1, boxed). Among these molecules, anti-
HIV activity, anti-inflammatory or anti-proliferative properties
are just a few of the discovered biological characteristics.[21]
With respect to the very potent structure–activity relation-
ships of the aforementioned, independently in nature occur-
ring entities, it is a logical step to fuse them on a molecular
level to form chiral phosphonic g-(hydroxyalkyl)butenolides
1 (Scheme 1). Given the fact that 1) most of the biogenic
g-(hydroxyalkyl)butenolides exist only as one stereoisomer and
2) in a-hydroxy phosphonates the vicinal stereogenic carbon
center of the phosphorus is able to dictate the extent of bio-
logical activity,[11c,22] an enantioselective access towards the ter-
tiary a-hydroxy phosphonates 1 is desirable. Synthetically chal-
lenging, these structures bear a quaternary stereogenic center
adjacent to the tertiary one of the lactone core. In the past,
the number of syntheses for optically active secondary a-hy-
droxy phosphonates increased continuously by means of
chemical or enzymatic methods.[23,24] On the contrary, enan-
tioenriched tertiary a-hydroxy phosphonates were untouched
in literature until 2006. Zhao and Samanta pointed the way
with a seminal work on asymmetric organocatalytic aldol addi-
tions of a-keto phosphonates 2 and acetone, and later others
followed by synthesizing non-racemic tertiary a-hydroxyphos-
phono esters by asymmetric aldol, allylation, or hydrophospho-
nylation reactions.[25]
Figure 1. Chiral sulfoximines L1–L8 probed as ligands in this study. L1–L6
and L8 bear (S) configuration at sulfur, L7 is derived from a (R)-configured
sulfoximine.
Table 1. Influence of solvents and additives on the enantioselective VMA
reaction of a-keto phosphonate 2a and dienol silane 3a.[a]
Results and Discussion
Solvent
Additive
Yield
[%][b]
de
[%][c]
ee
[%][d]
The search for novel lead structures with hopefully promising
pharmacological qualities requires a reliable protocol to quickly
build up complex molecules from a suitable library of precur-
sors. To construct the framework of 1, asymmetric vinylogous
Mukaiyama aldol (VMA) reactions between readily available ke-
tonic phosphonate derivatives 2 and 2-(trimethylsilyloxy)furan
(3a) appeared an obvious solution.[26,27] Nevertheless, it must
be emphasized that catalytic aldol additions to ketones with
high levels of enantioselectivity are still rare and comparatively
difficult to achieve.[28]
1
2
3
4
5
6
7
8
9
10
11
12[e]
13
DCM
toluene
THF
1,4-dioxane
TBME
iPr2O
none
none
none
none
none
none
none
HFIP
TFE
91
71
62
83
43
69
79
74
85
80
89
92
91
96
99
98
97
94
98
97
97
98
89
97
98
99
89
94
93
95
96
94
95
96
85
68
81
98
91
Et2O
1,4-dioxane
1,4-dioxane
1,4-dioxane
Et2O
Et2O
Et2O
iPrOH
HFIP
TFE
A recent publication of Miao, Chen and co-workers describes
a copper-catalyzed diastereoselective VMA reaction between
a-keto phosphonates 2 and dienol silane 3a to give the race-
mic adducts 1 in variable yields from 46–89% and with diaste-
reomeric ratios starting at 83:17.[29] Their work stimulated us to
investigate a method for the enantiocontrolled preparation of
phosphonic g-(hydroxyalkyl)butenolides 1 that, at the outset of
our studies, had not been reported, yet.[30] Based on our expe-
rience in the use of chiral C1-symmetric sulfoximines as ligands
in copper-catalyzed stereoselective Mukaiyama-type reac-
tions,[31,32] we assumed that such hemilabile N,N’-chelators
(Figure 1) could also be useful in catalytic asymmetric VMA re-
actions of ketonic phosphonates 2 and nucleophile 3a to give
the desired optically active aldol products 1.
iPrOH
[a] Additional reaction conditions: 2a (0.20 mmol), additive (1.2 equiv), 3a
(1.1 equiv), solvent (2 mL), over night. Only the major stereoisomer of 1a is
displayed. [b] Total yield of unseparated diastereomers of 1a after column
chromatography. [c] The de values always refer to the anti product of 1a.
For details on its determination, see the Supporting Information. [d] The ee
values always refer to (S,R)-1a as the major enantiomer. [e] Full conversion
of 2a within 30 min.
catalyzed by a complex derived from 10 mol% of Cu(OTf)2 and
enantiopure amino sulfoximine L1 (Table 1, entries 1–7). Al-
ready the first experiment with dichloromethane (DCM) pro-
ceeded with promising stereocontrol, and the desired tertiary
a-hydroxy phosphonate 1a was formed in high yield (91%)
with a diastereomeric excess (de) of 96% and an enantiomeric
excess (ee) of 89% (entry 1). By switching the solvent to tolu-
To test this hypothesis, we initially explored addition reac-
tions of furan derivative 3a to diethyl benzoylphosphonate
(2a) in various solvents at room temperature (RT), which were
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1692
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ene the major stereoisomer of 1a could further be enriched
(99% de and 94% ee), but unfortunately at the expense of a re-
duced yield of 71% (entry 2).
Furthermore, another aspect of TFE, that has not been prop-
erly addressed in this discussion yet, has to be mentioned. For
this additive, a very pronounced rate accelerating effect was
observed in the solvents diethyl ether, toluene, and THF. Track-
ing the course of each reaction by thin layer chromatography
(TLC) revealed that all three reactions were completed within
30 min and a-keto phosphonate 2a was fully consumed. If no
fluorinated alcohol was present, the catalysis proceeded slower
and 2a could still be detected by TLC a day later. Thus, the ad-
ditive plays a dual role. Apart from facilitating the catalyst turn-
over and thus giving higher yields in shorter times, it is able to
affect the asymmetric induction. Although we are still lacking
a mechanistic model of our catalyst structure we propose that
TFE coordinates to the chiral copper-sulfoximine complex gen-
erating a more stereoselective species, which could be ex-
plained by the concept of Brønsted acid assisted Lewis acid cat-
alysis.[34,35] This might also be reflected by the fact that in each
solvent the reaction mixture undergoes a very characteristic
color change when the alcoholic additive is absent. Coordina-
tion of amino sulfoximine L1 to Cu(OTf)2 forms a green solu-
tion which, upon adding the reactants 2a, TFE and 3a in exact-
ly the given order, keeps this color until the end. If the second
step (the injection of the additive) is skipped, the green solu-
tion with the chiral catalyst and electrophile inside turns black-
ish blue within seconds after final addition of furan 3a. Since
diethyl ether and TFE worked best together in terms of yield
and ee, these two parameters were fixed for the following ex-
periments.
Next, various cyclic and linear ethers were applied. Fully
aware of the fact that these solvents can interact with the
chiral copper catalyst because of their ability to coordinate to
the Lewis acid we anticipated finding a beneficial influence.
With tetrahydrofuran (THF) the remarkable stereochemical dif-
ferentiation could be maintained (98% de and 93% ee), yet 1a
was isolated only in a moderate yield of 62% (entry 3). As
demonstrated by entry 4, the VMA reaction proceeded much
better in 1,4-dioxane. With almost no decline in diastereoselec-
tivity (97% de), 1a was obtained in a satisfactory yield (83%)
and the highest ee (95%) up to this point. Regarding the for-
mation of 1a, the series of linear ethers showed a clear trend
(entries 5–7). From the results observed with tert-butyl methyl
ether (TBME), diisopropyl ether and diethyl ether it can be
seen that the extent of the yields of 1a (43, 69, and 79%, re-
spectively) diametrically depended on the decreasing steric
demand of the ether’s alkyl group. Apparently, the asymmetric
induction (94–98% de and 94–96% ee) was not altered much
by using these ethers. Although TBME furnished the best ee, it
provided by far the lowest yield. Consequently, this solvent
was considered to be not viable for the future study of the re-
action.
To further optimize the conditions a few alcoholic additives
were scrutinized, for they are known to be beneficial in metal-
catalyzed Mukaiyama-type reactions.[29–31] Since 1,4-dioxane
and diethyl ether proved equally well in terms of stereocontrol
(Table 1, entries 4 and 7), the reactions were repeated with 1.2
equivalents of 1,1,1,3,3,3-hexafluoroisopropanol (HFIP), 2,2,2-tri-
fluoroethanol (TFE) or isopropanol. Concerning the cyclic ether,
no substantial improvement was found for HFIP (Table 1,
entry 8). The ee of 1a attained again the previous maximum of
96%, albeit to the detriment of a lower yield (74%). Interest-
ingly, both the chiral catalyst’s activity and its selectivity were
reversed in the presence of TFE because 1a was provided in
a slightly better yield of 85%, but with a considerably de-
creased ee of 85% (Table 1, entry 9). As revealed by entry 10,
the surprisingly strong decrease in enantioselectivity (68% ee)
continued when isopropanol was added. Regarding the enan-
tiocontrol in this dioxane-based system, the alcohol’s acidity
presumably takes precedence over steric or other factors of
the additive. Furthermore, isopropanol caused an unexpectedly
low diastereopreference (89% de), which was not observed
when fluorinated additives were applied (97 or 98% de in
Table 1, entries 8 and 9). In diethyl ether all three additives no-
ticeably enlarged the yield of 1a from 79% to a range of 89–
92% (Table 1, entry 7 vs. 11–13). The original level of diastereo-
selectivity either remained intact (97% de) or could even be
raised (98 or 99% de). Once again a strong, yet different rela-
tionship than before between each additive and the catalyst’s
asymmetric induction was evident. While especially HFIP and
isopropanol decreased the ee of 1a (81 and 91%, respectively),
TFE supported the enantioselective pathway to a great extent,
and its use culminated in an ee of 98% (Table 1, entry 12).[33]
Subsequently, the chiral copper catalyst was varied by utiliz-
ing other copper salts or sulfoximines (Table 2). Entry 2 states
that the asymmetric VMA reaction also worked with an analo-
gous copper(I) precatalyst providing the a-hydroxy phospho-
Table 2. Exploration of copper salts and sulfoximines L1–8 in the enan-
tioselective VMA reaction of a-keto phosphonate 2a and dienol silane 3a
to provide 1a.[a]
Copper salt
Ligand
Yield
[%][b]
de
[%][c]
ee
[%][d]
1[e]
2
3
4
5
6
7
8
9
10
11
12
13[g]
14[h]
15[i]
Cu(OTf)2
(CuOTf)2·toluene
Cu(SbF6)2
Cu(ClO4)2·6H2O
L1
L1
L1
L1
L1
L2
L3
L4
L5
L6
L7
L8
L1
L1
L1
92
75
93
83
97
88
87
99
98
88
69
91
98
99
95
98
98
82
98
98
94
98
99
97
95
98
98
>99
>99
99
98
96
9
93
96
67
97
94
96
11
[f]
f
Cu(ClO4)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
0
78
98
>99
97
[a] Reaction conditions: 2a (0.20 mmol), TFE (1.2 equiv), 3a (1.1 equiv),
copper salt (10 mol%), ligand (10 mol%), Et2O (2 mL), RT, over night.
[b–d] Same as in Table 1. [e] Identical with entry 12 of Table 1. [f] Prepared
in situ from CuCl2 and the respective silver salt. [g] Conducted at 108C.
[h] Conducted at 08C. [i] Performed with 5 mol% of the copper-sulfoxi-
mine complex at 08C.
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1693
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
nate under admirable stereocontrol (98% de and 96% ee).
Compared to the result obtained with Cu(OTf)2 the yield of 1a
decreased to 75% and it took significantly longer to fully con-
sume the starting material 2a. This could possibly indicate that
copper(I) is not the active species in the catalytic cycle and
that it needs to be transformed into a higher oxidation state
before. Cu(SbF6)2 delivered 1a in a high yield of 93% but large-
ly reduced the diastereochemical differentiation (82% de).
Above all, it rendered the enantioselective pathway almost
void and gave the product with an ee of just 9% (entry 3). The
commercially available hexahydrate of Cu(ClO4)2 served as
basis for a catalyst with rather acceptable activity (83% yield)
and good enantioselectivity (93% ee), but the corresponding
anhydrous salt was more appropriate (Table 2, entries 4 and 5,
respectively). Indeed, the yield reached a new peak (97%) with
1a being isolated highly enantioenriched (96% ee). Besides,
both types of Cu(ClO4)2 allowed the formation of the major
diastereomer with excellent selectivity (98% de). Although the
latter copper(II) source actually represents an attractive alterna-
tive to Cu(OTf)2 we decided to renounce further optimization
steps with anhydrous Cu(ClO4)2 because it required a tedious
in situ preparation of the metal salt.
estingly, when the adjacent methylene unit of the secondary
amine in L1 was virtually substituted for a sulfone group the
enantioregulatory pathway was negated. With the consequen-
tial sulfonamido sulfoximine L7 butenolide 1a was obtained as
a racemate (Table 2, entry 11). However, ligand L8 with con-
strained flexibility at the non-sulfonimidoyl nitrogen (which
was realized by the rigid imine unit) turned out to be applica-
ble affording 1a with an ee of 78% (Table 2, entry 12).
These findings open room for discussion. The change from
CH2 to SO2 in bidentate sulfoximines L1 and L7 obviously en-
forces an arrangement of the latter N,N’-ligand around the
central metal atom, which leads to a completely unselective
chiral copper-sulfoximine complex. The extreme difference be-
tween the ee values, which were achieved by L1 (98% ee for
1a) and L7 (0% ee for 1a) with secondary amino and sulfon-
amido groups, respectively, disproves the initial assumption
that only an NH group was crucial for asymmetric induction.
This is also supported by the fact that sulfoximine chelators L6
(11% ee for 1a) and L8 (78% ee for 1a), which possess no hy-
drogen at the second nitrogen coordination site, were able to
accumulate one enantiomer of 1a. Actually, the NH substitu-
tion pattern in the ligand backbone does not seem to be es-
sential, but can be very supportive for an efficient asymmetric
induction, depending on its chemical environment, though. Se-
lectivity issues caused by neighboring groups like the CH2 (as
in L1) versus the bulkier SO2 (as in L7) indicate that steric fac-
tors probably contribute. A comparison of the results coming
from the NH (as in L1) versus the NMe (as in L6) substitution
pattern supports this hypothesis. From another perspective,
however, a proper electron density at the second nitrogen
donor could be the major responsible factor. This would ulti-
mately be reflected in the nitrogen’s basicity towards the
cupric Lewis acid. Indeed, sulfoximine ligands with donor sites
of medium basicity (an aldimine in L8, a secondary amine in
L1) deliver the product 1a with good or excellent ee values
whereas those with strong (a tertiary amine in L6) or very
weak basicity (a sulfonamide in L7) fail to give reasonable
enantiocontrol. Another test reaction with (R)-N-(2-aminophen-
yl)-S-methyl-S-phenylsulfoximine as ligand strengthens this as-
sumption. This anilino sulfoximine, which lacks of any addition-
al steric bulk at the primary amine and of which the basicity
should be closer to the ones of L1 and L8 than of L6 or L7 fur-
nished ent-1a with a moderate ee of 60%.
Next, applications of copper complexes with sulfoximine-
based ligands L2–L8, which contain altered steric or electronic
features, were tested. L2 produced a less selective and reactive
catalyst (Table 2, entry 6). Obviously, the higher steric demand
of the two tert-butyl substituents and a competing additional
coordination site for the copper ion which is offered by the
phenyl ring’s hydroxy function reduced both the stereocontrol
(94% de, 67% ee) and the yield (88%). The three amino sulfoxi-
mines L3–L5 (Table 2, entries 7–9) showed a better overall per-
formance, giving phosphonate 1a in good to almost quantita-
tive yields (87–99%) and with excellent de values (97–99%). Al-
though they were able to form the major enantiomer of 1a
with a respectable excess (94–97% ee), none of them was su-
perior to L1.
Before the substrate scope of the reaction was investigated,
another point was addressed. Bearing the generally high ste-
reoselectivity in mind, which the amino sulfoximines L1–L5
had exhibited under almost all conditions so far, we wondered
which features of the ligand framework were responsible for
this exceptional behavior. Compared to L1 neither a branched
alkyl chain next to sulfur in the “western part” (as in L4) nor
a less bulky or strongly electron-deficient phenyl ring in the
“eastern part” (as in L3 or L5) significantly affected the enantio-
selection (vide supra). Thus, we hypothesized that major con-
tributions would arise from the nitrogen atom of the amino
function,[36] and that modifying this moiety should affect its
ability to coordinate to the copper ion. Thus, sulfoximines L6–
L8 with variations of the advantageous core structure of L1
were applied in the catalysis, and their effects on the stereo-
chemical pathway towards product 1a were studied. Exchange
of the nitrogen’s hydrogen atom with a methyl group in L6
demonstrated that for very high asymmetric induction, a secon-
dary amine was better than a tertiary (Table 2, entry 10 vs. 1).
The enantioselectivity diminished when sulfoximine L6 was ap-
plied and 1a was isolated with a deteriorated ee of 11%. Inter-
Finally, the effect of temperature was briefly scrutinized and
the results are delineated in Table 2, entries 13 and 14. Appa-
rently, 08C was found to be the optimal temperature. Virtually
only one of the four possible stereoisomers of 1a was selec-
tively obtained in a yield of 99% and with both de and ee
above 99% (Table 2, entry 14). If the catalyst loading was re-
duced to 5 mol%, the same efficiency could no longer be
maintained at 08C (entry 15). Nevertheless, the general reactivi-
ty and stereoselectivity remained acceptable, and 1a was iso-
lated in 95% yield with a de of 99% and an ee of 97%.
Under the optimized conditions that comprised 10 mol% of
a catalyst stemming from Cu(OTf)2 and amino sulfoximine L1
with TFE as additive in diethyl ether at 08C (Table 2, entry 14),
we evaluated the scope of the transformation with respect to
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1694
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
than butenolide 1g (97%), which bears an electron-withdraw-
ing 4-chloro substituent. A more distinct effect concerning the
enantioselectivity was found for the position of the substitu-
ent. As shown by entries 5 and 6, methyl groups in 3- or 4-po-
sition at the aromatic system allowed the formation of highly
enantioenriched products 1e and f (each >99% ee). In con-
trast, the analogous 2-methyl-substituted 1d (entry 4) was pro-
duced less selectively and had an ee of 95%. Such a result can
be explained by steric interactions of the methyl group in this
position, which interfere with the chiral catalyst. This assump-
tion was confirmed when the scope was expanded to a-keto
phosphonates with fused aromatic rings (Table 3, entries 9 and
10). Reactants 2i and j that possess either a 1- or 2-naphthyl
rest adjacent to the carbonyl group reacted nearly equally well
in terms of yield (89 and 86%) and both gave the major diaste-
reomers of 1i and j in almost complete manner (each 99% de).
As expected, 1-naphthyl product 1i had a comparably low ee
of 90% while the 2-naphthyl derivative 1j was isolated with an
excellent ee of 99%. Indeed, these different enantioselectivities
in the formation of 1i and j are in full line with the previous
data of entries 4 and 5 (Table 3).
Table 3. Reactions of various a-keto phosphonates 2 with furan 3a.[a]
2
R1
R2
1
Yield
[%][b]
de
[%][c]
ee
[%][d]
1[e]
2
3
4
5
6
7
8
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
Et
Me
iPr
Et
Et
Et
Et
Et
Et
Et
Et
Et
Me
Et
iPr
Et
Et
Et
Et
Ph
Ph
Ph
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
99
50
99
83
90
93
96
93
89
86
81
83
95
98
81
90
89
88
62
>99
96
>99
>99
98
99
97
97
99
>99
>99
98
2-Me-C6H5
3-Me-C6H5
4-Me-C6H5
4-Cl-C6H5
4-MeO-C6H5
1-naphthyl
2-naphthyl
2-thienyl
2-furyl
Me
Me
Me
iPr
nBu
cyclohexyl
Bn
95
>99
>99
97
>99
90
9
10
99
93
93
99
94
97
99
98
99
11
12
13
14
15
16
17
18
19
>99
>99
>99
>99
>99
97 (>99)[f]
98
>99
97
>99
90
Just recently it has been revealed that a-heteroaryl
a-hydroxy phosphonates that possess a 2-furyl or 2-thienyl
unit serve as precursors for compounds with certain plant-re-
lated biological activities.[37] All compounds, however, were
prepared as racemates. In order to demonstrate the functional
group tolerance of the chiral copper-sulfoximine catalyst and
to establish a route to similar but enantioenriched heteroaro-
matic structures, thiophene- and furan-based a-keto phospho-
nates 2k and l were utilized (Table 3, entries 11 and 12). These
reactions proceeded quite flawlessly leading to the two corre-
sponding products in acceptable yields (81 and 83%, respec-
tively) and with an identical de of 93% for each. To our delight,
a high level of enantioselectivity could be maintained in both
experiments, providing butenolide 1k with an ee of 94% and
1l with an even better ee of 97%.
The recently ISO approved pesticide clacyfos,[38] derived
from a (1-hydroxyethyl)phosphonate ester, served as stimulus
to further elaborate the substrate scope with aliphatic a-keto
phosphonates. Like before, the role of a methyl, ethyl and iso-
propyl group R1 on the ester position of acetylphosphonates
2m–o was studied first (Table 3, entries 13–15). In contrast to
the analogous aromatic products 1b and 1c, the yield of the
dimethyl phosphonate 1m (95%) was much higher than the
one of diisopropyl phosphonate 1o (81%). Nonetheless, the
best yield (98%) could again be obtained in case of the diethyl
a-hydroxy phosphonate 1n. Irrespective of the three sterically
different alkoxy groups at phosphorus all stereoselectivities
proved outstanding (>99% de for products 1m–o, 98% ee for
1n and 99% ee for 1m and o, respectively). Based on the excel-
lent performance of the ethyl ester group in substrate 2n
(Table 3, entry 14), a few other aliphatic diethyl a-keto phos-
phonates 2 were consequently applied with furan 3a. In this
context we found that the ketonic alkyl part R2 could be pro-
longed from methyl to n-butyl as in 2q without hampering the
catalysis (Table 3, entry 17). Furthermore, it can also be sterical-
ly increased to isopropyl as in 2p or cyclohexyl as in 2r
s
s
[a] Reaction conditions: Cu(OTf)2 (10 mol%), L1 (10 mol%), 2 (0.20 mmol),
TFE (1.2 equiv), 3a (1.1 equiv), Et2O (2 mL), 08C, overnight. [b–d] Same as in
Table 1. [e] Identical with entry 14 of Table 2. [f] Before and after (in paren-
theses) column chromatography.
different a-keto phosphonates 2. The results are described in
Table 3. Firstly, the role of alkoxy substituents at phosphorus
was examined for the up-coming series of aromatic ketones.
Hence, two dialkyl benzoylphosphonates 2b (R1 =Me) and 2c
(R1 =iPr), which are of smaller or bigger size than 2a (R1 =Et)
were tried. Indeed, the ester moieties of the electrophiles con-
tributed significantly to the general success of the catalysis.
Compared to the excellent yields (each 99%) and de values
(each>99%) of the products 1a and c with bulkier ethyl and
isopropyl esters, respectively, dimethyl a-hydroxy phosphonate
1b was obtained only with a barely tolerable yield of 50%, but
a good de of 96% (entries 1–3). Interestingly, the enantioselec-
tive pathway for 1b was not negatively affected furnishing an
ee higher than 99%. Merely in case of diisopropyl benzoyl-
phosphonate (2c) as substrate the ee of the corresponding
product 1c decreased negligibly (98%).
Taking into account that an ethyl ester group like in com-
pound 2a worked best, the scope was extended to a compre-
hensive spectrum of aromatic diethyl phosphonates. Entries 4–
8 of Table 3 disclose that the developed catalytic system easily
accepted various phenyl-substituted substrates 2d–h to pro-
duce the VMA products 1d–h in good to high yields (83–96%)
with de values over 97%. The electronic nature of the substitu-
ents had only a subtle influence on the stereochemistry
(Table 3, entries 6–8). Nevertheless, products 1f and h with
electron-donating 4-methyl or 4-methoxy groups at the phenyl
ring were obtained with slightly better ee values (each >99%)
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1695
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(Table 3, entries 16 and 18, respectively). Albeit the reactivities
of the aliphatic electrophiles 2p–r were somewhat lower com-
pared to those of aromatic ones, they still afforded the corre-
sponding tertiary a-hydroxy phosphonates 1p–r with accepta-
ble yields in a close range of 88–90%. Although no significant
trend in diastereoselectivity seemed obvious (97–99% de for
1p–r), increasing the steric bulk at the ketonic alkyl rest im-
plied a better enantiocontrol during the catalysis (Table 3, en-
tries 16 and 18 vs. entry 17). This was ultimately demonstrated
by the salient ee values of over 99% for VMA products 1p
and r bearing isopropyl and cyclohexyl substituents, respec-
tively. However, the ee (97%) of a-hydroxy phosphonate 1q,
which contained a linear C4 alkyl chain, was reasonably high
as well. Unpleasantly, g-(hydroxyalkyl)butenolide 1s with
a benzyl substituent was isolated in a rather low yield of 62%
(Table 3, entry 19). Besides, the enantiofacial differentiation
during the asymmetric catalysis was less efficient as indicated
by an ee of 90%. The diastereoselectivity (98% de) was not af-
fected, though. Finally, it is worth mentioning that under no
circumstances, neither during the initial optimization process
nor while exploring the substrate scope, a/g-regioselectivity
issues have occurred with silyloxy furan 3a. All reactions were
completely g-selective and regioisomeric a-(hydroxyalkyl)bute-
nolides have never been observed.[39]
Table 4. Applicability and scope of substituted nucleophiles 3.[a]
2
3
4
Yield
[%][b]
de
[%][c]
ee
[%][d]
1
2
3
4
5
6[e]
a
a
a
a
n
n
b
c
d
e
e
f
a
b
c
d
e
f
86
88
70
77
61
83
99
>99
99
99
85
89
92
98
71
82
88
94 (>99)[f]
[a] Reaction conditions: 2 (0.20 mmol), TFE (1.2 equiv), 3 (1.1 equiv), Et2O
(2 mL). [b–d] Same as in Table 1. [e] Conducted at À208C. [f] Same as in
Table 3.
Studies in the field of a-keto phosphonates 2 are usually fo-
cused only on these electrophiles themselves. Besides, often
the reported scope is narrow and restricted to trivial or minor
variation patterns of 2 while the applicable spectrum of their
nucleophilic reaction partners is scarcely explored.[25] This is
particularly true for heterocyclic silyloxy dienes in catalytic
enantioselective aldol-type additions, which are mostly con-
fined to the simple and unsubstituted 2-(trimethylsilyloxy)furan
(3a).[27–29] By altering the steric or electronic demands of bute-
nolide precursors 3 problems of selectivity and reactivity may
arise, and an efficient catalyst must show its proficiency also
under such conditions. Table 4 shows that the chiral copper-
sulfoximine complex is largely able to deal with such issues.
As anticipated, the standard electrophile diethyl benzoyl-
phosphonate (2a) reacted differently in catalyses with the sily-
loxy furans 3b–d, which have a single methyl group as sub-
stituent R2 at each possible position (Table 4, entries 1–3).
These bulkier nucleophiles—compared to unsubstituted furan
3a from entry 1 of Table 3—led to slightly declined yields of
70–88% for the products 4a–c. While the diastereoselection re-
mained unchanged, the enantiocontrol was noticeably influ-
enced by the position of the methyl group in furan derivatives
3b–3d (from a to b to g, see Scheme of Table 4). The closer
the methyl substituent was located to the newly connected
g-carbon in this series, the higher was the ee value of products
4a–c (from 89 to 92 to 98%). The formation of two vicinal qua-
ternary stereogenic centers with high asymmetric induction is
still considered to be a tremendous challenge. Noteworthy, the
chiral catalyst we are reporting here successfully accomplished
this task yielding phosphorus butenolide 4c with an excellent
ee of 98%.
These seem to be of higher relevance than just simple steric
interactions. When benzoylphosphonate 2a reacted with sily-
loxy furan 3e, which bears the stronger electron-donating 4-
methoxy substituent instead of the 4-methyl group (as in 3c),
the yield of product 4d reached only 77%. Moreover, a pro-
nounced regression of the ee down to 71% was detected for
4-methoxy butenolide 4d (compared to the ee of 92% for 4-
methyl butenolide 4b). The entries 1–4 of Table 4 together
with entry 1 of Table 3 unequivocally emphasize how the nu-
cleophilicity at the g-carbon atom in silyloxy furans 3a–e affects
the enantioselectivity of the catalysis. This can be rationalized
as follows: The enantioselection improves in the same way as
silyloxy furans 3 become less nucleophilic towards phospho-
nate 2a. In turn, reduced nucleophilicity originates from
a lower electron density at the respective g-carbon of 3, which
is controlled by the nature and position of substituents R2.
Since the electron densities at the g-carbon atoms of 3 directly
correlate with their chemical shifts in nuclear magnetic reso-
nance (NMR) spectra, a tendency is visible. The more the NMR
signal of the g-carbon atom in the sequence from 3e to c
to a moved into the downfield region of the spectrum (indicat-
ing a lower nucleophilic behavior), the higher were the ee
values of corresponding VMA products.[40]
We also discovered that 4-methoxy silyloxy furan 3e is more
sensitive to its addition partners 2. As delineated in entry 5 of
Table 4, the VMA reaction of 3e with acetylphosphonate 2n
gave product 4e, but both yield (61%) and de (85%) were re-
duced. To our delight, the synthesis of aliphatic 4e proceeded
with higher enantioselectivity (82% ee) than the one for aro-
matic 4d.[41] Decreasing the temperature from 08C to À308C
was not helpful for the specific combination of substrates 2n
and 3e. In this case product 4e was obtained with the same
A comparison of entries 2 and 4 of Table 4 reveals how elec-
tronic factors in heterocycles 3 have to be taken into account.
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1696
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
yield and stereoselectivity as before (not shown in the Table).
A nitrogen nucleophile could also be applied, confirming the
functional group tolerance of the chiral catalyst (Table 4,
entry 6). When acetylphosphonate 2n reacted with N-methyl
protected silyloxy pyrrole 3f, g-(hydroxyalkyl)lactam 4f was
formed in reasonable yield (83%) and with good stereocontrol
(94% de and 88% ee).
Figure 3. Phosphonates 2t and u and silyl nucleophile 3g failed to react.
Analyzing the X-ray crystal structures of compounds 1a
and o allowed the unambiguous determination of their relative
and/or absolute configurations (Figure 2).[42] The stereochemi-
due to the steric hindrance of the ketonic tert-butyl group,
which prevented reactivity. With respect to heterocyclic pat-
terns and their presence in electrophiles or nucleophiles, some
unexpected reactivity issues were observed. The attempt to
utilize phosphonate 2u bearing the N-methyl pyrrole core was
unsuccessful, and no conversion was observed with silyloxy
furan 3a. This was surprising because in reactions with 3f as
nucleophile this heterocyclic substituent was unproblematic.
The pyrrole-2-carbonyl phosphonate 2u was fully recovered,
even when the reaction was conducted in diethyl ether at
room temperature or in 1,4-dioxane at 1008C. Interestingly,
2-(trimethylsilyloxy)thiophene (3g) did not act as nucleophile
with benzoylphosphonate 2a, although silyloxy thiophenes are
prone to undergo Mukaiyama-type transformations and the
thiophene moiety itself did not hamper the asymmetric cataly-
sis when it was featured in phosphonate 2k.[31e,45]
In order to demonstrate the robustness and reliability of the
catalytic system, an up-scaling experiment (from 0.2 mmol to
10 mmol) using acetylphosphonate 2n and silyloxy furan 3a to
yield 1n was performed (Scheme 2). Guided by the results pre-
sented in entry 15 of Table 2, the amount of Cu(OTf)2 and
Scheme 2. Scale-up of the synthesis of 1n with reduced catalyst loading.
Figure 2. Figure ORTEP plots[43] of the X-ray crystal structures of (S,R)-1a
(top) and rac,anti-1o (bottom). Ellipsoids are shown at 50% probability level.
amino sulfoximine L1 was reduced to 5 mol%. Under those
conditions, exactly the same stereoselectivities (99% de and
92% ee, respectively) were observed for both reactions.
Worthy of note, the yield of 1n was even higher on larger
scale (94 vs. 83%). Besides, sulfoximine ligand L1 could suc-
cessfully be recovered with a yield of 72% in (stereo-)chemical-
ly homogenous manner as confirmed by NMR spectroscopy
and analysis by high performance liquid chromatography on
chiral stationary phase. Comparing the result of the small scale
experiment with the one achieved by using 10 mol% of cata-
lyst (entry 14 of Table 3), a slightly lower yield and ee of 1n
was observed.
cally almost homogeneous product 1a (>99% de and>99%
ee) that was available by amino sulfoximine ligand (S)-L1 dis-
played an anti relationship. In addition, the absolute configura-
tion of anti-1a was established to be (S) for the quaternary ste-
reogenic center at C1 and (R) for the tertiary stereogenic
center at C2. For aliphatic product 1o suitable crystals could
be obtained within the context of its racemate synthesis (>
99% de) with ligand rac-L1. Here, the relative configuration
was confirmed to be anti again. Based on these data all other
configurations for products 1 were assigned in analogy.
Regarding the importance of the butyrolactone motif in nat-
urally occurring compounds,[20] enantioenriched 1n was trans-
formed into the g-(hydroxyalkyl)butanolide 5 (Scheme 3). Hy-
drogenation of the CÀC double bond succeeded with full con-
Finally, the limitations of the reaction scope shall be dis-
cussed (Figure 3).[44] Electrophile 2t failed to give its corre-
sponding product with silyloxy furan 3a. Presumably, this is
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1697
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
16.48 ppm (d, J=5.6 Hz); 31P NMR (243 MHz, CDCl3): d=
18.59 ppm; IR (ATR): n˜ =3195, 2980, 2924, 1754, 1603, 1490, 1448,
1397, 1299, 1224, 1159, 1102, 1011, 888, 827, 696; MS (EI): m/z (%):
327 ([M+H]+, 1), 308 (1), 243 (19), 151 (11), 121 (14), 111 (15), 105
(100), 93 (8), 83 (15), 77 (37), 65 (12), 55 (6), 51 (8); MS (CI): m/z (%):
355 ([M+C2H5]+, 2), 327 ([M+H]+, 63), 309 (14), 243 (97), 189 (21),
151 (17), 139 (100), 121 (22), 111 (45), 105 (44), 85 (56), 83 (13); ele-
mental analysis calcd (%) for C15H19O6P (326.28): C 55.22, H 5.87;
found: C 55.04, H 5.88; HPLC: tR =24.3 min [minor], 26.6 [major]
(AD-H column, flow rate 0.7 mLminÀ1, n-heptane/iPrOH=93:7, l=
210 nm, 208C); ee: >99%.
Scheme 3. Synthesis of butyrolactone 5 by hydrogenation of butenolide 1n.
version of 1n, and saturated lactone 5 could be isolated with-
out chromatographic work-up in 97% yield.
Acknowledgements
Conclusion
This work was supported by the Fonds der Chemischen Indus-
trie. We express our gratitude to Dr. Jun Wang (School of
Chemistry and Chemical Engineering, Sun Yat-Sen University,
China) and Dr. Christoph Rꢂuber (RWTH Aachen University) for
very stimulating and fruitful discussions.
A catalytic and highly stereoselective vinylogous Mukaiyama
aldol reaction with substantial variations of a-keto phospho-
nates 2 and heterocyclic dienol silanes 3 has been developed
to give products 1 of potential biomedical interest.[46,47] Note-
worthy, an additional cleavage step to remove the TMS group
under acidic conditions is not needed, and unprotected aldol
products 1 are directly obtained. The catalyst derived from
Cu(OTf)2 and amino sulfoximine L1[48] overcomes commonly
observed obstacles associated with such a process, namely the
lower reactivity of ketonic substrates plus the control of regio-,
diastereo- and enantioselectivity. Furthermore, possible key
factors in the structure of C1-symmetric sulfoximines are dis-
cussed being responsible for the high asymmetric induction.
Keywords: aldol reaction · asymmetric catalysis · butenolides ·
a-hydroxy phosphonates · sulfoximines
[1] M. Horiguchi, M. Kandatsu, Nature 1959, 184, 901–902.
[2] For reviews see: a) R. L. Hilderbrand, T. O. Henderson in The Role of
Phosphonates in Living Systems (Ed.: R. L. Hilderbrand), CRC, Boca Raton,
1983, pp. 5–30; b) M. Horiguchi in Biochemistry of Natural CÀP Com-
pounds (Eds.: T. Hori, M. Horiguchi, A. Hayashi), Japanese Association for
Research on the Biochemistry of CÀP Compounds, Shiga, 1984, pp. 24–
52.
Experimental Section
[3] For an overview see: W. W. Metcalf, W. A. van der Donk, Annu. Rev. Bio-
chem. 2009, 78, 65–94.
General procedure for the Cu-catalyzed VMA reaction
[4] For a review see: S. C. Fields, Tetrahedron 1999, 55, 12237–12273.
[5] For selected examples see: a) E. D. Korn, D. G. Dearborn, H. M. Fales,
E. A. Sokoloski, J. Biol. Chem. 1973, 248, 2257–2259; b) T. Ogita, S.
Gunji, Y. Fukazawa, A. Terahara, T. Kinoshita, H. Nagaki, T. Beppu, Tetra-
hedron Lett. 1983, 24, 2283–2286; c) Y. Watanabe, M. Nakajima, T. Hosh-
ino, K. Jayasimhulu, E. E. Brooks, E. S. Kaneshiro, Lipids 2001, 36, 513–
519.
A Schlenk tube under argon atmosphere was charged with
Cu(OTf)2 (7.4 mg, 0.020 mmol) and amino sulfoximine L1 (9.3 mg,
0.020 mmol). Et2O (2.0 mL) was added, and the green solution was
stirred at room temperature for 30 min. After cooling to 08C the
corresponding a-keto phosphonate 2 (0.20 mmol), 2,2,2-trifluoro-
ethanol (17 mL, 0.24 mmol) and heterocyclic dienol silane
3
´
´
[6] A. Szymanska, M. Szymczak, J. Boryski, J. Stawinski, A. Kraszewski, G.
Collu, G. Sanna, G. Giliberti, R. Loddo, P. La Colla, Bioorg. Med. Chem.
2006, 14, 1924–1934.
(0.22 mmol) were added in the order given. The Schlenk tube was
sealed with a rubber septum, and the reaction mixture was stirred
until the electrophile was consumed as determined by TLC. The
cold reaction mixture was directly subjected to flash column chro-
matography to give products 1.
[7] a) N. Zuo, H.-W. He, Phosphorus Sulfur Silicon Relat. Elem. 2008, 183,
621–622; b) H. Song, H. Mao, D. Shi, Chin. J. Chem. 2010, 28, 2020–
2024.
[8] K. R. M. Naidu, K. S. Kumar, P. Arulselvan, C. B. Reddy, O. Lasekan, Arch.
Pharm. Chem. Life Sci. 2012, 345, 957–963.
[9] For a recent review see: W. M. Abdou, A. A. Shaddy, ARKIVOC 2009, ix,
143–182.
Synthesis of a-hydroxy phosphonate 1a
Prepared according to the general procedure from diethyl benzoyl-
phosphonate (2a) and 2-(trimethylsilyloxy)furan (3a). The product
was purified by flash column chromatography two times (first:
n-pentane/acetone=2:1; second: EtOAc) to give product 1a as
a white solid. Yield: 99%; [a]=69.2 (c=1.0 in CHCl3); m.p. 109–
111 8C (racemate: 115–1168C); de: >99%; 1H NMR (600 MHz,
CDCl3): d=7.65 (dd, J=5.8 Hz, 0.9 Hz, 1H), 7.62–7.59 (m, 2H), 7.36–
7.33 (m, 2H), 7.31–7.27 (m, 1H), 6.05 (dd, J=5.8 Hz, 1.7 Hz, 1H),
5.64 (dt, J=5.3 Hz, 1.6 Hz, 1H), 4.33 (s br, 1H), 4.14–4.03 (m, 3H),
3.95 (ddq, J=14.3 Hz, 10.1 Hz, 7.1 Hz, 1H), 1.24 (t, J=7.1 Hz, 3H),
1.23 ppm (t, J=7.1 Hz, 3H); 13C NMR (151 MHz, CDCl3): d=172.80,
153.77 (d, J=4.1 Hz), 135.81, 128.43, 128.40 (d, J=2.1 Hz), 126.26
(d, J=4.3 Hz), 123.28, 85.39 (d, J=10.8 Hz), 76.64 (d, J=160.9 Hz),
64.51 (d, J=7.6 Hz), 64.44 (d, J=7.6 Hz), 16.49 (d, J=5.6 Hz),
[10] a) C. J. Dunn, A. Fitton, E. M. Sorkin, Drugs Aging 1994, 5, 446–474; b) I.
Shinkai, Y. Ohta, Bioorg. Med. Chem. 1996, 4, 3–4; c) L. Widler, K. A.
Jaeggi, J. R. Green, Phosphorus Sulfur Silicon Relat. Elem. 1999, 144, 5–8;
d) C. J. Dunn, K. L. Goa, Drugs 2001, 61, 685–712.
[11] a) D. V. Patel, K. Rielly-Gauvin, D. E. Ryono, Tetrahedron Lett. 1990, 31,
5587–5590; b) J. F. Dellaria, Jr., R. G. Maki, H. H. Stein, J. Cohen, D. Whit-
tern, K. Marsh, D. J. Hoffman, J. J. Plattner, T. J. Perun, J. Med. Chem.
1990, 33, 534–542; c) D. V. Patel, K. Rielly-Gauvin, D. E. Ryono, C. A.
Free, W. L. Rogers, S. A. Smith, J. M. DeForrest, R. S. Oehl, E. W. Petril-
lo, Jr., J. Med. Chem. 1995, 38, 4557–4569; d) M. Tao, R. Bihovsky, G. J.
Wells, J. P. Mallamo, J. Med. Chem. 1998, 41, 3912–3916.
[12] R. F. Frechette, C. Ackerman, S. Beers, R. Look, J. Moore, Bioorg. Med.
Chem. Lett. 1997, 7, 2169–2172.
[13] T. R. Burke Jr., Z.-H. Li, J. B. Bolen, V. E. Marquez, J. Med. Chem. 1991, 34,
1577–1581.
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1698
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[14] a) D. L. Pompliano, E. Rands, M. D. Schaber, S. D. Mosser, N. J. Anthony,
J. B. Gibbs, Biochemistry 1992, 31, 3800–3807; b) R. J. Hohl, K. A. Lewis,
D. M. Cermak, D. F. Wiemer, Lipids 1998, 33, 39–46.
[15] a) A. M. MacLeod, R. Baker, M. Hudson, K. James, M. B. Roe, M. Knowles,
G. MacAllister, Med. Chem. Res. 1992, 2, 96–101; b) A. J. Ganzhorn, J.
Hoflack, P. D. Pelton, F. Strasser, M.-C. Chanal, S. R. Piettre, Bioorg. Med.
Chem. 1998, 6, 1865–1874.
[16] J. A. Sikorski, M. J. Miller, D. S. Braccolino, D. G. Cleary, S. D. Corey, J. L.
Font, K. J. Gruys, C. Y. Han, K. C. Lin, P. D. Pansegrau, J. E. Ream, D.
Schnur, A. Shah, M. C. Walker, Phosphorus Sulfur Silicon Relat. Elem.
1993, 76, 115–118.
[17] L. C. Pettigrew, F. Bieber, J. Lettieri, D. P. Wermeling, F. A. Schmitt, A. J.
Tikhtman, J. W. Ashford, C. D. Smith, D. R. Wekstein, W. R. Markesbery, J.
Orazem, B. B. Ruzicka, J. Mas, B. Gulanski, J. Clin. Pharmacol. 1998, 38,
236–245.
Korean Chem. Soc. 2011, 32, 785–786; k) S. Kong, W. Fan, G. Wu, Z.
Miao, Angew. Chem. 2012, 124, 8994–8997; Angew. Chem. Int. Ed. 2012,
51, 8864–8867; l) C. Wang, C. Xu, X. Tan, H. Peng, H. He, Org. Biomol.
Chem. 2012, 10, 1680–1685; m) M. Hatano, T. Horibe, K. Ishihara,
Angew. Chem. 2013, 125, 4647–4651; Angew. Chem. Int. Ed. 2013, 52,
4549–4553.
[26] For a summary on a-keto phosphonates 2 see: C. E. McKenna, B. A. Ka-
shemirov, in Topics in Current Chemistry, Vol. 220 (Ed.: J.-P. Majoral),
Springer, Berlin, 2002, pp. 201–238.
[27] For selected reviews concerning silyloxy furan 3a see: a) S. E. Denmark,
J. R. Heemstra, Jr., G. L. Beutner, Angew. Chem. 2005, 117, 4760–4777;
Angew. Chem. Int. Ed. 2005, 44, 4682–4698; b) G. Casiraghi, F. Zanardi,
L. Battistini, G. Rassu, Synlett 2009, 1525–1542; c) G. Casiraghi, L. Battis-
tini, C. Curti, G. Rassu, F. Zanardi, Chem. Rev. 2011, 111, 3076–3154;
d) S. V. Pansare, E. K. Paul, Chem. Eur. J. 2011, 17, 8770–8779; e) V. Bisai,
Synthesis 2012, 44, 1453–1463; f) Z. Miao, F. Chen, Synthesis 2012, 44,
2506–2514.
[18] For overviews see: a) R. Engel, Chem. Rev. 1977, 77, 349–367; b) L.
Azema, R. Baron, S. Ladame, Curr. Enzym. Inhib. 2006, 2, 61–72.
[19] For a recent summary see: S. Sobhani, Z. Tashrifi, Tetrahedron 2010, 66,
1429–1439.
[28] For a review see: S. Adachi, T. Harada, Eur. J. Org. Chem. 2009, 3661–
3671.
[20] For reviews see: a) S. S. C. Koch, A. R. Chamberlin in Studies in Natural
Products Chemistry, Vol. 16, Part J (Ed.: Atta-ur-Rahman), Elsevier, Amster-
dam, 1995, pp. 687–725; b) M. Seitz, O. Reiser, Curr. Opin. Chem. Biol.
2005, 9, 285–292; c) T. M. Ugurchieva, V. V. Veselovsky, Russ. Chem. Rev.
2009, 78, 337–373.
[21] For recent examples see: a) R.-T. Li, Q.-B. Han, Y.-T. Zheng, R.-R. Wang, L.-
M. Yang, Y. Lu, S.-Q. Sang, Q.-T. Zheng, Q.-S. Zhao, H.-D. Sun, Chem.
Commun. 2005, 2936–2938; b) Y.-W. Hsu, L.-C. Hsu, Y.-H. Liang, Y.-H.
Kuo, T.-M. Pan, J. Agric. Food Chem. 2011, 59, 4512–4518; c) R. Chib,
B. A. Shah, N. Anand, A. Pandey, K. Kapoor, S. Bani, V. K. Gupta, Rajni-
kant, V. K. Sethi, S. C. Taneja, Bioorg. Med. Chem. Lett. 2011, 21, 4847–
4851; d) R. M. Centko, S. Ramꢃn-Garcꢄa, T. Taylor, B. O. Patrick, C. J.
Thompson, V. P. Miao, R. J. Andersen, J. Nat. Prod. 2012, 75, 2178–2182;
e) L. Hua, W.-Y. Qi, S. H. Hussain, K. Gao, M. Arfan, Steroids 2012, 77,
811–818.
[22] a) J. G. Allen, F. R. Atherton, M. J. Hall, C. H. Hassall, S. W. Holmes, R. W.
Lambert, L. J. Nisbet, P. S. Ringrose, Nature 1978, 272, 56–58; b) F. R.
Atherton, M. J. Hall, C. H. Hassall, R. W. Lambert, P. S. Ringrose, Antimi-
crob. Agents Chemother. 1979, 15, 677–683; c) T. Kametani, K. Kigasawa,
M. Hiiragi, K. Wakisaka, S. Haga, H. Sugi, K. Tanigawa, Y. Suzuki, K.
Fukawa, O. Irino, O. Saita, S. Yamabe, Heterocycles 1981, 16, 1205–1242.
[23] For reviews, see: a) D. F. Wiemer, Tetrahedron 1997, 53, 16609–16644;
b) S. Failla, P. Finocchiaro, G. A. Consiglio, Heteroat. Chem. 2000, 11,
493–504; c) H. Grçger, B. Hammer, Chem. Eur. J. 2000, 6, 943–948; d) P.
Kafarski, B. Lejczak, J. Mol. Catal. B 2004, 29, 99–104; e) O. I. Kolodiazh-
nyi, Tetrahedron: Asymmetry 2005, 16, 3295–3340; f) O. I. Kolodiazhnyi,
Russ. Chem. Rev. 2006, 75, 227–253; g) Ł. Albrecht, A. Albrecht, H.
Krawczyk, K. A. Jørgensen, Chem. Eur. J. 2010, 16, 28–48; h) P. Merino,
E. Marquꢀs-Lꢃpez, R. P. Herrera, Adv. Synth. Catal. 2008, 350, 1195–
1208; i) O. I. Kolodiazhnyi, Russ. Chem. Rev. 2011, 80, 883–910; j) V. Bisai,
A. Bisai, V. K. Singh, Tetrahedron 2012, 68, 4541–4580.
[24] For recent examples see: a) N. S. Goulioukina, G. N. Bondarenko, A. V.
Bogdanov, K. N. Gavrilov, I. P. Beletskaya, Eur. J. Org. Chem. 2009, 510–
515; b) O. O. Kolodyazhnaya, O. I. Kolodyazhnyi, Russ. J. Gen. Chem.
2010, 80, 1718–1719; c) D. Uraguchi, T. Ito, T. Ooi, J. Am. Chem. Soc.
2009, 131, 3836–3837; d) K. Suyama, Y. Sakai, K. Matsumoto, B. Saito, T.
Katsuki, Angew. Chem. 2010, 122, 809–811; Angew. Chem. Int. Ed. 2010,
49, 797–799; e) W. Chen, Y. Hui, X. Zhou, J. Jiang, Y. Cai, X. Liu, L. Lin, X.
Feng, Tetrahedron Lett. 2010, 51, 4175–4178.
[29] J. Yu, X. Zhao, Z. Miao, R. Chen, Org. Biomol. Chem. 2011, 9, 6721–6726.
[30] In the midst of our research, Miao and co-workers reported a copper-
catalyzed enantioselective approach with a bis(oxazoline) ligand: G.
Hou, J. Yu, C. Yu, G. Wu, Z. Miao, Adv. Synth. Catal. 2013, 355, 589–593.
[31] a) M. Langner, C. Bolm, Angew. Chem. 2004, 116, 6110–6113; Angew.
Chem. Int. Ed. 2004, 43, 5984–5987; b) M. Langner, P. Rꢀmy, C. Bolm,
Chem. Eur. J. 2005, 11, 6254–6265; c) P. Rꢀmy, M. Langner, C. Bolm, Org.
Lett. 2006, 8, 1209–1211; d) J. Sedelmeier, T. Hammerer, C. Bolm, Org.
Lett. 2008, 10, 917–920; e) M. Frings, I. Atodiresei, J. Runsink, G. Raabe,
C. Bolm, Chem. Eur. J. 2009, 15, 1566–1569; f) M. Frings, D. Goedert, C.
Bolm, Chem. Commun. 2010, 46, 5497–5499; g) M. Frings, I. Atodiresei,
Y. Wang, J. Runsink, G. Raabe, C. Bolm, Chem. Eur. J. 2010, 16, 4577–
4587.
[32] For reviews on sulfoximine ligands in asymmetric metal catalysis see:
a) H. Okamura, C. Bolm, Chem. Lett. 2004, 33, 482–487; b) C. Bolm in
Asymmetric Synthesis with Chemical and Biological Methods (Eds.: D.
Enders, K.-E. Jaeger), Wiley-VCH, Weinheim, 2007, pp. 149–175; c) H.
Pellissier, Tetrahedron 2007, 63, 1297–1330; d) C. Worch, A. C. Mayer, C.
Bolm in Organosulfur Chemistry in Asymmetric Synthesis (Eds.: T. Toru, C.
Bolm), Wiley-VCH, Weinheim, 2008, pp. 209–229.
[33] Although the use of toluene and THF led to good stereoselectivities,
the total value of these syntheses suffered from rather moderate yields
of the g-(hydroxyalkyl)lactone 1a (Table 1, entries 2 and 3). Therefore,
we also investigated the impact of TFE in these solvents and found
that it significantly increased the yields of 1a up to 87%. Unfortunately,
the stereoselectivities were not improved (96% de and 88% ee with tol-
uene, 99% de and 93% ee with THF).
[34] For computational studies of a related system see: R. Gotoh, M. Yama-
naka, Molecules 2012, 17, 9010–9022.
[35] For a review see: H. Yamamoto, K. Futatsugi, Angew. Chem. 2005, 117,
1958–1977; Angew. Chem. Int. Ed. 2005, 44, 1924–1942.
[36] For a review see: B. Zhao, Z. Han, K. Ding, Angew. Chem. 2013, 125,
4844–4889; Angew. Chem. Int. Ed. 2013, 52, 4744–4788.
[37] a) T. Wang, D. Y. Lei, Y. Huang, L. H. Ao, Phosphorus Sulfur Silicon Relat.
Elem. 2009, 184, 2777–2785; b) H.-W. He, H. Peng, T. Wang, C. Wang, J.-
L. Yuan, T. Chen, J. He, X. Tan, J. Agric. Food Chem. 2013, 61, 2479–
2488.
[38] a) H.-W. He, J.-L. Yuan, H. Peng, T. Chen, P. Shen, S.-Q. Wan, Y. Li, H.-L.
Tan, Y.-H. He, J.-B. He, Y. Li, J. Agric. Food Chem. 2011, 59, 4801–4813;
b) AGROW 2012, 640, 20.
[39] For an a-selective asymmetric metal catalysis with silyloxy furan 3a and
aldehydes see: M. Woyciechowska, G. Forcher, S. Buda, J. Mlynarski,
Chem. Commun. 2012, 48, 11029–11031.
[40] Enumerated in increasing order, we measured the following chemical
shifts for the g-carbon atoms of silyloxy furans 3a–e in CDCl3 (in paren-
theses are given the ee values of the corresponding catalysis products
4 from phosphonate 2a): 3e: d=112.8 ppm (71% ee for 4d), 3c: d=
128.9 ppm (92% ee for 4b), 3b: d=131.6 ppm (89% ee for 4a), 3a: d=
132.6 ppm (>99% ee for 1a), 3d: d=141.7 ppm (98% ee for 4c). For
analogous findings with linear silyloxy dienes in VMA reactions see:
S. E. Denmark, J. R. Heemstra, Jr., J. Org. Chem. 2007, 72, 5668–5688.
[25] a) S. Samanta, C.-G. Zhao, J. Am. Chem. Soc. 2006, 128, 7442–7443; b) T.
Mandal, S. Samanta, C.-G. Zhao, Org. Lett. 2007, 9, 943–945; c) J.
Huang, J. Wang, X. Chen, Y. Wen, X. Liu, X. Feng, Adv. Synth. Catal.
2008, 350, 287–294; d) X. Chen, J. Wang, Y. Zhu, D. Shang, B. Gao, X.
Liu, X. Feng, Z. Su, C. Hu, Chem. Eur. J. 2008, 14, 10896–10899; e) J. Liu,
Z. Yang, Z. Wang, F. Wang, X. Chen, X. Liu, X. Feng, Z. Su, C. Hu, J. Am.
Chem. Soc. 2008, 130, 5654–5655; f) V. B. Gondi, K. Hagihara, V. H.
Rawal, Angew. Chem. 2009, 121, 790–793; Angew. Chem. Int. Ed. 2009,
48, 776–779; g) J. Jiang, X. Chen, J. Wang, Y. Hui, X. Liu, L. Lin, X. Feng,
Org. Biomol. Chem. 2009, 7, 4355–4357; h) F. Wang, X. Liu, X. Cui, Y.
Xiong, X. Zhou, X. Feng, Chem. Eur. J. 2009, 15, 589–592; i) S. Perera,
V. K. Naganaboina, L. Wang, B. Zhang, Q. Guo, L. Rout, C.-G. Zhao, Adv.
Synth. Catal. 2011, 353, 1729–1734; j) H. J. Lee, J. H. Kim, D. Y. Kim, Bull.
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1699
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[41] The major stereoisomer of syn-4e was formed with a low ee of 23%.
See the Supporting Information.
Mor, N. Culeddu, G. Casiraghi, J. Org. Chem. 1998, 63, 1368–1369; c) H.
Yanai, A. Takahashi, T. Taguchi, Chem. Commun. 2010, 46, 8728–8730.
[46] Tests on biological activity are carried out currently.
[47] Compared to the reactions conditions of Miao and co-workers in
Ref. 30, the catalyst reported here gives higher yields, is more stereose-
lective and shows better reactivity for aliphatic phosphonates. Addi-
tionally, the reactions proceed at much milder temperature and less
amount of the nucleophile is required.
[42] CCDC 950558 and CCDC 950559 contain the crystallographic data for
(S,R)-1a and rac,anti-1o, respectively. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. Regarding the publication of
Miao and co-workers in ref. 30, we must note that several discrepancies
were found in the stereochemical representation of their crystal struc-
ture and the chemical drawings of their catalysis products.
[48] The (R)- and (S)-enantiomer of amino sulfoximine L1 are commercially
available, hence allowing easy access to both product enantiomers.
[43] A. L. Spek, Acta Crystallogr. D 2009, 65, 148–155.
[44] For recent comments on upholding scientific standards see: a) W. F. van
Gunsteren, Angew. Chem. 2013, 125, 128–132; Angew. Chem. Int. Ed.
2013, 52, 118–122; b) R. Noyori, J. P. Richmond, Adv. Synth. Catal. 2013,
355, 3–8.
Received: November 5, 2013
[45] a) G. Rassu, P. Spanu, L. Pinna, F. Zanardi, G. Casiraghi, Tetrahedron Lett.
1995, 36, 1941–1944; b) F. Zanardi, L. Battistini, G. Rassu, L. Pinna, M.
Published online on January 8, 2014
Chem. Eur. J. 2014, 20, 1691 – 1700
www.chemeurj.org
1700
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim