Practical Method for the Rhodium-Catalyzed Addition of Aryl- and Alkenylboronic Acids to Aldehydes

Article abstract of DOI:10.1002/1615-4169(20010430)343:4<343::AID-ADSC343>3.0.CO;2-Z

The arylation or alkenylation of aldehydes will boronic acids is conveniently effected by a catalyst system comprising RhCl₃ · 3 H₂O (1 mol %), the sterically hindered imidazolium salt 2 (1 mol %), and a base. The N-heterocyclic carbene 6 derived from 2 is believed to be the actual ligand to the catalytically active rhodium species formed in situ. The method is compatible with various functional groups in both reaction partners and follows a non-chelation controlled pathway in additions to the Garner aldehyde 25.

Full text of DOI:10.1002/1615-4169(20010430)343:4<343::AID-ADSC343>3.0.CO;2-Z

FULL PAPERS
Practical Method for the Rhodium-Catalyzed Addition of
Aryl- and Alkenylboronic Acids to Aldehydes
Alois FuÈ rstner,* Helga Krause
Max-Planck-Institut fuÈ r Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 MuÈ lheim/Ruhr, Germany
Fax: (+49) 208-306-2994, e-mail: fuerstner@mpi-muelheim.mpg.de
Received: March 9, 2001; Accepted March 16, 2001
Keywords:
aldehy-
des; arylation; boron;
carbene ligands; imi-
dazolium salts; rho-
dium
Abstract: The arylation or alkenylation of aldehydes method is compatible with
with boronic acids is conveniently effected by a cat- various functional groups
alyst system comprising RhCl3 ´3 H2O (1 mol %), the in both reaction partners
sterically hindered imidazolium salt 2 (1 mol %), and follows a non-chela-
and a base. The N-heterocyclic carbene 6 derived tion controlled pathway in
from 2 is believed to be the actual ligand to the cata- additions to the Garner aldehyde 23.
lytically active rhodium species formed in situ. The
phosphines, preferentially P(t-Bu)3 or PCy3, lead to
Introduction
better results.[3]
This finding has attracted our attention because of
our excellent experience with N-heterocyclic carbenes
(NHC's) as superior surrogates for the air-sensitive
and rather expensive PCy3. NHC's constitute a class of
ligands that exhibits pronounced s-donor but very
poor p-acceptor properties.[7] They are easily gener-
ated from the corresponding imidazolium salts and
have been employed with considerable success in var-
ious catalytic transformations involving the turn-over
of an electron-rich transition metal template. In parti-
cular, this refers to olefin metathesis[8±10] and palla-
dium-catalyzed cross-coupling reactions.[11,12] De-
scribed below is the successful extension of this
concept to the rhodium-catalyzed addition of organo-
boron compounds to aldehydes which is significantly
upgraded in practical terms if carried out in the pre-
sence of the sterically hindered imidazolium salt 2.
Although the addition of carbon nucleophiles to alde-
hydes is usually a facile process, limits are encoun-
tered where functionalized organometallic reagents
are required. Since organomagnesium and organo-
lithium derivatives, which are most frequently used
for this purpose, tolerate only few electrophilic
groups on themselves,[1] there remains ample room
for further methodological advancements. In spite of
the progress that has been achieved by using functio-
nalized organozinc, copper, chromium, tin, manga-
nese, and related organometallic species,[1] recent
publications describing the addition of arylboronic
acid derivatives to aldehydes in the presence of cata-
lytic amounts of Rh(I) and phosphine additives de-
serve particular mention.[2±6] These methods com-
bine a high efficiency with a reasonable tolerance
towards polar substituents in the substrates and ben-
efit from the stability and ready accessibility of the re-
quired boron derivatives. It is believed that the reac-
tion involves a transmetallation of the boronic acid
with formation of an organorhodium(I) species
which is nucleophilic enough to transfer its aryl sub-
stituent to an aldehyde. Originally, Rh(acac)(coe)2
(coe = cyclooctene) or Rh(acac)(CO)2 in combination
with bidentate phosphine ligands such as dppb [1,4-
bis(diphenylphosphino)butane] or dppf [1,1'-bis(di-
phenylphosphino)ferrocene] have been recom-
mended for the in situ preparation of the yet elusive
catalyst.[2,4] Later on, however, it was noticed that
sterically hindered and strongly basic monodentate
Results and Discussion
The data summarized in Table 1 show the efficiency
of Rh(acac)(coe)2 (3 mol %)[3] in combination with
different ligands or ligand precursors (3 mol % each)
in catalyzing the addition of phenylboronic acid to p-
methoxybenzaldehyde in basic medium (Scheme 1).
As can be seen, the imidazolium chlorides 1 and 2
bearing bulky aromatic substituents on their N-atoms
gave excellent results, whereas the corresponding
imidazolium salts 3 and 4 containing N-alkyl groups
turned out to be less efficient.[13] Imidazolium salts
Adv. Synth. Catal. 2001, 343, No. 4
Ó WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2001
1615-4150/01/34301-343±350 $ 17.50-.50/0
343

asc.wiley-vch.de
are known to generate NHC's in the presence of non-
nucleophilic bases; therefore, a control experiment
using free carbene 6 as the additive has been carried
out (entry 8). The reaction took the expected course
affording alcohol 7 in 55% yield. Although this result
suggests that carbenes are in fact involved as the ac-
tual donor ligands to the rhodium center in all of
these reactions, it is particularly fortunate from the
preparative point of view that the in situ release of this
ligand from the imidazolium salt 2 and base followed
by its interception by the admixed rhodium pre-cata-
lyst leads to significantly better results than the use of
isolated 6 (cf. entries 4/8). Furthermore, it was found
that compound 2 delivering an ``unsaturated'' car-
bene gives better results than the use of the corre-
sponding dihydroimidazolium salt 5 that furnishes
an even more basic ``saturated'' NHC (entry 7).
Scheme 1.
A survey of different bases showed that K2CO3 (15%,
4 h) and Et3N (11%, 8 h) are inappropriate, likely be-
cause they are unable to generate the NHC in situ by
deprotonation of 2, whereas all stronger bases such as
aqueous NaOH (90%, 6 h), NaOMe (77%, 1.5 h), KOt-
Bu (83%, 1 h), Cs2CO3 (50%, 3.5 h) and even TBAF
(67%, 14 h) gave product 7 in reasonable to excellent
yields.
Next, various rhodium and other transition metal
complexes have been screened for catalytic activity
(Table 2). We were pleased to learn that the replace-
[3]
ment of Rh(acac)(coe)2 ± a compound that is not
commercially available ± by simple, robust and fully
air-stable salts such as [Rh(OAc)2]2 or RhCl3 ´3 H2O
increases the overall efficiency. Use of the latter re-
sults in the shortest reaction time and the highest
yield even if the catalyst loading is reduced to
1 mol % (entry 4). This screening, however, has also
revealed that rhodium is rather unique in catalyzing
the addition reaction. None of the other late transition
metal salts that have been tested showed any appreci-
able activity. This includes different cobalt salts as
well as IrCl3 ´n H2O, i.e. the elements located above
and below rhodium in the periodic table (entry 5,
footnote b).
Table 1. Screening of different additives in the rhodium-
catalyzed addition of phenylboronic acid to p-methoxybenz-
aldehyde depicted in Scheme 1.
The results summarized above lead to an optimized
and very user-friendly procedure for the addition of
arylboronic acids to aldehydes which is distinguished
by its high efficiency, a low catalyst loading, short re-
action times, mild conditions (50±80 °C), the use of in-
expensive bases in aqueous solvents, and the fact that
only fully air-stable and well accessible ingredients
Table 2. Screening of the activity of different metal salts
(3 mol % each, unless stated otherwise) in combination with
imidazolium chloride 2 (3 mol %) in catalyzing the addition
of phenylboronic acid to p-methoxybenzaldehyde. All reac-
tions were carried out in aqueous DME at 80 °C in the pre-
sence of NaOMe (2 equivalents).
Entry
Metal Salt
Time (h)
Yield (%)
1
2
3
4
5
Rh(acac)(coe)2
[RhCl(cod)]2
[Rh(OAc)2]2
1.5
6.5
0.5
0.2
77
< 10
79
RhCl3 ´3 H2O
93[a]
< 5
[b]
±
4±21
[a] Using only 1 mol % of the rhodium salt and of the imida-
zolium chloride.
[b] The following metal salts showed no appreciable catalytic
activity: Pd(OAc)2, PtCl2, CoCl2 ´6 H2O, Co(acac)2, RuCl3 ´n H2O,
IrCl3 ´n H2O.
344
Adv. Synth. Catal. 2001, 343, 343±350

FULL PAPERS
Table 3. Rhodium-catalyzed addition of boronic acids to aldehydes. All reactions were carried out using RhCl3 ´3 H2O
(1 mol %), imidazolium chloride 2 (1 mol %), and NaOMe (1±2 equivalents) in aqueous DME at 80 °C unless stated other-
wise.
Adv. Synth. Catal. 2001, 343, 343±350
345

asc.wiley-vch.de
Table 3. (Continued)
are required. The results compiled in Table 3 reveal these rhodium-catalyzed transformations (Table 4).
the wide scope of this method which is compatible Gratifyingly, a very high selectivity for the anti-con-
with acetal, amide, urethane, ether, trifluoromethyl, figurated product 24 was noted (entry 1), which likely
unprotected hydroxy as well as bromide functions in originates from a non-chelation controlled pathway
both reaction partners. Importantly, entries 5 and 11 caused by the low affinity of the electron-rich rho-
show that the addition is highly chemoselective for al- dium center to the ``hard'' donor sites in 23. Alcohol
dehydes, while keto groups in either component re- 24 and derivatives thereof have recently attracted at-
main unaffected. Electron-rich and electron-poor
aromatic aldehydes react with similar ease, and ali-
phatic ones are also suitable even if they are sterically
hindered (entries 4, 16). Moreover, the reaction is not
limited to arylboronic acids but can also be applied to
alkenylboronic acids as nucleophiles which are read-
ily accessible by hydroboration of alkynes with cate-
cholborane and hydrolysis of the resulting boronates
(entries 14±16).[14,15] Limitations, however, were en-
countered with arylboronic acid derivatives bearing
strongly electron-withdrawing groups. Thus, at-
tempted addition of p-nitro- or p-cyanophenylboronic
acid to benzaldehyde by means of the present catalyst
system has been unsuccessful. This failure is ascribed
to the reduced nucleophilicity of these compounds
which makes the transmetalation to the rhodium cat-
alyst unfavorable.
Table 4. Rhodium-catalyzed addition of boronic acids to the
Garner aldehyde 23.
R
Yield (%)
anti : syn
Ph
71[a]
78[b]
97 : 3
CH3(CH2)5CH=CH-
82 : 18
[a]
Using 3 mol % each of RhCl3 ´3 H2O and imidazolium
chloride 2 at 80 °C.
Addition reactions to the Garner aldehyde 23[16]
provide information on the stereochemical course of
[b]
Using 5 mol % each of RhCl3 ´3 H2O and imidazolium
chloride 2 at 55 °C.
346
Adv. Synth. Catal. 2001, 343, 343±350

FULL PAPERS
tention as sphingosine mimics[17] as well as important ceptor carbonyl oxygen O3 only takes part in a much
building blocks for drug development programs aim- weaker C±H. . .O interaction.[21,22] However, in this
ing at the design of selective inhibitors of protein ki- case there is a competing C±H. . .p interaction invol-
nase C isozymes.[18] It is worth mentioning that the ving the phenyl ring. In the crystal the phenyl rings
selectivity for the desired anti-configurated isomer between neighboring molecules form a dihedral an-
of 24 is significantly higher than that observed in the gle of 77.2°. The shortest distance between a hydro-
[23]
Ê
addition of phenylmagnesium bromide to aldehyde gen atom and the phenyl ring plane is 2.70 A.
Ad-
23 (anti : syn = 83 : 17).[19] Since an unambiguous as- justing the carbon-hydrogen bond length to the
signment of the relative stereochemistry of such com- value of 1.08 A, obtained from neutron diffraction ex-
pounds by NMR is hampered by the observation of ro- periments, shortens this distance to 2.60 A. The hy-
Ê
Ê
tamers in solution, the crude product mixture was drogen bond pattern forms a zig-zag stack parallel to
recrystallized from hexane, thus affording crystals of the c axis of the crystal, while the hydrophobic parts
the major isomer that were suitable for X-ray analy- of the molecule separate individual stacks (Figure 2).
sis. Its molecular structure in the solid state (Fig-
The addition of (E)-octenylboronic acid to Garner
ure 1) confirms the anti-relationship between the aldehyde 23 had to be carried out at lower tempera-
newly formed stereocenter and the C-N bond. The ture (55 °C) in order to avoid competing proto-debor-
1,3-oxazolidine ring adopts a flattened twist confor- ylation. The reduced reaction rate was compensated
Ê
mation. The puckering amplitude of 0.31 A is slightly for by a somewhat higher catalyst loading (5 mol %);
Ê
smaller than the average of 0.36 A calculated from a this resulted in the formation of the desired truncated
selection of 105 1,3-oxazolidine rings contained in sphingosine derivative 25 in 78% yield after 2 h reac-
the Cambridge Structure Database.[20] The ring nitro- tion time, although the stereochemical bias towards
gen atom is planar (sum of the bond angles 359.3°)
Ê
and is less than 0.01 A from the ring mean plane. The
distances of the other ring atoms from the ring mean
plane are in agreement with approximate C2 symme-
try. Further analysis shows interesting crystal pack-
ing effects because an intermolecular hydrogen bond
between the hydroxy proton and the ring oxygen is
observed, whereas the stronger hydrogen bond ac-
Figure 1. Molecular structure diagram of anti-24 plus sym-
metry equivalent i molecule. Anisotropic displacement para-
meters are drawn at the 50% probability level, hydrogen
atoms are drawn with an arbitrary radius. Hydrogen bonds:
O(2)±H(O2) 0.88(4), H(O2). . .O(1)i 1.90(4), O(2). . .O(1)i
2.770(2) A, Ð(O(2)H(O2)O(1)i) 171(3)°; C(2)±H(2A)
Ê
1.01(2), H(2A). . .O(3)ii 2.58(2), C(2). . .O(3)ii 3.441(3) A,
Figure 2. Crystal packing diagram of anti-24 viewed along
the c direction. The phenyl ring C±H. . .p herring-bone pat-
tern is parallel to the hydroxy. . .O-oxazolidine hydrogen
bond. The weak C±H. . .O bond links the stacks horizontally
(b direction).
Ê
Ð(C(2)H(2A)O(3)ii) 142.6(17)°. Symmetry operators i) ±x,
1/2 + y, 1/2 ± z ; ii) ±x + 1, ±y + 1, ±z + 1. The shortest C±H con-
i
Ê
tacts between the phenyl rings are H(11) . . .C(15) 2.91 A,
H(11) . . .C(14) 2.93 A.
i
Ê
Adv. Synth. Catal. 2001, 343, 343±350
347

asc.wiley-vch.de
128.7, 128.0, 127.6, 126.6, 114.1, 75.9, 55.5; MS: m/z (rel. in-
tensity) = 214 ([M+], 76), 213 (18), 197 (14), 137 (42), 136
(13), 135 (50), 109 (100), 108 (39), 105 (53), 94 (13), 77 (39).
The analytical and spectroscopic data are in agreement with
those previously reported in the literature.[24]
the anti-isomer is somewhat lower under these con-
ditions (Table 4, entry 2).
Conclusion
All other compounds were prepared analogously. The
analytical and spectroscopic data of products 9,[25] 10,[26]
12,[27] 14,[28] 17,[29] 24,[19] were in agreement with those of
authentic samples prepared according to literature proce-
dures; the data of new compounds are compiled below.
Compound 8: IR: n = 3219, 3059, 3027, 2884, 1692, 1569,
1493, 1454, 1418, 1313, 1180, 1091, 1039, 1024, 778, 768,
702 cm±1; 1H NMR (300 MHz, CD2Cl2): d = 7.20±7.59 (m, 9H),
5.76 (s, 1H), 2.72 (br. s, 1H, -OH); 13C NMR (75 MHz, CD2Cl2):
d = 146.8, 143.8, 130.7, 130.4, 129.7, 128.9, 128.2, 126.8, 125.5,
122.8, 75.7; MS: m/z (rel. intensity) = 264/262 ([M+], 20), 185
(19), 183 (32), 165 (12), 107 (13), 106 (11), 105 (100), 79 (27),
78 (25), 77 (44), 76 (10), 51 (15); anal.: calcd. for C13H11BrO:
C, 59.3; H 4.2; found: C, 59.22; H, 4.08.
A convenient and highly user-friendly method for the
addition of aryl- or alkenylboronic acids to aldehydes
is presented which employs a catalyst formed in situ
from RhCl3 ´ 3 H2O (1 mol %), the readily accessibly
and fully air-stable imidazolium salt 2 (1 mol %) and
inexpensive aqueous bases. The addition is highly
chemoselective for aldehydes, it turned out to be
compatible with many functional groups in both reac-
tion partners, and exhibits a high preference for a
non-chelation controlled pathway in reactions with
Garner aldehyde. Future investigations are aiming at
the development of an asymmetric version of this
process and at exploiting its favorable profile in a
combinatorial set-up using resin-bound reagents.
Compound 11: IR: n = 3480, 3088, 3028, 2922, 2847, 1703,
1604, 1495, 1464, 1452, 1404, 1372, 1357, 1334, 1244, 1163,
1108, 1083, 1055, 1032, 758, 701 cm±1 1H NMR (300 MHz,
;
CD2Cl2): d = 7.24±7.35 (m, 5H), 4.63 (dd, 1H, J = 6.2, 5.8 Hz),
2.39 (t, 2H, J = 7.3 Hz), 2.19 (br. s, 1H, OH), 2.08 (s, 3H), 1.50±
1.74 (m, 4H), 1.27 (br. m, 14H); 13C NMR (75 MHz, CD2Cl2):
d = 209.3, 145.8, 128.6, 127.6, 126.2, 74.7, 44.0, 39.7, 29.9,
29.8, 29.7, 29.5, 26.2, 24.2; MS: m/z (rel. intensity) = 290
([M+], 12), 184 (12), 107 (100), 91 (13), 79 (35), 77 (14), 71
(33), 59 (12), 58 (32), 43 (34), 41 (10); anal.: calcd. for
Experimental Section
General Remarks
All reactions were carried out under Ar. The DME used was
purified by distillation over Na/K alloy prior to use. Flash
chromatography: Merck silica gel 60 (230±400 mesh). NMR:
spectra were recorded on DPX 300 spectrometer in the sol-
vents indicated; chemical shifts (d) are given in ppm relative
to TMS, coupling constants (J) in Hz. IR: Nicolet Magna 750
FT-IR, wave numbers in cm±1. MS (EI): Finnigan MAT 8200
(70 eV), HR-MS: Finnigan MAT 95. All arylboronic acids
were purchased (Lancaster, Aldrich) and used as received.
Aldehydes were distilled under reduced pressure prior to
use.
C19H30O2: C, 78.6; H, 10.4; found: C, 78.78; H, 10.44.
Compound 13: IR: n = 3387, 3193, 3027, 2897, 1613, 1601,
1511, 1452, 1421, 1334, 1238, 1222, 1175, 1163, 1122, 1105,
1007, 859, 836, 818 cm±1 1H NMR (300 MHz, THF-d8):
;
d = 8.17 (br, 1 H, OH); 7.55 (s, 4 H), 7.13 (d, 2 H, J = 8.5 Hz),
6.66 (d, 2 H, J = 8.5 Hz), 5.68 (s, 1 H), 4.80 (br. s, 1 H, OH);
13C NMR (75 MHz, THF-d8): d = 158.0, 151.6, 136.7, 129.2 (q,
J = 32 Hz), 128.7, 127.6, 125.6, 125.5 (q, J = 269 Hz), 115.8,
75.4; MS: m/z (rel. intensity) = 268 ([M+], 54), 267 (11), 173
(40), 145 (18), 123 (43), 122 (14), 121 (32), 95 (100), 94 (15),
77 (16); anal.: calcd. for C14H11F3O: C, 62.7; H, 4.1; found: C,
62.57; H, 4.04.
Compound 15: IR: n = 3359, 2896, 1619, 1504, 1489, 1444,
1327, 1248, 1164, 1124, 1067, 1040, 1017, 930, 853, 811, 785,
Representative Procedure for the Rhodium-
Catalyzed Addition of Boronic Acids to Aldehydes;
(4-Methoxyphenyl)phenylmethanol (7)
764 cm±1 1H NMR (300 MHz, CD2Cl2): d = 7.53 (AB, 4 H,
;
J = 8.3 Hz), 6.76±6.86 (m, 3 H), 5.93 (m, 2 H), 5.78 (s, 1 H),
2.61 (br. s, 1 H, OH); 13C NMR (75 MHz, CD2Cl2): d = 148.4,
147.7, 137.9, 129.4 (q, J = 32 Hz), 126.9, 125.7, 124.9 (q,
J = 270 Hz), 120.5, 108.4, 107.4, 101.7, 75.7; MS: m/z (rel. in-
tensity) = 296 ([M+], 100), 279 (13), 197 (10), 173 (32), 152
(11), 151 (39), 149 (22), 145 (19), 127 (10), 123 (88), 121
(10), 93 (63), 65 (25); anal.: calcd. for C15H11F3O3: C, 60.8; H,
3.7; found: C, 60.76; H, 3.73.
Phenylboronic acid (2.40 g, 19.6 mmol), imidazolium chlor-
ide 2 (42 mg, 0.098 mmol, 1 mol %),[13a] NaOMe (0.53 g,
9.8 mmol), 4-methoxybenzaldehyde (1.33 g, 9.8 mmol) and
water (10 mL) were successively added to a suspension of
RhCl3 ´ 3 H2O (26 mg, 0.098 mmol, 1 mol %) in DME (40
mL). The resulting mixture was heated for 30 min at 80 °C,
cooled to ambient temperature, diluted with ethyl acetate
(50 mL) and extracted with water. After drying over Na2SO4,
the organic phase was evaporated and the residue was puri-
fied by flash chromatography (hexane/ethyl acetate, 6/1)
thus affording the title compound as a colorless syrup that
slowly crystallized upon standing at ambient temperature;
yield: 1.96 g (93%); IR: n = 3408, 3065, 3008, 2951, 2909,
2836, 1612, 1587, 1517, 1494, 1445, 1305, 1265, 1254, 1178,
Compound 16: IR: n = 3080, 3020, 2993, 2941, 1750, 1719,
1690, 1622, 1607, 1590, 1486, 1423, 1371, 1329, 1248, 1223,
1167, 1120, 1069, 1031, 1017, 984, 928, 861, 707 cm±1
;
1H NMR (300 MHz, CD2Cl2): d = 7.12±7.66 (m, 8 H), 6.91 (s,
1 H), 2.22 (s, 6 H), 2.17 (s, 3 H); 13C NMR (75 MHz, CD2Cl2):
d = 173.0, 170.0, 144.3, 141.9, 140.4, 130.4 (q, J = 32 Hz),
130.3, 129.1, 127.7, 125.9, 124.5 (q, J = 270 Hz), 75.8, 27.1,
21.1; MS: m/z (rel. intensity) = 393 ([M+], 15), 351 (65), 291
(75), 250 (32), 181 (53), 43 (100); anal.: calcd. for
1110, 1034, 1018, 1008, 841, 810, 727, 697 cm±1 1H NMR
;
(300 MHz, CDCl3): d = 7.26±7.37 (m, 7 H), 6.86 (d, 2 H,
J = 8.0 Hz), 5.77 (br. s, 1 H, OH), 3.78 (s, 3 H), 2.34 (br. s, 1 H,
OH); 13C NMR (75 MHz, CDCl3): d = 159.4, 144.8, 136.8,
C20H18F3NO4: C, 61.1; H, 4.6; found: C, 61.28; H, 4.53.
Compound 18: IR: n = 3570, 3475, 3002, 2968, 2941, 2840,
348
Adv. Synth. Catal. 2001, 343, 343±350

FULL PAPERS
1617, 1594, 1477, 1459, 1437, 1417, 1329, 1245, 1221, 1168,
1130, 1111, 1068, 1033, 1016, 873, 841, 783, 766, 731 cm±1
3 H, J = 6 Hz); 13C NMR (75 MHz, CD2Cl2): d = 133.3, 128.9,
94.8, 81.2, 74.4, 65.4, 62.8, 32.7, 32.1, 29.5, 29.3, 28.5, 26.5,
24.8, 23.0, 14.2; MS: m/z (rel. intensity) = 341 ([M+], 1), 200
(31), 144 (26), 100 (72), 57 (100); anal.: calcd. for
;
1H NMR (300 MHz, CD2Cl2): d = 7.52 (AB, 4 H, J = 8.5 Hz),
7.28 (t, 1 H, J = 8.5 Hz), 6.65 (d, 2 H, J = 8.5 Hz), 6.33 (br. d,
1 H, J = 11.3 Hz), 4.31 (br. d, 1 H, J = 11.3 Hz), 3.80 (s, 6 H);
13C NMR (75 MHz, CD2Cl2): d = 158.0, 149.9, 129.7, 128.5 (q,
J = 32 Hz), 126.2, 125.1, 124.9 (q, J = 270 Hz), 119.3, 104.9,
68.2, 56.1 (2´); MS: m/z (rel. intensity) = 313 (16), 312 ([M+],
94), 295 (17), 294 (85), 293 (10), 279 (11), 263 (15), 173 (21),
168 (12), 167 (100), 165 (24), 159 (16), 151 (19), 149 (29), 145
(22), 139 (23), 137 (24), 135 (20), 127 (16), 122 (17), 107 (25),
77 (16); anal.: calcd. for C16H15F3O3: C, 61.5; H, 4.8; found: C,
61.57; H, 4.84.
C19H35NO4: C, 66.8; H, 10.3; found: C, 66.59; H, 10.15.
X-Ray Crystallographic Study of anti-24
C17H25NO4, M = 307.38 g ´ mol±1, colorless, crystal dimen-
sions 0.50 ´ 0.25 ´ 0.05 mm, monoclinic P21/c (no. 14), at
Ê
100 K a = 12.7491(8), b = 18.2535(12), c = 7.4578(4) A,
3
±3
Ê
b = 99.750(2), V = 1710.48(18) A , Z = 4, r = 1.194 Mg ´ m
,
m = 0.084 mm±1, l = 0.71073 A. X-ray diffraction data were
collected using a Nonius KappaCCD diffractometer employ-
ing w-scans to cover reciprocal space up to 33.14° q with
99.8% completeness, integration of raw data yielded a total
of 17533 reflections, merged into 6510 unique reflections
with Rint = 0.1900 after applying Lorentz, polarization and
absorption corrections. The structure was solved by direct
method using SHELXS-97,[30] and atomic positions and dis-
placement parameters were refined using full matrix least-
squares based on F2 using SHELXL-97.[30] Refinement of 299
parameters using all reflections converged at R = 0.0836,
Ê
Compound 19: IR: n = 3347, 3108, 2882, 1620, 1417, 1327,
1165, 1125, 1067, 1016, 854, 838, 790, 761, 710 cm±1; 1H NMR
(300 MHz, CD2Cl2): d = 7.57 (AB, 4 H, J = 8.1 Hz), 7.32 (dd,
1 H, J = 3.0, 5.1 Hz), 7.21 (dd, 1 H, J = 1.0, 3.0 Hz), 6.99 (dd,
1 H, J = 1.0, 5.0 Hz), 5.93 (s, 1 H), 2.71 (br. s, 1 H, OH);
13C NMR (75 MHz, CD2Cl2): d = 148.0, 145.2, 129.8 (q,
J = 32 Hz), 127.0, 126.5, 125.8, 124.7 (q, J = 270 Hz), 122.4,
72.3; MS: m/z (rel. intensity) = 258 ([M+], 58), 225 (14), 173
(32), 145 (21), 127 (14), 113 (24), 112 (15), 111 (38), 85
(100); anal.: calcd. for C12H9F3OS: C, 55.8; H, 3.5; found: C,
55.90; H, 3.51.
wR = 0.2376, highest residual electron density peak 0.698
Compound 20: IR: n = 3373, 2998, 2956, 2926, 2855, 1667,
1611, 1586, 1512, 1465, 1442, 1302, 1248, 1173, 1038, 968,
3
Ê
A . Crystallographic data (excluding structure factors) have
been deposited with the Cambridge Crystallographic Data
Centre as supplementary publication no. CCDC-159665. Co-
pies of the data can be obtained free of charge on application
to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [Fax:
(+44) 1223±336±033; E-mail: deposit@ccdc.cam.ac.uk].
;
831 cm±1 1H NMR (300 MHz, CD2Cl2): d = 7.28 (d, 2 H,
J = 6.8 Hz), 6.88 (d, 2 H, J = 6.8 Hz), 5.65±5.74 (m, 2 H), 5.08
(d, 1 H, J = 4.9 Hz), 3.79 (s, 3 H), 2.03±2.11 (m, 3 H), 1.28±
1.40 (m, 8 H), 0.91 (t, 3 H, J = 7.0 Hz); 13C NMR (75 MHz,
CD2Cl2): d = 159.4, 136.4, 133.1, 132.4, 127.7, 114.0, 74.9,
55.6, 32.6, 32.1, 29.5, 29.3, 23.0, 14.3; MS: m/z (rel. intensity):
248 ([M+], 35), 247 (11), 217 (10), 164 (13), 163 (100), 150
(49), 137 (26), 135 (60), 121 (47), 109 (33), 108 (18), 77 (14),
55 (39); anal.: calcd. for C16H24O2: C, 77.4; H, 9.7; found: C,
77.67; H, 9.93.
Acknowledgements
Generous financial support by the Deutsche Forschungsge-
meinschaft (Leibniz award to A. F.) and the Fonds der Che-
mischen Industrie is gratefully acknowledged. We thank Mr.
J. Rust and Dr. C. W. Lehmann for the X-ray structure of com-
pound anti-24.
Compound 21: IR: n = 3348, 2956, 2925, 2855, 1670, 1637,
1466, 1378, 1306, 1055, 1004, 967, 723 cm±1
;
1H NMR
(300 MHz, CD2Cl2): d = 5.59 (dt, 1 H, J = 6.1, 15.4 Hz), 5.54
(ddt, 1 H, J = 1.3, 6.7, 15.4 Hz), 3.98 (q, 1 H, J = 6.7 Hz), 2.02
(q, 2 H, J = 6.5 Hz), 1.28±1.51 (m, 22 H), 0.88 (t, 6 H,
J = 6 Hz); 13C NMR (75 MHz, CD2Cl2): d = 133.8, 132.1, 73.4,
37.9, 32.6, 32.3, 32.1, 30.0, 29.7, 29.6, 29.2, 26.1, 25.9, 23.1,
23.0, 14.3; MS: m/z (rel. intensity) = 254 ([M+], 3), 169 (20),
156 (11), 141 (99), 123 (22), 96 (12), 81 (41), 71 (24), 69 (13),
67 (26), 58 (13), 57 (100); anal.: calcd. for C17H34O: C, 80.2; H,
13.5; found: C, 80.41; H, 13.30.
References and Notes
[1] (a) For a recent progress report, however, on the pre-
paration of functionalized Grignard reagents and re-
lated organometallic species see: A. Boudier, L. O.
Bromm, M. Lotz, P. Knochel, Angew. Chem. 2000, 112,
4584±4606; Angew. Chem. Int. Ed. 2000, 39, 4414±4435;
for selected reviews on other types of functionalized
C-nucleophiles see the following for leading refer-
ences: (b) Zinc: P. Knochel, R. D. Singer, Chem. Rev.
1993, 93, 2117±2188; (c) Copper: R. D. Rieke, M. S.
Sell, W. R. Klein, T. Chen, J. D. Brown, M. V. Hanson
in Active Metals. Preparation, Characterization, Appli-
cations (Ed.: A. FuÈ rstner), VCH, Weinheim, 1996,
pp. 1±59; (d) A. FuÈ rstner, Angew. Chem. 1993, 105,
171±197; Angew. Chem. Int. Ed. Engl. 1993, 32, 164±
189; (e) Chromium: A. FuÈ rstner, Chem. Rev. 1999, 99,
991±1045; (f) Tin: H. Nozaki in Organometallics in
Synthesis. A Manual (Ed.: M. Schlosser), Wiley, Chi-
chester, 1994, pp. 535±578; (g) Manganese: A. FuÈ rst-
Compound 22: IR: n = 3429, 2956, 2927, 2857, 1667, 1478,
1464, 1393, 1379, 1363, 1098, 1037, 997, 970 cm±1 1H NMR
;
(300 MHz, CD2Cl2): d = 5.60 (dt, 1H, J = 6.8, 15.4 Hz), 5.49
(ddt, 1 H, J = 1.1, 7.2, 15.4 Hz), 3.65 (d, 1 H, J = 7.2 Hz), 2.04
(q, 2 H, J = 6.7 Hz), 1.47 (br. s, 1 H, OH), 1.27±1.38 (m, 8 H),
0.86±0.91 (m, 12 H); 13C NMR (75 MHz, CD2Cl2): d = 133.9,
130.4, 81.3, 35.0, 32.8, 32.1, 30.0, 29.2, 25.9, 23.0, 14.3; MS:
m/z (rel. intensity) = 198 ([M+], 1), 141 (41), 123 (15), 81
(30), 67 (16), 57 (100); anal.: calcd. for C13H26O: C, 78.7; H,
13.2; found: C, 78.96; H, 13.08.
Compound anti-25: [a]2D0 = 2.1° (c 0.73, CH2Cl2); IR:
n = 3443, 2977, 2959, 2928, 2873, 2857, 1700, 1478, 1457,
1390, 1366, 1256, 1175, 1098, 1071, 1052, 966, 849, 767 cm±1
;
1H NMR (300 MHz, CD2Cl2): d = 5.70 (dt, 1 H, J = 6.2,
15.5 Hz), 5.43 (ddt, 1 H, J = 1.3, 6.2, 15.5 Hz), 3.80±4.20 (br.
m, 4 H), 2.03 (q, 2 H, J = 6.5 Hz), 1.26±1.51 (m, 23 H), 0.88 (t,
Adv. Synth. Catal. 2001, 343, 343±350
349

asc.wiley-vch.de
ner, H. Brunner, Tetrahedron Lett. 1996, 37, 7009±7012
and references cited therein.
[14] H. C. Brown, S. K. Gupta, J. Am. Chem. Soc. 1975, 97,
5249±5255.
[2] M. Sakai, M. Ueda, N. Miyaura, Angew. Chem. 1998,
110, 3475±3477; Angew. Chem. Int. Ed. 1998, 37, 3279±
3281.
[3] M. Ueda, N. Miyaura, J. Org. Chem. 2000, 65, 4450±
4452.
[15] In contrast, attempted addition of butylboronic acid to
p-methoxybenzaldehyde was unsuccessful.
[16] (a) P. Garner, J. M. Park, J. Org. Chem. 1987, 52, 2361±
2364; (b) See also: P. Garner, J. M. Park, E. Malecki, J.
Org. Chem. 1988, 53, 4395±4398.
[4] R. A. Batey, A. N. Thadani, D. V. Smil, Org. Lett. 1999,
1, 1683±1686.
[5] For related transmetalations of organotin reagents to
rhodium see: S. Oi, M. Moro, Y. Inoue, Chem. Com-
mun. 1997, 1621±1622.
[6] For related rhodium-catalyzed additions of boronic
acids to imines or 1,4-additions to enones or nitro-
alkenes see: (a) M. Ueda, A. Saito, N. Miyaura, Synlett
2000, 1637±1639; (b) Y. Takaya, M. Ogasawara,
T. Hayashi, M. Sakai, N. Miyaura, J. Am. Chem. Soc.
1998, 120, 5579±5580; (c) T. Hayashi, T. Senda,
M. Ogasawara, J. Am. Chem. Soc. 2000, 122, 10716±
10717.
[17] For recent reviews on sphingosine and analogues see:
(a) P. M. Koskinen, A. M. P. Koskinen, Synthesis 1998,
1075±1091; (b) R. R. Schmidt in Synthesis in Lipid
Chemistry (Ed.: J. H. P. Tyman), Spec. Publ. ± Royal So-
ciety of Chemistry, Cambridge, UK, Vol. 180, 1996, 93±
118.
[18] For leading references see the following and refer-
ences therein: (a) J. Chun, L. He, H.-S. Byun, R. Bitt-
man, J. Org. Chem. 2000, 65, 7634±7640; (b) L. Zhao,
L. Qiao, S.-B. Rong, A. P. Kozikowski, Tetrahedron Lett.
2000, 41, 8711±8715.
[19] L.Williams, Z. Zhang, F. Shao, P. J. Carroll, M. M.
JoullieÂ, Tetrahedron 1996, 52, 11673±11694.
[7] Reviews: (a) D. Bourissou, O. Guerret, F. P. Gabbai,
G. Bertrand, Chem. Rev. 2000, 100, 39±91; (b) W. A.
Herrmann, C. KoÈcher, Angew. Chem. 1997, 109, 2256±
2282; Angew. Chem. Int. Ed. Engl. 1997, 36, 2162±
2187.
[8] Recent reviews: (a) T. M. Trnka, R. H. Grubbs, Acc.
Chem. Res. 2001, 34, 18±29; (b) A. FuÈ rstner, Angew.
Chem. 2000, 112, 3140±3172; Angew. Chem. Int. Ed.
2000, 39, 3012±3043.
[9] (a) M. Scholl, T. M. Trnka, J. P. Morgan, R. H. Grubbs,
Tetrahedron Lett. 1999, 40, 2247±2250; (b) J. Huang, E.
D. Stevens, S. P. Nolan, J. L. Petersen, J. Am. Chem.
Soc. 1999, 121, 2674±2678; (c) L. Ackermann, A. FuÈ rst-
ner, T. Weskamp, F. J. Kohl, W. A. Herrmann, Tetrahe-
dron Lett. 1999, 40, 4787±4790.
[20] Cambridge Structural Database Ver. 5.20 October
2000. The search was limited to monocyclic oxazoli-
dine rings with sp3-ring carbon atoms, the third bond
of the nitrogen atom was restricted to C and H (D).
Database entries were required to be error-free and
limited to organic molecules only.
[21] T. Steiner, Crystallogr. Rev. 1996, 6, 1±57.
[22] In contrast, the solid state structure of a related syn-
configurated compound formed by addition of
lithiated 2-phenyl-5,6-dihydro-1,4-dithiin to Garner al-
dehyde shows a strong intramolecular hydrogen bond
between the OH and the carbonyl oxygen of the NBoc
group, cf. R. Caputo, L. Longobardo, G. Palumbo, S.
Pedatella, F. Giordano, Tetrahedron 1996, 52, 11857±
11866.
[10] (a) A. FuÈ rstner, O. R. Thiel, L. Ackermann, H.-J.
Schanz, S. P. Nolan, J. Org. Chem. 2000, 65, 2204±2207;
(b) L. Ackermann, D. El Tom, A. FuÈ rstner, Tetrahe-
dron 2000, 56, 2195±2202; (c) A. FuÈ rstner, O. R. Thiel,
N. Kindler, B. Bartkowska, J. Org. Chem. 2000, 65,
7990±7995; (d) A. FuÈ rstner, O. R. Thiel, G. Blanda,
Org. Lett. 2000, 2, 3731±3734.
[11] (a) C. Zhang, J. Huang, M. L. Trudell, S. P. Nolan, J.
Org. Chem. 1999, 64, 3804±3805; (b) W. A. Herrmann,
C.-P. Reisinger, M. Spiegler, J. Organomet. Chem.
1998, 557, 93±96; (c) J. Huang, S. P. Nolan, J. Am.
Chem. Soc. 1999, 121, 9889±9890.
[23] For a discussion of the geometric analysis of
(Csp2)2C±H. . .p interactions and parameter definitions
see: Y. Umezawa, S. Tsuboyama, K. Honda, J. Uzawa,
M. Nishio, Bull. Chem. Soc. Jpn. 1998, 71, 1207±1213.
[24] P. Strazzolini, A. G. Giumanini, G. Verardo, Tetrahe-
dron 1994, 50, 217±254.
[25] N.-S. Li, S. Yu, G. W. Kabalka, J. Organomet. Chem.
1997, 531, 101±105.
[26] A. Guijarro, D. J. Ramon, M. Yus, Tetrahedron 1993,
49, 469±482.
[27] D. Touiti, R. Jost, J. Sommer, J. Chem. Soc. Perkin
Trans. 2 1986, 1793±1797.
[12] A. FuÈ rstner, A. Leitner, Synlett 2001, 290±292.
[13] For the preparation of imidazolium salts and the
synthesis, isolation and structure of the corresponding
NHC's see: (a) A. J. Arduengo, R. Krafczyk, R. Schmut-
zler, H. A. Craig, J. R. Goerlich, W. J. Marshall, M. Un-
verzagt, Tetrahedron 1999, 55, 14523±14534; (b) A. J.
Arduengo, Acc. Chem. Res. 1999, 32, 913±921.
[28] B. Ancian, F. Membrey, J. P. Doucet, J. Org. Chem.
1978, 43, 1509±1518.
[29] J.-S. Shiue, M.-H. Lin, J.-M. Fang, J. Org. Chem. 1997,
62, 4643±4649.
[30] G. M. Sheldrick, SHELXS-97 Program for Crystal
Structure Solution and Refinement, University of GoÈt-
tingen, Germany, 1997.
350
Adv. Synth. Catal. 2001, 343, 343±350