Germanium and Silicon Linking Strategies for Traceless Solid-Phase Synthesis

Article abstract of DOI:10.1021/jo961889y

Here we describe the development and application of germanium and silicon linkage strategies for the solid-phase synthesis of aromatic compounds, as demonstrated for 1,4-benzodiazepine derivatives. The metal-aryl bond that attaches the benzodiazepine to t

Full text of DOI:10.1021/jo961889y

J . Org. Chem. 1997, 62, 2885-2893
2885
Ger m a n iu m a n d Silicon Lin k in g Str a tegies for Tr a celess
Solid -P h a se Syn th esis
Matthew J . Plunkett1 and J onathan A. Ellman*
Department of Chemistry, University of California, Berkeley, California 94720
Received October 7, 1996X
Here we describe the development and application of germanium and silicon linkage strategies for
the solid-phase synthesis of aromatic compounds, as demonstrated for 1,4-benzodiazepine deriva-
tives. The metal-aryl bond that attaches the benzodiazepine to the polymeric support may be
cleaved by electrophilic reagents such as H+ and Br2 to provide the corresponding substituted
derivatives. A number of compounds have been prepared in good overall yield for the eight- or
nine-step process. This approach can also be applied to the solid-phase synthesis of libraries of
other classes of aromatic compounds.
In tr od u ction
site of linker attachment.6 In many cases the previous
two methods may not be applicable to a desired com-
pound class or may limit the chemistry that can be
performed. The third strategy, which is often necessary,
is the introduction of an auxiliary functional group (such
as a phenol, alcohol, amide, or carboxylate) as a handle
for linker attachment. After cleavage from the solid
support at the end of a synthesis sequence, the linking
functional group can have a negligible, positive, or
negative effect on the biological or chemical activity of
the target molecule, depending on where it is situated
and on the specific receptor/ligand interaction.
The synthesis and screening of small-molecule libraries
is a powerful method for the discovery of new molecules
with a particular desired property.2 A library of organic
compounds can be made in solution or on a solid support;
both approaches offer distinct advantages. When a
library is made in solution, the development of chemistry
is sometimes more straightforward than that on a solid
support. Purity of the final product can be improved by
the use of support-bound reagents,3 and extractions can
be used to remove charged impurities.4 However, most
of the chemical combinatorial libraries reported to date
have been constructed using solid-phase methods. Reac-
tions performed on a solid support can be driven to
completion by the use of excess reagents, and purifica-
tions between reaction steps are simplified because
noncovalently bound material may simply be rinsed
away. For this reason, the multistep solution synthesis
of a library of architecturally complex compounds is more
difficult than that on a solid phase. A key aspect of any
solid-phase synthesis strategy is the linkage element,
which acts as a tether to the polymeric support material.
Ideally, the linker should be stable to all reactions used
in a synthesis sequence and should be cleaved quanti-
tatively under conditions that do not degrade the desired
target molecule.
An alternative to the above approaches is linkage
through a functional group that can be excised efficiently
and quantitatively when desired, leaving behind no trace
or “memory” of the solid-phase synthesis. For aromatic
compounds, we felt that linkage by a bond between a
Group IV metal and the aromatic ring would fulfill these
requirements.7 Periodic trends in the reactivity of Group
IV metals are well documented.8 Silylaromatic com-
pounds are used to block reactivity at specific sites within
an aromatic ring,9,10 and arylsilanes can be removed by
an acid or basic fluoride ion. The effect of aromatic
substitution11 and silicon substituents12 on reaction rate
has been studied extensively. Silicon-directed ipso sub-
stitution of arylsilanes is frequently used for the high-
yielding, regiospecific introduction of electrophilic func-
tional groups.13 Germanium is used less frequently in
organic synthesis, although an extensive compilation of
organogermanium chemistry is available.14 One notable
use of germanium chemistry is Moerlein’s synthesis of
aryltrimethylgermanium substrates for the clean, high-
yielding preparation of regiospecifically radiolabeled halo-
Linking methods may be roughly grouped into three
categories. The first strategy is to link through func-
tionality already present in the desired target molecule,
such as in solid-phase peptide and oligonucleotide syn-
thesis. The second approach is a cleavage where the
linking functional group is incorporated into the final
molecule.5 Additional diversity may be introduced at this
(6) Backes, B. F.; Virgilio, A. A.; Ellman, J . A. J . Am. Chem. Soc.
1996, 118, 3055.
(7) Sucholeiki has described the use of a benzyl thioether that
cleaves upon irradiation (350 nm) to provide a phenylmethyl group.
Sucholeiki, I. Tetrahedron Lett. 1994, 35, 7307.
(8) Eaborn, C.; Pande, K. C. J . Chem. Soc. 1960, 1566.
(9) Mills, R. J .; Taylor, N. J .; Snieckus, V. J . Org. Chem. 1989, 54,
4372.
(10) Sengupta, S.; Snieckus, V. Tetrahedron Lett. 1990, 31, 4270.
(11) Deans, F. B.; Eaborn, C. J . Chem. Soc. 1959, 2299.
(12) Bott, R. W.; Eaborn, C.; J ackson, P. M. J . Organomet. Chem.
1967, 7, 79.
(13) Chan, T. C.; Fleming, I. Synthesis 1979, 761.
(14) (a) MacDiarmid, A. G. Organometallic Componds of the Group
IV Elements, Vol. 1: The Bond to Carbon; Marcel Dekker: New York,
1968. (b) MacDiarmid, A. G. Organometallic Componds of the Group
IV Elements, Vol. 2: The Bond to Halogens and Halogenoids; Marcel
Dekker: New York, 1972.
X Abstract published in Advance ACS Abstracts, April 1, 1997.
(1) Present address: Arris Pharmaceutical Corp., 385 Oyster Point
Blvd., Suite 3, South San Francisco, CA 94080.
(2) (a) Thompson, L. A.; Ellman, J . A. Chem. Rev. 1996, 96, 555. (b)
Terrett, N. K.; Gardner, M.; Gordon, D. W.; Kobylecki, R. J .; Steele, J .
Tetrahedron 1995, 51, 8135. (c) Gordon, E. M.; Barrett, R. W.; Dower,
W. J .; Fodor, S. P. A.; Gallop, M. A. J . Med. Chem. 1994, 37, 1385.
(3) (a) Desai, M. C.; Stephens Stramiello, L. M. Tetrahedron Lett.
1993, 34, 7685. (b) Virgilio, A. A.; Schu¨rer, S. C.; Ellman, J . A.
Tetrahedron Lett. 1996, 37, 6961. (c) Kaldor, S. W.; Siegel, M. G.; Fritz,
J . E.; Dressman, B. A. Tetrahedron Lett. 1996, 37, 7193.
(4) Boger, D. L.; Tarby, C. M.; Myers, P. L.; Caporale, L. H. J . Am.
Chem. Soc. 1996, 118, 2109.
(5) (a) Crowley, J . I.; Rapoport, H. J . Am. Chem. Soc. 1970, 92, 6363.
(b) DeWitt, S. H.; Kiely, J . S.; Stankovic, C. J .; Schroeder, M. C.;
Reynolds Cody, D. M.; Pavia, M. R. Proc. Natl. Acad. Sci. U.S.A. 1993,
90, 6909.
S0022-3263(96)01889-0 CCC: $14.00 © 1997 American Chemical Society

2886 J . Org. Chem., Vol. 62, No. 9, 1997
Plunkett and Ellman
Sch em e 1
aromatic compounds.15 Electrophilic demetalations with
iodine, bromine, and chlorine were found to proceed in
high yield; for this system trimethylsilicon was judged
to be too unreactive and trialkyltin was felt to be not
stable enough.
silane22 gave the arylsilane 2 in high yield. The tri-
methyltin group was introduced using the directed ortho
metalation reaction followed by addition of trimethyltin
chloride to give 3.23 While we were not satisfied with
the yield obtained in this reaction, variation of reaction
stoichiometry, solvent, and lithium reagent failed to
improve the amount of product obtained. In part, the
low yield was due to alkene lithiation and subsequent
alkene stannylation. Hydroboration of arylstannane 3
with 9-BBN and an oxidative workup with basic perox-
ide24 provided the primary alcohol 4. Mitsunobu reac-
tion25 of this alcohol with the cyanomethyl ester of (4-
hydroxyphenoxy)acetic acid afforded the preactivated
ester derivative 5 in good yield. Use of the cyanomethyl
ester eliminates a saponification step from the synthesis
sequence and also minimizes exposure of the acid-
sensitive arylstannane to the free phenoxyacetic acid and
to the acidic coupling additives that are often used for
amide bond formation. Finally, (aminomethyl)polysty-
rene resin is acylated with 5 using DMAP and i-Pr2EtN.26
The solid-phase synthesis route to benzodiazepine
derivatives is shown in Scheme 2. (For clarity, we
omitted the spacer in this scheme.) Stille coupling reac-
tions of stannane 6a with both aromatic and aliphatic
acid chlorides are performed using the ligandless catalyst
Pd2dba3‚CHCl3 to give 7a . The Bpoc protecting group is
removed by brief treatment with 3% trifluoroacetic acid
in CH2Cl2 to give 2-aminoaryl ketones 8a . This com-
pound is acylated with an R-N-Fmoc-amino acid fluo-
ride,27 and the protecting group is removed with 20%
piperidine in DMF to give 9a . This intermediate is
cyclized with mild acidic catalysis at 65 °C to afford
support-bound benzodiazepine derivative 10a . Depro-
tonation and alkylation give the fully functionalized
derivative 11a .
We have previously developed a method for the solid-
phase synthesis of chemically diverse 1,4-benzodiazepine
derivatives.16 The benzodiazepine derivatives are syn-
thesized from a support-bound, protected (aminoaryl)-
stannane and three variable components: acid chlorides,
Fmoc amino acids, and alkylating agents. The synthesis
sequence includes transition metal catalyzed bond forma-
tion, an acylation, and acidic and basic reaction condi-
tions. We felt that the solid-phase synthesis of this class
of pharmaceutically important molecules would be an
ideal test of a metal-based linker, and we have published
a preliminary communication on the application of a silyl
linkage approach.17 Since our initial communication of
a silyl linker, two reports of silicon linking strategies for
the synthesis of biaryl compounds have appeared.18,19
Here, we expand on our previous report and document
our investigations of alternate metal linkage approaches
and different electrophile-mediated cleavages.
Resu lts a n d Discu ssion
Silyl-Lin k ed Ben zod ia zep in es. The synthesis of an
appropriately functionalized silicon-containing (amino-
aryl)stannane is shown in Scheme 1. Synthesis began
with protection of 4-bromoaniline (1) using the ((2-(4-
biphenyl)isopropyl)oxy)carbonyl (Bpoc) group.20 The re-
duced basicity of the aniline (compared to an aliphatic
amine), and the thermal instability of the Bpoc reagent,
required the use of potassium hydride to generate the
anilide anion, which then added rapidly to biphenyl-
isopropyl phenyl carbonate. Two equivalents of KH was
used since the product is more acidic than the starting
material. Carbamate deprotonation with potassium hy-
dride, lithium-halogen exchange with tert-butyllithi-
um,21 and quenching with 3-butenylchlorodimethyl-
Deprotection and cleavage of the derivatives are shown
in Scheme 3. Any amino acid side-chain protecting
groups are removed by treatment of 11a with trifluoro-
acetic acid/dimethyl sulfide/water. The silicon-aryl bond
is stable to TFA (even at 60 °C) because of the extremely
(15) Moerlein, S. M. J . Org. Chem. 1987, 52, 664.
(16) Plunkett, M. J .; Ellman, J . A. J . Am. Chem. Soc. 1995, 117,
3306.
(17) Plunkett, M. J .; Ellman, J . A. J . Org. Chem. 1995, 60, 6006.
(18) (a) Chenera, B.; Finkelstein, J . A.; Veber, D. F. J . Am. Chem.
Soc. 1995, 117, 11999.
(22) Steinmetz, M. G.; Yu, C. J . Org. Chem. 1992, 57, 3107.
(23) (a) Stanetty, P.; Koller, H.; Mihovilovic, M. J . J . Org. Chem.
1992, 57, 6833. (b) Salituro, F. G.; McDonald, I. A. J . Org. Chem. 1988,
53, 6138.
(24) Fleming, I.; Lawrence, N. J . J . Chem. Soc., Perkin Trans. 1
1992, 3309.
(19) Han, Y.; Walker, S. D.; Young, R. N. Tetrahedron Lett. 1996,
37, 2703.
(20) Sieber, P.; Iselin, B. Helv. Chim. Acta 1968, 51, 622.
(21) KH is necessary because lithium-halogen exchange is faster
than proton transfer. Moyer, M. P.; Shiurba, J . F.; Rapoport, H. J .
Org. Chem. 1986, 51, 5106.
(25) Hughes, D. L. Org. React. (N.Y.) 1992, 42, 335.
(26) The active ester derived from (4-hydroxyphenyl)acetic acid
acylates at least 10-fold more slowly under these conditions, perhaps
due to reduced electrophilicity of the carbonyl carbon.
(27) Carpino, L. A.; Satat-Aalaee, D.; Chao, H. G.; DeSelms, R. H.
J . Am. Chem. Soc. 1990, 112, 9651.

Germanium and Silicon Linking Strategies
J . Org. Chem., Vol. 62, No. 9, 1997 2887
Sch em e 2
Sch em e 3
electron-poor nature of the protonated benzodiazepine.
The aryl-silicon bond is then cleaved with anhydrous
HF to give the product benzodiazepines 12. When
cleavage of 11a with basic fluoride was attempted (DMF,
CsF, 18-crown-6, 100 °C), no benzodiazepine product 12
was observed. The low reactivity under these conditions
is not surprising because the benzodiazepine is no longer
electron deficient when the imine is not protonated.
Although basic fluoride-mediated cleavage of electron-
poor arylsilanes proceeds cleanly,18 other researchers
have reported a lack of reactivity to fluoride for nonelec-
tron deficient aromatic systems.19
Using this general route we have synthesized deriva-
tives 12a -d , shown in Figure 1. These compounds
incorporate amide, phenol, ether, aromatic, and aliphatic
functionality. The unpurified products appear more pure
by TLC and NMR than the yields alone would suggest.
One reason for this discrepancy is that the loading to give
6a was not quantitative. After acylation of 5, the
unacylated aminomethyl groups were capped with Fmoc-
alanine using hydroxybenzotriazole (HOBt) and diiso-
propylcarbodiimide. Fmoc quantitation then gives the
percent of unreacted aminomethyl groups. If based upon
the loading level of the protected (aminoaryl)stannane,
the yields in Figure 1 would be higher by approximately
10%.
F igu r e 1. Benzodiazepine derivatives prepared using the silyl
linker. The yields are based on the initial aminomethyl
substitution level of the polystyrene resin and are determined
by the mass balance of purified material. For compound 12b
an additional 11% of the debenzylated product was isolated.
F igu r e 2. Possible mechanism for formation of silicon-
containing benzodiazepines 14.
The initial silyl linker that we evaluated incorporated
the three-carbon linker 13, derived from allylchlorodi-
methylsilane and analogous to our successful system 11a .
However, upon HF treatment of some derivatives, this
linker gives significant amounts of silicon-containing
benzodiazepine products 14, as determined by proton
NMR and mass spectrometry (Figure 2). The alternative
cleavage pathway is probably similar to the cyclopropane-
forming decomposition of (3-chloropropyl)diethylmethyl-
silane.28 We observed increased amounts of this side
reaction when the substituent at R1 is not strongly
electron donating. As R1 becomes less electron donating,
more electron density is pulled from the silicon-appended
aromatic ring. Since the rate-limiting step for the desired
protodesilation reaction is ipso protonation of the aro-
matic ring, lower electron density in this ring will slow
protodesilation. The observed trends are then explained
if the rate of the undesired fragmentation is independent
of substitution at R1.
When we began these investigations, we had hoped
that the products would be cleaved by the volatile reagent
trifluoroacetic acid. Cleavage with HF will present
problems with certain functional groups, such as the
debenzylation seen for compound 12b. However, the
commonly used cleavage reagent trifluoroacetic acid
should suffice for most classes of aromatic compounds.29
Biaryl compounds synthesized using a silyl linker have
been cleaved by TFA, fluoride ion, Br2, and ICl.18,19
Modification of the alkyl substituents in an aryl-
trialkylsilane dramatically changes the rate of proto-
desilation, with smaller substituents accelerating the rate
of cleavage. However, because we employ a dimethyl-
substituted silicon-linker, substituent modifications to
increase the lability of our system are not straightfor-
ward.30 We were inspired to investigate an analogous
(29) The protodesilation of less electron-poor aromatic compounds
is quite facile in TFA. Funk, R. L.; Vollhardt, K. P. C. J . Am. Chem.
Soc. 1980, 102, 5253.
(28) Ponomarenko, V. A., et al. Dokl. Akad. Nauk. SSSR 1956, 106,
76. Chem. Abstr. 1956, 50, 13726e.

2888 J . Org. Chem., Vol. 62, No. 9, 1997
Plunkett and Ellman
Sch em e 4
Sch em e 5
germanium linker by the periodic trends in the rate of
protodemetalation.8
Ger m a n iu m -Lin k ed Ben zod ia zep in es. To obtain
a linking element that could be cleaved with trifluoro-
acetic acid, we decided to investigate a similar linker
based on germanium. The synthesis of appropriately
functionalized germanium-linked resin is shown in Scheme
4. Synthesis began with protection of 4-bromoaniline (1)
as the 2-(4-methylphenyl)isopropyl carbamate (Mpc group).
Deprotonation, lithium-halogen exchange, and then
trapping of the aryl anion with chlorodimethyl(6-((tri-
ethylsilyl)oxy)hexyl)germane gave silyl ether 15. The
directed orthometalation reaction was used as before to
introduce the trimethyltin group to give 16. The silyl
ether is removed cleanly with tetrabutylammonium
fluoride (TBAF) to give alcohol 17. Mitsunobu reaction
with cyanomethyl (4-hydroxyphenoxy)acetate gives the
preactivated ester 18. (Aminomethyl)polystyrene resin
is then directly acylated with 18 in N-methylpyrrolidi-
none (NMP) to give resin 6b. The resin was again capped
as for the silyl system, and Fmoc quantitation showed
that 15% of the aminomethyl sites on the resin did not
react with the cyanomethyl ester.
The choice of tether length merits some discussion.
Although a three-carbon spacer derived from the allyl-
germanyl chloride would be suitable for trifluoroacetic
acid cleavage conditions, the synthesis of allyldimethyl-
germanium chloride is three steps from commercially
available materials.31 The four-carbon tether, derived
from a homoallyl Grignard reagent, is difficult to form
because of facile HBr elimination. We initially used a
five-carbon spacer derived from 5-chloro-1-pentene, but
the orthometalation reaction gave some products (as with
the silyl system) that seemed to correspond to alkene
stannylation. This prompted the investigation of a
different alcohol masking group, one that could be
removed using conditions that would not affect the
germanium or tin but would be stable to distillation,
Grignard reagents, and tert-butyllithium. The triethyl-
silyl ether was chosen because of its ease of cleavage,
stability to carbon bases, and low molecular weight. The
six-carbon linker, derived from 1-chloro-6-hexanol, was
used because of cost considerations.
The solid-phase route to germanium-linked benzo-
diazepines is shown in Scheme 2. The sequence is
analogous to the silyl route except for the carbamate
deprotecton of 7b to give 8b. While treatment of silyl-
linked 7a with 3% trifluoroacetic acid caused no detect-
able 2-aminobenzophenone cleavage from the resin, 7b
is less stable to acidic conditions. We initially protected
the aniline as the Bpoc carbamate as with the silicon
system but were unable to obtain less than 10-11%
cleavage of the 2-aminobenzophenone using a variety of
acids and reaction times. Substitution of the Mpc car-
bamate (which cleaves approximately 4× faster32 ) brought
the amount of undesired cleavage to 4-5%.33 Although
this is only a small amount of material, we wanted to
entirely eliminate any premature cleavage. We found
that treatment of 7b with the Lewis acid B-chloro-
catecholborane and 0.5 equiv of i-Pr2EtN cleaves the Mpc
group rapidly34 without any undesired 2-aminobenzo-
phenone cleavage. It is important to maintain anhydrous
conditions and to use pure reagent for this deprotection.
Acylation of the aniline, Fmoc deprotection, cyclization,
(30) Myers has demonstrated an innovative approach for increasing
the reactivity of silyl enol ethers by constraining the silicon as a
silacyclobutane. When the silicon is predisposed to formation of a
pentavalent intermediate, nucleophile-mediated silyl transfer reactions
are accelerated. It is not clear if this approach would facilitate acid-
catalyzed protodesilation reactions. Myers, A. G.; Kephart, S. E.; Chen,
H. J . Am. Chem. Soc. 1992, 114, 7922.
(31) Massol, M.; Cabadi, Y.; Satge´, J . Bull. Soc. Chim. Fr. 1971,
3235. Attempts to prepare the allyl-substituted reagent by addition of
allylmagnesium chloride to dimethylgermanium dichloride gave very
poor selectivity for mono- vs diaddition of the Grignard reagent, even
at -78 °C.
(32) Geiger, R.; Ko¨nig, W. In The Peptides: Analysis, Synthesis, and
Biology; Gross, E., Meienhofer, J ., Eds.; Academic Press: New York,
1981; Vol. 3, p 35.
(33) Complete carbamate deprotection can be obtained by treatment
of 7b with 0.2% of TFA in CH2Cl2 for 1 min or by 5% of dichloroacetic
acid in CH2Cl2 for 5 min, as determined by cleavage of the germanium-
aryl bond with bromine.
(34) Boeckman, R. K.; Potenza, J . C. Tetrahedron Lett. 1985, 26,
1411.

Germanium and Silicon Linking Strategies
J . Org. Chem., Vol. 62, No. 9, 1997 2889
reaction is allowed to proceed for several hours. No
overbromination was detected. Compound 20d was
synthesized with the phenol protected as the tert-butyl
ether, and the protecting group was removed after the
electrophilic cleavage from the resin. When the phenol
of tyrosine-containing benzodiazepines is deprotected
before electrophilic cleavage from the resin, most of the
product reattaches to the resin as the germanyl ether.
Optimization of this cleavage/reattachment could allow
for further functionalization of alcohol-containing prod-
ucts at the linkage site by a variety of palladium-
mediated methods.
Numerous functional groups are present in the deriva-
tives we prepared by the germanium route, including
amides, esters, acids, phenols, acetals, ethers, and ha-
lides. The tolerance of numerous functional groups, and
the ease of electrophilic demetalation, makes the ger-
manium route a significant improvement over the silyl
method for traceless 1,4-benzodiazepine synthesis. Al-
though dimethylgermanium dichloride is expensive (∼$25/
g), only small amounts would be needed for the synthesis
of a thousand compounds on a 10 µmol scale per
compound.
F igu r e 3. Figure 3. Germanium benzodiazepines prepared
by proteolytic cleavage from resin. Yields are based on the
mass balance of analytically pure material.
Con clu sion
As demonstrated in this work, metal linking strategies
are a viable approach for the solid-phase synthesis of
aromatic compounds. The judicious selection of the metal
or its appendant alkyl groups can give cleavage condi-
tions for an aromatic structure that are straightforward
to use and are compatible with a range of chemical
functionality.
Exp er im en ta l Section
Gen er a l. Unless otherwise noted, materials were obtained
from commercial suppliers and used without further purifica-
tion. The R-N-Fmoc-amino acid fluorides were prepared by
the method of Carpino et al.35 with slight modifications.36
Tetrahydrofuran and diethyl ether were distilled under N2
from sodium/benzophenone immediately prior to use. Organic
layers were dried over Na2SO4 before concentration. Flash
column chromatography was carried out using Merck 60 230-
400 mesh silica gel according to the procedure described by
Still.37 Thin layer chromatography (TLC) analyses were
performed with Merck 60 F254 0.25 µm silica gel plates. Unless
otherwise stated, TLC Rf values given were determined with
the solvent used for column chromatography. IR spectra were
recorded neat (for oils) and as films from CHCl3 or CH2Cl2 (for
crystalline compounds). 1H NMR spectra were obtained with
a UCB Bruker AM-400 or AM-500 FT spectrometer. Proton-
decoupled 13C spectra were obtained at 101 or 126 MHz with
the same instruments. Chemical shifts are reported in ppm.
J values are in hertz. High-resolution mass spectra were
obtained at the University of California mass spectral labora-
tory using fast atom bombardment (FAB) with 3-nitrobenzyl
alcohol as matrix solvent. Elemental analyses were performed
by M-H-W Labs, Phoenix, AZ. Gel form (aminomethyl)-
polystyrene resin was purchased from Bachem California (no.
RMIS12, 100-200 mesh, cross-linked with 1% divinylbenzene).
A Safe-Lab peptide flask (no. M2570, a glass cylinder with a
frit and a three-way valve at the bottom) was often used with
the resin. Solvents may be forced through by nitrogen
pressure at the top, and reaction mixtures may be gently
agitated by bubbling in nitrogen from below. A filtration
F igu r e 4. Figure 4. Germanium-linked benzodiazepines
prepared by electrophilic cleavage with bromine. Yields are
based on the mass balance of analytically pure material.
and alkylation were performed as for the silyl system.
The two different cleavage options are shown in Scheme
5.
Derivatives 19a -d were obtained by treatment of the
fully functionalized benzodiazepine derivatives 11b with
neat trifluoroacetic acid at 60 °C (Scheme 5). As with
the silicon system, side-chain protecting groups can be
removed before the compound is cleaved from the resin,
obviating the need for cation scavengers (such as water
and dimethyl sulfide) in the final cleavage. No debenzyl-
ation (<2%) upon cleavage was seen for compound 19c
(Figure 3). This is in contrast to the significant amount
of debenzylated product that was observed upon HF
cleavage of the silyl-linked benzodiazepine 12b.
We also investigated electrophilic cleavage of the
germanium-linked benzodiazepines. Cleavage of silicon-
linked biaryl compounds from resin by electrophiles such
as Br2 and ICl has recently been shown to proceed in good
yield.19 Shown in Figure 4 are several benzodiazepines
prepared by electrophilic cleavage of the germanium-
aryl bond with elemental bromine. The reaction is
complete in 5 min, although some decomposition of the
products (probably imine oxidation) is seen when the
(35) Carpino, L. A.; Sadat-Aalaee, D.; Chao, H. G.; DeSelms, R. H.
J . Am. Chem. Soc. 1990, 112, 9651. The fluorides are synthesized using
2.0 equiv of cyanuric fluoride and are purified by extraction.
(36) Bunin, B. A.; Plunkett, M. J .; Ellman, J . A. Methods Enzymol.
1996, 267, 486.
(37) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.

2890 J . Org. Chem., Vol. 62, No. 9, 1997
Plunkett and Ellman
cannula (Pharmacia, Inlet Tubing with Filter, catalog no. 19-
4448-01; Filter Inserts, catalog no. 19-5086-01) was used for
filtration from round bottom flasks.
NMR:38 δ -8.6, -3.0, 14.8, 27.9, 28.8, 81.6, 112.8, 121.5 (b),
124.8, 127.0, 127.1, 127.1, 128.6, 134.6, 134.9, 139.9, 140.8,
141.4, 141.7, 143.8, 144.9, 152.7. Anal. Calcd for
C31H41NO2SiSn: C, 61.40; H, 6.81; N, 2.31, found C, 61.39; H,
7.00; N, 2.40.
Cya n om eth yl (4-Hyd r oxyp h en oxy)a ceta te. To (4-hy-
droxyphenoxy)acetic acid (1.68 g, 10.0 mmol) were added
chloroacetonitrile (10 mL) and diisopropylethylamine (1.74 mL,
10.0 mmol). The reaction mixture was stirred overnight. The
solvent was removed by rotary evaporation. Column chroma-
tography (50:50 ethyl acetate/hexanes) gave 1.10 g (53%) of
the desired product as a white solid: mp 90.5-92 °C. TLC:
N-Bp oc-4-(d im eth yl(4-h yd r oxybu tyl)silyl)-2-(tr im eth -
ylsta n n yl)a n ilin e (4). Stannane 3 (1.23 g, 2.03 mmol) and
9-BBN (6 mL, 0.5 M in THF) were added to an oven-dried
Schlenk flask under nitrogen. The reaction mixture was
stirred until the starting material was completely consumed,
as determined by TLC analysis. The flask was cooled to 0 °C,
and 2.0 M aqueous KOH (5.00 mL, 10.0 mmol) was added,
followed by 5.00 mL of 30% aqueous H2O2. After 1 min the
mixture was taken up in 50 mL of ethyl acetate, washed with
water (3 × 50 mL) and brine (50 mL), dried, and concentrated.
Column chromatography (25:75 ethyl acetate/hexanes) gave
0.91 g (72%) of the desired compound as a clear oil. TLC: Rf
Rf ) 0.33. IR: 1771, 1510 cm-1
.
1H NMR: δ 4.68 (s, 2H),
4.92 (s, 2H), 6.71 (d, 2H, J ) 9.1), 6.79 (d, 2H, J ) 9.1). 13C
NMR: δ 48.4, 65.3, 114.4, 115.4, 115.6, 151.0, 151.8, 168.5.
Anal. Calcd for C10H9NO4: C, 57.97; H, 4.38; N, 6.76, found
C, 58.18; H, 4.54; N, 6.77.
N-Bp oc-4-br om oa n ilin e. To a Schlenk flask under nitro-
gen were added 2-(4-biphenylyl)isopropyl phenyl carbonate
(14.95 g, 45.0 mmol), 4-bromoaniline (1) (8.60 g, 50.0 mmol),
and tetrahydrofuran (200 mL). Potassium hydride (35%, 18.0
g, 150 mmol) was then slowly added. The reaction mixture
was stirred for 15 min, then carefully quenched with water,
and diluted with ethyl acetate (300 mL). The organic layer
was washed with water (3 × 200 mL) and brine (200 mL),
dried, and concentrated. Column chromatography (10:90 ethyl
acetate/hexanes) gave 14.68 g (80%) of the desired product as
white crystals: mp 152-153 °C. TLC: Rf ) 0.23. IR: 1716,
) 0.23. IR: 3300 (b), 2927, 1712, 1487, 1245, 1140 cm-1 1H
.
NMR: δ 0.23 (s, 6H), 0.38 (s, 9H), 0.73 (m, 2H), 1.38 (m, 2H),
1.57 (m, 2H), 1.88 (s, 6H), 3.60 (m, 2H), 6.49 (s, broad, 1H),
7.33 (t, 1H, J ) 7.4), 7.40-7.45 (m, 3H), 7.46-7.49 (m, 3H),
7.53-7.59 (m, 5H). 13C NMR: δ -8.7, -3.1, 15.5, 20.0, 28.8,
36.5, 62.6, 81.5, 121.5 (b), 124.8, 127.0, 127.0, 127.1, 128.6,
134.8, 134.8, 139.8, 140.8, 141.6, 143.7, 144.9, 152.7. Anal.
Calcd for C31H43NO3SiSn: C, 59.63; H, 6.94; N, 2.24, found C,
59.62; H, 7.08; N, 2.25.
1634 cm-1
.
1H NMR: δ 1.89 (s, 6H), 6.74 (s, broad, 1H), 7.26
Cya n om eth yl Ester 5. To an oven-dried Schlenk flask
under nitrogen were added the cyanomethyl ester of (4-
hydroxyphenoxy)acetic acid (269 mg, 1.30 mmol), primary
alcohol 4 (740 mg, 1.18 mmol), triphenylphosphine (341 mg,
1.30 mmol), and tetrahydrofuran (7 mL). The flask was cooled
to 0 °C. Diethyl azodicarboxylate (197 µL, 1.25 mmol) was
added dropwise, and the ice bath was removed. After 1 h the
reaction mixture was concentrated, and column chromatog-
raphy (25:75 ethyl acetate/hexanes) gave 0.82 g (85%) of the
desired product as a clear oil. TLC: Rf ) 0.27. IR: 2932, 1780,
(d, 2H, J ) 9.2), 7.33-7.39 (m, 3H), 7.45 (t, 2H, J ) 7.7), 7.50
(d, 2H, J ) 8.4), 7.57-7.62 (m, 4H). 13C NMR: δ 28.8, 81.9,
115.5, 119.9, 124.7, 127.0, 127.1, 127.2, 128.6, 131.8, 137.1,
139.9, 140, 144.6, 151.7. Anal. Calcd for C22H20NO2Br: C,
64.40; H, 4.91; N, 3.41, found C, 64.60; H, 5.00; N, 3.44.
N-Bp oc-4-(3-bu ten yld im eth ylsilyl)a n ilin e (2). To an
oven-dried Schlenk flask under nitrogen were added N-Bpoc-
4-bromoaniline (3.40 g, 8.29 mmol) and tetrahydrofuran (40
mL). Potassium hydride (35%, 1.20 g, 10.0 mmol) was then
added slowly. The reaction mixture was cooled to -78 °C, and
tert-butyllithium (1.7 M in pentane, 10.0 mL, 17.0 mmol) was
added (the mixture turned red). After 10 min, 3-butenyldi-
methylchlorosilane (2.97 g, 20.0 mmol) was added in one
portion. The reaction mixture was carefully quenched with
water and diluted with ethyl acetate (100 mL). The organic
layer was washed with water (3 × 100 mL), 1 M NaHSO4 (2
× 100 mL), and brine (100 mL), dried, and concentrated.
Column chromatography (5:95 ethyl acetate/hexanes) gave
2.39 g (65%) of the desired product as white crystals: mp 104-
105 °C. TLC: Rf ) 0.16. IR: 1708, 1588, 1517, 1141, 1102
1718, 1507, 1242 cm-1 1H NMR: δ 0.28 (s, 6H), 0.40 (s, 9H),
.
0.83 (m, 2H), 1.78 (m, 2H), 1.90 (s, 6H), 3.85 (t, 2H, J ) 6.7),
6.53 (s, broad, 1H), 6.81 (d, 2H, J ) 9.2), 6.84 (d, 2H, J ) 9.2),
7.35 (t, 1H, J ) 7.3), 7.42-7.47 (m, 3H), 7.48-7.53 (m, 3H),
7.56-7.62 (m, 3H). 13C NMR: δ -8.6, -3.0, 15.5, 20.4, 28.8,
32.9, 48.6, 65.8, 68.1, 81.5, 113.7, 115.5, 115.9, 121.5 (b), 124.8,
127.0, 127.0, 127.1, 128.6, 134.7, 134.8, 139.8, 140.8, 141.6,
143.7, 144.9, 151.4, 152.7, 154.4, 167.8. Anal. Calcd for
C41H50N2O6SiSn: C, 60.53; H, 6.19; N, 3.44, found C, 60.85;
H, 6.41; N, 3.40.
2-(4-Meth ylph en yl)isopr opyl P h en yl Car bon ate. Dichlo-
romethane (200 mL), 2-(4-methylphenyl)isopropyl alcohol (30.0
g, 200 mmol), and pyridine (40.5 mL, 500 mmol) were
combined in a round bottom flask. The reaction mixture was
cooled to -78 °C, and phenyl chloroformate (26.4 mL, 210
mmol) was added in one portion. Large amounts of white
solids immediately formed. The mixture was stirred for 3 h
at room temperature and then quenched with water (200 mL).
The organic layer was washed with 1 M NaHSO4 (3 × 200
mL) and 1 M NaOH (3 × 200 mL), dried, and concentrated.
The mixed carbonate was used without further purification.
N-Mp c-4-br om oa n ilin e. To a Schlenk flask under nitro-
gen were added 2-(4-methylphenyl)isopropyl phenyl carbonate
(14.0 g, 50.0 mmol), 4-bromoaniline (1) (8.60 g, 50.0 mmol),
and tetrahydrofuran (100 mL). The reaction mixture was
cooled to 0 °C, and potassium hydride (35%, 13.2 g, 110 mmol)
was added slowly. The reaction mixture was stirred for 30
min, then carefully quenched with water, and diluted with
ethyl acetate (150 mL). The organic layer was washed with
water (3 × 200 mL) and brine (200 mL), dried, and concen-
trated. Column chromatography (10:90 ethyl acetate/hexanes)
gave 14.7 g (84%) of the desired product as white crystals: mp
151-152 °C. TLC (5:95 ethyl acetate/hexanes): Rf ) 0.10.
cm-1 1H NMR: δ 0.22 (s, 6H), 0.80 (m, 2H), 1.86 (s, 6H), 2.02
.
(m, 2H), 4.85 (dd, 1H, J ) 1.7, 10.1), 4.95 (dd, 1H, J ) 1.7,
17.1), 6.52 (tdd, 1H, J ) 6.2, 10.1, 17.1), 6.63 (s, broad, 1H),
7.30-7.36 (m, 3H), 7.37-7.43 (m, 4H), 7.50 (d, 2H, J ) 8.3),
7.57-7.62 (m, 4H). 13C NMR: δ -3.0, 14.8, 27.8, 28.9, 81.6,
112.7, 117.7, 124.7, 127.0, 127.1, 127.1, 128.6, 133.2, 134.3,
138.7, 139.9, 140.7, 141.4, 144.8, 151.8. Anal. Calcd for
C28H33NO2Si: C, 75.80; H, 7.50; N, 3.16, found C, 75.66; H,
7.43; N, 3.15.
N-Bp oc-4-(3-b u t en yld im et h ylsilyl)-2-(t r im et h ylst a n -
n yl)a n ilin e (3). Bpoc-protected aniline 2 (2.45 g, 5.52 mmol)
and diethyl ether (12 mL) were added to an oven-dried Schlenk
flask under nitrogen. The flask was cooled to -78 °C and
stirred vigorously, and n-butyllithium (2.50 mL, 6.25 mmol)
was added, followed by tert-butyllithium (4.00 mL, 6.80 mmol).
The reaction mixture was allowed to warm to -10 °C (main-
tained by a salt bath); the mixture slowly turned red. After 2
h the mixture was cooled to -78 °C and trimethyltin chloride
(10.0 mL, 1.0 M in hexane) was added; the solution rapidly
turned a cloudy yellow color. After reaction mixture was
warmed to room temperature, it was quenched with 50 mL of
water, diluted with 50 mL of ethyl acetate, washed with water
(3 × 50 mL) and brine (100 mL), dried, and concentrated.
Column chromatography (5:95 ethyl acetate/hexanes) gave
1.43 g (43%) of the desired product as a clear oil. TLC: Rf )
IR: 1701, 1526, 1230, 1145 cm-1 1H NMR (400 MHz): δ 1.84
.
(s, 6H), 2.35 (s, 3H), 6.71 (s, broad, 1H), 7.18 (d, 2H, J ) 8.1),
7.24 (d, 2H, J ) 8.8), 7.32 (d, 2H, J ) 8.1), 7.36 (d, 2H, J )
0.25. IR: 1717, 1486, 1270 cm-1
.
1H NMR: δ 0.27 (s, 6H),
0.41 (s, 9H), 0.85 (m, 2H), 1.90 (s, 6H), 2.09 (m, 2H), 4.90 (dd,
1H, J ) 1.7, 7.4), 5.00 (dd, 1H, J ) 1.7, 17.0), 5.89 (tdd, 1H, J
) 6.2, 10.1, 17.0), 6.52 (s, broad, 1H), 7.35 (t, 1H, J ) 7.4),
(38) For compounds 3-5 and 16-18 one of the expected carbon
peaks was not detected, even with very long acquisition times. This is
presumably because two carbon atoms have nearly the same resonance
frequency.
7.42-7.47 (m, 3H), 7.49-7.53 (m, 3H), 7.57-7.61 (m, 5H). 13
C

Germanium and Silicon Linking Strategies
J . Org. Chem., Vol. 62, No. 9, 1997 2891
8.8). 13C NMR (101 MHz): δ 21.0, 28.8, 82.0, 115.4, 119.9,
124.1, 129.0, 131.7, 136.7, 137.2, 142.6, 151.8. Anal. Calcd
for C17H18NO2Br: C, 58.63; H, 5.21; N, 4.02, found C, 58.70;
H, 5.17; N, 4.06.
the mixture was cooled to -78 °C and trimethyltin chloride
(2.0 mL, 1.0 M in hexane) was added in one portion. The
mixture rapidly turned a cloudy yellow color and was allowed
to warm to room temperature. The reaction mixture was
quenched with 25 mL of water, diluted with 50 mL of ethyl
acetate, washed with water (3 × 50 mL) and brine (50 mL),
dried, and concentrated. Column chromatography (5:95 ethyl
acetate/hexanes) gave 0.56 g (50%) of the desired product as
a clear oil. TLC: Rf ) 0.30. 1H NMR: δ 0.37 (s, 6H), 0.41 (s,
9H), 0.64 (q, 6H, J ) 8.0), 0.95 (m, 2H), 1.00 (t, 9H, J ) 8.0),
1.30-1.45 (m, 6H), 1.51-1.56 (m, 2H), 1.87 (s, 6H), 2.37 (s,
3H), 6.48 (s, broad, 1H), 7.18 (d, 2H, J ) 8.0), 7.34 (d, 2H, J )
8.0), 7.40 (d, 1H, J ) 8.0), 7.47 (s, 1H), 7.58 (d, 1H, J ) 8.0).
13C NMR: δ -8.7, -3.7, 4.4, 6.7, 16.0, 20.9, 24.9, 25.4, 28.8,
32.8, 33.0, 62.9, 51.6, 122.0 (b), 124.2, 128.9, 134.3, 136.5,
137.3, 141.0, 143.0, 143.2, 152.8. Anal. Calcd for
C34H59GeNOSiSn: C, 54.51; H, 7.94; N, 1.81, found C, 54.67;
H, 7.89; N, 1.86.
6-Ch lor o-1-((tr ieth ylsilyl)oxy)h exa n e. To a round bot-
tom flask were added CH2Cl2 (150 mL), 6-chloro-1-hexanol
(25.0 g, 183 mmol), imidazole (34.0 g, 500 mmol), and triethyl-
silyl chloride (28.63 g, 190 mmol). The reaction mixture was
stirred for 10 min, then washed with water (3×), dried, and
concentrated. Vacuum distillation (0.1 mmHg, 80-88 °C) gave
41.2 g (90%) of the desired product as a clear oil. IR: 1458,
1238, 1099 cm-1 1H NMR (400 MHz): δ 0.59 (q, 6H, J ) 7.9),
.
0.95 (t, 9H, J ) 7.9), 1.33-1.57 (m, 6H), 1.77 (quintet, 2H, J
) 7.0), 3.52 (t, 2H, J ) 6.7), 3.60 (t, 2H, J ) 6.5). 13C NMR
(101 MHz): δ 4.4, 6.7, 25.1, 26.7, 32.6, 32.7, 44.9, 62.6. Anal.
Calcd for C12H27ClOSi: C, 57.45; H, 10.85, found C, 57.29; H,
10.68.
C h lo r o d i m e t h y l(6-((t r i e t h y ls i ly l)o x y )h e x y l)g e r -
m a n e.39 To a flame-dried Schlenk flask under nitrogen were
added 1-chloro-6-((triethylsilyl)oxy)hexane (18.1 g, 72.0 mmol),
magnesium turnings (2.44 g, 100 mmol), and tetrahydrofuran
(60 mL). A drop of 1,2-dibromoethane was added to initiate
the reaction, and the reaction mixture was heated to reflux.
Heating was continued for 1 h, and then the flask was cooled
to 0 °C. In a separate flame-dried Schlenk flask THF (30 mL)
and dichlorodimethylgermane (10.0 g, 57.6 mmol) were com-
bined and cooled to -78 °C. The Grignard reagent was
transferred to the germanium flask by cannula; gray mag-
nesium salts precipitated out. The mixture was allowed to
warm to ambient temperature and stirred for 1 h. The mixture
was concentrated by rotary evaporation and then taken up in
hexane (40 mL). The salts were removed by vacuum filtration,
and the organic layer was reconcentrated. Vacuum distillation
(0.1 mmHg, 110-130 °C) gave 14.2 g (70%) of the desired
compound as a clear oil. 1H NMR (400 MHz): δ 0.60 (q, 6H,
J ) 7.9), 0.68 (s, 6H), 0.96 (t, 9H, J ) 7.9), 1.18 (m, 2H), 1.34-
1.40 (m, 4H), 1.47-1.56 (m, 4H), 3.60 (t, 2H, J ) 6.6). 13C
NMR (101 MHz): δ 3.3, 4.4, 6.8, 21.6, 23.9, 25.4, 32.3, 32.7,
62.8.
N-Mp c-4-(d im eth yl(6-h yd r oxyh exyl)ger m a n yl)-2-(tr i-
m eth ylsta n n yl)a n ilin e (17). Silyl ether 16 (3.90 g, 5.21
mmol) was treated with 8 mL of 1.0 M Bu4NF in THF for 5
min. The reaction mixture was diluted with 50 mL of ethyl
acetate, washed with water (3×) and brine, dried, and con-
centrated. Column chromatography (20:80 ethyl acetate/
hexanes) gave 3.14 g (95%) of the desired compound as a clear
oil. TLC (25:75 ethyl acetate/hexanes): Rf ) 0.21. IR (film
from CH2Cl2): 1705, 1504, 1236, 1140 cm-1
.
1H NMR (400
MHz): δ 0.37 (s, 6H), 0.41 (s, 9H), 0.95 (m, 2H), 1.30-1.36
(m, 4H), 1.40 (m, 2H), 1.46 (t, 1H, J ) 5.3), 1.48-1.54 (m, 2H),
1.83 (s, 6H), 2.33 (s, 3H), 3.58 (dt, 2H, J ) 5.3, 6.5), 6.45 (s,
broad, 1H), 7.15 (d, 1H, J ) 8.0), 7.31 (d, 2H, J ) 8.0), 7.36 (d,
1H, J ) 7.9), 7.44 (s, 1H), 7.52 (d, 1H, J ) 7.9). 13C NMR
(101 MHz): δ -8.7, -3.7, 15.9, 20.9, 24.9, 25.2, 28.8, 32.6, 32.8,
62.9, 81.6, 122.1 (b), 124.2, 128.9, 134.3, 136.5, 137.3, 141.0,
143.0, 143.1, 152.9. Anal. Calcd. for C28H45GeNO3Sn: C,
52.97; H, 7.14; N, 2.21, found C, 52.95; H, 6.90; N, 2.14.
Cya n om eth yl Ester 18. To an oven-dried Schlenk flask
under nitrogen were added the cyanomethyl ester of (4-
hydroxyphenoxy)acetic acid (1.57 g, 7.60 mmol), primary
alcohol 17 (4.40 g, 6.93 mmol), triphenylphosphine (2.00 g, 7.60
mmol), and tetrahydrofuran (30 mL). The flask was cooled to
0 °C. Diethyl azodicarboxylate (1.15 mL, 7.30 mmol) was
added dropwise, and the ice bath was removed. After 1 h, the
reaction mixture was concentrated. Column chromatography
(50:50 CH2Cl2/hexanes) gave 4.22 g (74%) of the desired
product as a clear oil. TLC (25:75 ethyl acetate/hexanes): Rf
) 0.21. 1H NMR: δ 0.31 (s, 6H), 0.36 (s, 9H), 0.88-0.93 (m,
2H), 1.25-1.29 (m, 2H), 1.34-1.43 (m, 4H), 1.68-1.74 (m, 2H),
1.81 (s, 6H), 2.31 (s, 3H), 3.87 (t, 2H, J ) 6.6), 4.67 (s, 2H),
4.82 (s, 2H), 6.41 (s, broad, 1H), 6.81 (d, 2H, J ) 9.0), 6.84 (d,
2H, J ) 9.0), 7.13 (d, 2H, J ) 8.0), 7.29 (d, 2H, J ) 8.0), 7.34
(d, 1H, J ) 8.0), 7.41 (s, 1H), 7.52 (d, 1H, J ) 8.0). 13C NMR:
δ -8.7, -3.7, 15.9, 20.9, 24.9, 25.5, 28.8, 29.1, 32.9, 48.6, 65.8,
68.4, 81.6, 113.6, 115.4, 115.9, 122.0, 124.2, 128.9, 134.3, 136.5,
137.3, 141.0, 143.0, 143.2, 151.4, 152.8, 154.4, 167.8. Anal.
Calcd for C38H52GeN2O6Sn: C, 55.38; H, 6.36; N, 3.40, found
C, 55.26; H, 6.37; N, 3.19.
Cou p lin g Cya n om eth yl Ester 5 to th e (Am in om eth yl)-
p olystyr en e Resin . (The procedure for ester 18 is identical.)
To an oven-dried Schlenk flask under nitrogen were added the
activated ester 5 (0.72 g, 0.88 mmol), 4-(dimethylamino)-
pyridine (110 mg, 0.90 mmol), diisopropylethylamine (235 µL,
1.35 mmol), N-methylpyrrolidinone (8.0 mL), and (amino-
methyl)polystyrene resin (1.20 g, 0.624 mmol). The reaction
mixture was heated at 70 °C for 12 h, at which point the
Ninhydrin test indicated that the reaction had gone to >85%
completion. The resin was transferred to a peptide flask,
rinsed with ethyl acetate (5×) and CH2Cl2 (5×), and then dried
under vacuum. The resin 6a was calculated to have a loading
of 0.37 mmol/g, assuming the coupling reaction went to
completion. The resin 6b was calculated to have a loading of
0.46 mmol/g, assuming the coupling reaction went to comple-
tion. A description of the procedure used to determine the
percent completion of the acylation is contained in the Sup-
porting Information.
N -Mp c-4-(d im e t h yl(6-((t r ie t h ylsilyl)oxy)h e xyl)ge r -
m a n yl)a n ilin e (15). To an oven-dried Schlenk flask under
nitrogen were added Mpc-protected 4-bromoaniline (11.7 g,
33.5 mmol) and tetrahydrofuran (120 mL). The flask was
cooled to 0 °C, and KH (35%, 35 mmol, 4.2 g) was added
dropwise. The flask was cooled to -78 °C and stirred vigor-
ously, and tert-butyllithium (39.4 mmol) was added slowly by
syringe. After 2 h chlorodimethyl(6-((triethylsilyl)oxy)hexyl)-
germane (13.1 g, 37.0 mmol) was added dropwise. After the
reaction mixture was warmed to ambient temperature, the
reaction mixture was quenched with 50 mL of water, diluted
with 100 mL of ethyl acetate, washed with water (3 × 100
mL) and brine (100 mL), dried, and concentrated. Column
chromatography (eluting with 5:95 ethyl acetate/hexanes) gave
11.9 g (60%) of the desired product as a waxy solid: mp 43-
44 °C. TLC: Rf ) 0.21. IR: 1705, 1593, 1516, 1235, 1141
cm-1 1H NMR: δ 0.32 (s, 6H), 0.60 (q, 6H, J ) 8.0), 0.90 (m,
.
2H), 0.97 (t, 9H, J ) 8.0), 1.26-1.43 (m, 6H), 1.48-1.54 (m,
2H), 1.83 (s, 6H), 2.34 (s, 3H), 3.58 (t, 2H, J ) 6.8), 6.61 (s,
broad, 1H), 7.16 (d, 2H, J ) 8.0), 7.30-7.34 (m, 6H). 13C
NMR: δ -3.7, 4.4, 6.8, 16.0, 21.0, 25.0, 25.4, 28.9, 32.8, 33.0,
63.0, 81.6, 118.0, 124.2, 129.0, 133.8, 135.7, 136.6, 138.2, 142.9,
151.8. Anal. Calcd for C31H51GeNO3Si: C, 63.50; H, 8.77; N,
2.39, found C, 63.69; H, 9.00; N, 2.52.
N -Mp c-4-(d im e t h yl(6-((t r ie t h ylsilyl)oxy)h e xyl)ge r -
m a n yl)-2-(tr im eth ylsta n n yl)a n ilin e (16). To an oven-dried
Schlenk flask under nitrogen were added Mpc-protected
aniline 15 (0.88 g, 1.50 mmol) and diethyl ether (4.0 mL). The
flask was cooled to -78 °C and stirred vigorously, and
n-butyllithium (0.70 mL, 1.8 mmol) was added slowly by
syringe. tert-Butyllithium (1.10 mL, 1.87 mmol) was then
added slowly by syringe. The reaction mixture was allowed
to warm to -10 °C (maintained by a salt bath); the mixture
slowly turned the bright red color of the dianion. After 2 h
(39) Our procedure is based on the synthesis for ω-alkenyldimeth-
ylchlorogermanes. Mochida, K.; Asami, K. J . Organomet. Chem. 1982,
232, 13.
Gen er a l P r oced u r e for Stille Cou p lin gs on th e Solid
Su p p or t. The procedure for a typical experiment follows. To

2892 J . Org. Chem., Vol. 62, No. 9, 1997
Plunkett and Ellman
an oven-dried Schlenk flask under nitrogen were added the
resin 6 (0.100 mmol), Pd2dba3‚CHCl3 (30 mg, 0.030 mmol),
K2CO3 (10 mg), THF (2.0 mL), and diisopropylethylamine (20
µL, 0.10 mmol). The resin was stirred for 3 min to ensure
complete solvation, at which point the acid chloride (0.50
mmol) was added slowly. The reaction was allowed to proceed
for 1 h, after which the mixture was transferred to a peptide
flask and rinsed with CH2Cl2 (3×), DMSO saturated with KCN
(1×), methanol (1×), water (1×), methanol (3×), and CH2Cl2
(3×). For silyl-linked derivatives 7a the resin was treated with
97:3 CH2Cl2/trifluoroacetic acid for 5 min, then rinsed with
CH2Cl2 (5×) and methanol (5×), and dried under nitrogen to
provide support-bound 2-aminobenzophenone or 2-amino-
acetophenone 8a . For germanium-linked derivatives 7b the
resin was treated with 8 equiv of B-chlorocatecholborane and
4 equiv of i-Pr2EtN in CH2Cl2 (2 min, 0.05 M in borane). The
resin was rinsed with CH2Cl2 (5×), 10:1:1 CH2Cl2/methanol/
piperidine (1×), and methanol (5×) and dried under nitrogen
to give support-bound 2-aminobenzophenone or 2-amino-
acetophenone 8b.
Gen er a l P r oced u r e for 1,4-Ben zod ia zep in e Syn th esis.
Support-bound 2-aminoaryl ketone 8 was added to a 24/40
single-neck 50 mL round bottom flask with stir bar, followed
by CH2Cl2, 2,6-di-tert-butyl-4-methylpyridine (10 equiv), and
Fmoc-protected-amino acid fluoride (10 equiv, final concentra-
tion 0.2 M). After the resulting slurry was stirred for 15-24
h at ambient temperature, the solution was removed by
filtration cannula and the support-bound anilide was washed
with CH2Cl2 (3×) and DMF (3×). A mixture of 20% piperidine
in DMF (15 mL) was then added, and the slurry was stirred
for 20-30 min at room temperature. The solvent was removed
by filtration cannula, and the resin was rinsed with DMF (3×)
and with CH2Cl2 (3×) to give 9. At this point, 25 mL of 5%
acetic acid in DMF was added to the resin and the slurry was
stirred at 60 °C for 12 h to give 10. The solvent was removed
by filtration cannula, and the resin was rinsed with DMF (3×)
and CH2Cl2 (5×). The reaction flask was sealed with a fresh
rubber septum and was flushed with N2 followed by cooling
to 0 °C. Freshly distilled THF and 5-(phenylmethyl)-2-
oxazolidinone (12 equiv, 0.2 M) were added to an oven-dried
Schlenk flask. The flask was cooled to -78 °C, and n-
butyllithium in hexanes (10 equiv) was added. After the
mixture was stirred for 10 min, the solution was transferred
by cannula to the resin. The mixture was stirred at 0 °C for
10 min, and then the appropriate alkylating agent was added
by syringe (15 equiv) followed by addition of anhydrous DMF
to give a final solvent ratio of approximately 70:30 THF/DMF.
The slurry was allowed to warm to room temperature with
stirring. After the slurry was stirred for 1 h, the solvent was
removed by filtration cannula. The support was washed with
THF (1×), 1:1 THF/water (2×), THF (2×), and CH2Cl2 (2×).
Ben zod ia zep in e P r otectin g Gr ou p Clea va ge. For de-
rivatives incorporating side-chain protecting groups, the fully
derivatized 1,4-benzodiazepine 11 on solid support was treated
with 15 mL of 95:5:10 trifluoroacetic acid/water/dimethyl
sulfide for 15-30 min. No cleavage of the support-bound
benzodiazepine was seen. The cleavage solution was removed
by filtration cannula, and the resin was rinsed with CH2Cl2
(5×) and methanol (5×) and dried under vacuum.
tography gave the purified derivatives 12a -d . NMR spectra
for 12a -d were obtained using a mixture of 20% CD3OD in
CDCl3 (v/v).
Ben zod ia zep in e 12a . This compound was synthesized
from 0.240 g (0.089 mmol) of resin 6a , 3-methoxybenzoyl
chloride, racemic leucine acid fluoride, and ethyl iodide.
Column chromatography (20:80 ethyl acetate/hexanes) gave
20.5 mg (66%) of the desired compound as beige crystals: mp
105-106 °C. TLC (50:50 ethyl acetate/hexanes): Rf ) 0.53.
IR: 1676, 1562 cm-1 1H NMR: δ 0.69 (d, 3H, J ) 6.5), 0.85
.
(d, 3H, J ) 6.5), 0.97 (t, 3H, J ) 7.1), 1.74-1.88 (m, 2H), 2.06-
2.15 (m, 1H), 3.47 (dd, 1H, J ) 5.2, 8.7), 3.63 (qd, 1H, J ) 7.1,
14.0), 3.70 (s, 3H), 4.18 (qd, 1H, J ) 7.1, 14.0), 6.88 (dd, 1H,
J ) 1.7, 8.2), 6.93 (d, 1H, J ) 7.9), 6.99 (t, 1H, J ) 1.7), 7.11
(t, 1H, J ) 7.5), 7.15-7.20 (m, 2H), 7.33 (d, 1H, J ) 8.2), 7.44
(t, 1H, J ) 7.8). 13C NMR: δ 13.0, 21.7, 22.9, 24.5, 39.5, 55.0,
61.5, 114.6, 115.8, 121.9, 122.0, 124.3, 129.0, 130.0, 131.4,
139.8, 141.8, 159.2, 168.7, 169.5. HRMS: calcd for
C22H27N2O2: 351.2073, found 351.2075.
Ben zod ia zep in e 12b. This compound was synthesized
from 0.240 g (0.089 mmol) of resin 6a , 1-adamantanecarbonyl
chloride, racemic leucine acid fluoride, and ethyl iodide.
Column chromatography (20:80 ethyl acetate/hexanes) gave
20.1 mg (60%) of the desired compound as white crystals: mp
87-88 °C. TLC: Rf ) 0.45. IR: 2903, 1678, 1608, 1446 cm-1
.
1H NMR: δ 0.58 (d, 3H, J ) 6.5), 0.74 (d, 3H, J ) 6.5), 0.89 (t,
3H, J ) 7.1), 1.52-1.60 (m, 6H), 1.61-1.90 (m, 12H), 3.25 (dd,
1H, J ) 5.2, 8.6), 3.45 (qd, 1H, J ) 7.1, 14.0), 4.18 (qd, 1H, J
) 7.1, 14.0), 7.07 (dt, 1H, J ) 1.1, 7.8), 7.18 (dd, 1H, J ) 1.1,
7.8), 7.30 (dt, 1H, J ) 1.5, 7.8), 7.46 (dd, 1H, J ) 1.5, 7.8). 13C
NMR: δ 12.9, 21.6, 22.8, 24.4, 28.4, 36.3, 39.2, 41.2, 41.3, 41.7,
60.6, 122.1, 123.6, 126.8, 129.6, 130.9, 140.3, 171.5, 175.7.
HRMS: calcd for C25H35N2O 379.2749, found 379.2757.
Ben zod ia zep in e 12c. This compound was synthesized
from 0.240 g (0.089 mmol) of resin 6a , benzoyl chloride,
L-tyrosine (O-t-Bu) acid fluoride, and benzyl bromide. Column
chromatography (50:50 ethyl acetate/hexanes) gave 19.4 mg
(50%) of the desired compound as ivory crystals: mp 120-
120.5 °C. TLC: Rf ) 0.41. IR: 1672, 1602, 1515, 1448 cm-1
.
1H NMR: δ 3.36 (dd, 1H, J ) 7.6, 14.0), 3.46 (dd, 1H, J ) 6.2,
14), 3.72 (dd, 1H, J ) 6.2, 7.6), 4.68 (d, 1H, J ) 15.3), 5.52 (d,
1H, J ) 15.3), 6.63 (d, 2H, J ) 8.5), 6.86 (d, 2H, J ) 6.5), 6.95-
7.02 (m, 5H), 7.06 (d, 2H, J ) 8.5), 7.13 (dd, 2H, J ) 1.4, 8.5),
7.21 (t, 2H, J ) 7.7), 7.30-7.39 (m, 3H). 13C NMR: δ 36.5,
50.0, 65.2, 114.8, 122.3, 124.4, 127.0, 127.1, 127.9, 128.3, 129.4,
129.7, 130.1, 130.2, 130.5, 130.5, 131.3, 136.3, 138.1, 141.1,
154.9, 169.4, 169.5. HRMS: calcd for C29H25N2O2 433.1916,
found 433.1916. An additional 3.5 mg (11%) of the debenzyl-
ated product was obtained: Rf ) 0.24.
Ben zod ia zep in e 12d . This compound was synthesized
from 0.200 g (0.074 mmol) of resin 6a , 4-methoxybenzoyl
chloride, and L-alanine acid fluoride. Column chromatography
(50:50 ethyl acetate/hexanes) gave 14.0 mg (68%) of the desired
compound as white crystals: mp 247-248.5 °C dec. TLC: Rf
) 0.36. IR: 1669, 1597, 1564 cm-1 1H NMR: δ 1.54 (d, 3H,
.
J ) 6.5), 3.56 (q, 1H, J ) 6.5), 3.70 (s, 3H), 6.76 (d, 2H, J )
8.7), 6.99-7.08 (m, 2H), 7.19 (dd, 1H, J ) 1.5, 8.0), 7.30 (d,
2H, J ) 8.7), 7.38 (dt, 1H, J ) 1.5, 7.7). 13C NMR: δ 16.1,
55.0, 58.1, 113.3, 120.7, 123.0, 127.6, 131.1, 131.1, 131.4, 131.5,
138.4, 161.3, 169.2, 172.7. HRMS: calcd for C17H17N2O2
281.1290, found 281.1282.
HF Clea va ge for Silyl-Lin k ed Ben zod ia zep in es. The
resin 11a was transferred to a perfluorinated plastic reaction
vessel for HF cleavage. (CAUTION!! Hydrogen fluoride gas
is EXTREMELY toxic, and reactions with this gas should only
be performed with the proper equipment and training. There
are two excellent reviews40 that address safety issues with
respect to the handling and use of this reagent.) Anhydrous
hydrogen fluoride gas (Matheson) was condensed in the
reaction vessel and allowed to react (under some pressure since
HF boils at 19.5 °C) for 12 h, at which point the hydrogen
fluoride was removed with gentle nitrogen flow. The vapor
was removed with two sequential KOH traps. The resin was
rinsed with 20 mL of 1:4 methanol/CH2Cl2 (5 × 2 min).
Concentration of the combined filtrates and column chroma-
TF A Clea va ge for Ger m a n iu m -Lin k ed Ben zo-
d ia zep in es. To the dried resin 11b was added 15 mL of
anhydrous trifluoroacetic acid. The resin was heated at 60
°C for 24 h. The cleavage solution was removed by filtration
cannula, and the resin was rinsed with CH2Cl2 (3×) and
methanol (3×). Concentration of the combined filtrates and
column chromatography gave the final products 19a -d .
Ben zod ia zep in e 19a . This compound was made from
3-methoxybenzoyl chloride, L-leucine acid fluoride, and ethyl
iodide. Cleavage of 118 mg (0.0567 mmol) of resin and then
column chromatography gave 13.4 mg (62%) of the desired
product as ivory crystals: mp 60-62 °C. TLC: Rf ) 0.28. IR:
1675, 1599, 1448 cm-1 1H NMR (400 MHz): δ 0.82 (d, 3H, J
.
) 6.3), 0.99 (d, 3H, J ) 6.5), 1.10 (t, 3H, J ) 7.1), 1.90-2.01
(m, 2H), 2.28-2.36 (m, 1H), 3.59 (dd, 1H, J ) 4.8, 9.0), 3.75
(dq, 1H, J ) 14.0, 7.1), 3.83 (s, 3H), 4.31 (dq, 1H, J ) 14.0,
(40) (a) Sakakibara, S. In Chemistry and Biochemistry of Amino
Acids, Peptides, and Proteins; Weinstar, B., Ed.; New York: 1971; p
51. (b) Lenard, J . Chem. Rev. 1969, 69, 625.

Germanium and Silicon Linking Strategies
J . Org. Chem., Vol. 62, No. 9, 1997 2893
7.1), 6.99 (ddd, 1H, J ) 0.9, 2.6, 8.2), 7.09 (dt, 1H, J ) 7.7,
1.2), 7.17-7.22 (m, 2H), 7.30 (d, 1H, J ) 7.9), 7.33 (dd, 1H, J
) 1.5, 7.8), 7.41 (d, 1H, J ) 8.3), 7.55 (ddd, 1H, J ) 1.7, 7.2,
8.3). 13C NMR (101 MHz): δ 13.4, 22.0, 23.4, 24.7, 40.0, 42.5,
55.2, 61.7, 114.7, 115.7, 122.0, 122.1, 124.0, 129.1, 130.0, 130.5,
131.1, 140.4, 142.2, 159.4, 167.8, 169.5. Anal. Calcd for
C22H26N2O2: C, 75.40; H, 7.48; N, 7.99, found C, 75.60; H, 7.23;
N, 7.72.
7.56 (dd, 1H, J ) 2.3, 8.8). 13C NMR (101 MHz): δ 22.2, 23.3,
24.8, 40.1, 49.7, 55.7, 55.9, 61.5, 110.2, 111.6, 117.2, 123.3,
124.0, 127.2, 127.4, 128.5, 130.9, 132.4, 132.6, 133.9, 136.5,
140.8, 148.6, 151.1, 166.4, 169.9. Anal. Calcd for C28H29N2O3-
Br: C, 64.49; H, 5.61; N, 5.37, found C, 64.24; H, 5.73; N, 5.17.
Ben zod ia zep in e 20b. This compound was made from
3-methoxybenzoyl chloride, L-phenylalanine acid fluoride, and
iodoacetamide. Cleavage of 155 mg (0.0725 mmol) of resin and
then column chromatography (100% ethyl acetate) gave 21.2
mg (59%) of the desired product as white crystals: mp 110-
112 °C. TLC: Rf ) 0.39. IR (film from chloroform): 1698,
Ben zod ia zep in e 19b. This compound was made from
1-adamantanecarbonyl chloride, L-aspartate (O-t-Bu) acid
fluoride, and ethyl iodide. Cleavage of 160 mg (0.0736 mmol)
of resin and then column chromatography (50:50:1 ethyl
acetate/hexanes/acetic acid) gave 18.6 mg (66%) of the desired
product as white crystals: mp 86-90 °C. TLC: Rf ) 0.20. IR
1682, 1479, 1325 cm-1 1H NMR (400 MHz): δ 3.51 (dd, 1H,
.
J ) 5.3, 13.9), 3.60 (dd, 1H, J ) 8.6, 13.9), 3.83 (s, 3H), 3.86
(dd, 1H, J ) 5.3, 8.6), 4.27 (d, 1H, J ) 15.8), 4.58 (d, 1H, J )
15.8), 5.76 (s, broad, 1H), 6.28 (s, broad, 1H), 7.01 (dd, 2H, J
) 2.0, 8.0), 7.14 (t, 1H, J ) 2.0), 7.21 (t, 1H, J ) 7.2), 7.26-
7.31 (m, 3H), 7.34 (d, 2H, J ) 7.0), 7.38 (d, 1H, J ) 2.3), 7.47
(d, 1H, J ) 8.8), 7.64 (dd, 1H, J ) 2.3, 8.8). 13C NMR (101
MHz): δ 37.6, 51.8, 55.2, 64.6, 114.3, 116.8, 117.8, 122.3, 124.0,
126.2, 128.1, 129.3, 129.8, 130.9, 132.6, 134.8, 138.6, 139.3,
141.4, 159.5, 167.4, 170.0, 170.3. HRMS: calcd for C25H23N3O3-
Br 492.0910, found 492.0923.
(film from chloroform): 2904, 1710, 1674, 1448 cm-1 1H NMR
.
(400 MHz): δ 0.87 (t, 3H, J ) 7.1), 1.69 (d, 3H, J ) 12.2), 1.75
(d, 3H, J ) 12.2), 1.86 (d, 3H, J ) 12.2), 1.98 (d, 3H, J ) 12.2),
2.07 (s, 3H), 2.82 (dd, 1H, J ) 5.7, 16.1), 2.98 (dd, 1H, J ) 3.3,
16.1), 3.62 (dq, 1H, J ) 14.0, 7.1), 3.82 (dd, 1H, J ) 3.3, 5.7),
4.39 (dq, 1H, J ) 14.0, 7.1), 7.28 (dd, 1H, J ) 7.1, 7.9), 7.37
(d, 1H, J ) 7.9), 7.52 (dt, 1H, J ) 1.4, 7.1), 7.69 (dd, 1H, J )
1.4, 7.9). 13C NMR (101 MHz): δ 13.2, 28.3, 34.8, 36.2, 41.1,
41.6, 42.6, 58.8, 122.7, 124.3, 127.3, 129.8, 130.9, 140.4, 167.6,
172.6, 178.9. Anal. Calcd for C23H28N2O3: C, 72.61; H, 7.42;
N, 7.36, found C, 72.64; H, 7.33; N, 7.08.
Ben zod ia zep in e 20c. This compound was made from
3-methoxybenzoyl chloride, L-aspartate (O-t-Bu) acid fluoride,
and ethyl iodide. Cleavage of 176 mg (0.0828 mmol) of resin
and then column chromatography (25:75 ethyl acetate/hex-
anes) gave 22.7 mg (56%) of the desired product as white
crystals: mp 49-51 °C. TLC: Rf ) 0.22. IR: 1720, 1673, 1156
Ben zod ia zep in e 19c. This compound was made from
2-naphthoyl chloride, L-leucine acid fluoride, and benzyl
bromide. Cleavage of 128 mg (0.0589 mmol) of resin and then
column chromatography (25:75 ethyl acetate/hexanes) gave
17.4 mg (68%) of the desired product as white crystals: mp
cm-1
.
1H NMR (400 MHz): δ 1.10 (t, 3H, J ) 7.1), 1.46 (s,
9H), 3.13 (dd, 1H, J ) 6.7, 17.0), 3.35 (dd, 1H, J ) 7.3, 17.0),
3.71 (dq, 1H, J ) 14.0, 7.0), 3.84 (s, 3H), 4.06 (dd, 1H, J ) 6.7,
7.3), 4.30 (dq, 1H, J ) 14.0, 7.0), 7.00-7.05 (m, 2H), 7.17 (m,
1H), 7.29-7.34 (m, 2H), 7.45 (d, 1H, J ) 2.3), 7.65 (dd, 1H, J
) 2.3, 8.8). 13C NMR (101 MHz): δ 13.4, 28.1, 37.8, 42.7, 55.3,
60.7, 80.6, 114.4, 116.6, 117.3, 122.2, 123.9, 129.4, 132.1, 132.6,
134.4, 139.5, 141.2, 159.7, 167.1, 168.4, 171.3. Anal. Calcd
for C24H27N2O4Br: C, 59.14; H, 5.58; N, 5.75, found C, 58.81;
H, 5.61; N, 5.57.
60-62 °C. TLC: Rf ) 0.39. IR: 1682, 1603, 1449 cm-1 1H
.
NMR (400 MHz): δ 0.88 (d, 3H, J ) 6.3), 1.06 (d, 3H, J )
6.3), 2.02-2.14 (m, 2H), 2.39-2.46 (m, 1H), 3.80 (dd, 1H, J )
4.9, 8.9), 4.77 (d, 1H, J ) 15.3), 5.76 (d, 1H, J ) 15.3), 7.02-
7.17 (m, 6H), 7.25 (d, 1H, J ) 8.2), 7.46-7.58 (m, 5H), 7.67
(dd, 1H, J ) 1.5, 8.5), 7.74 (d, 1H, J ) 7.8), 7.83 (d, 1H, J )
8.5), 7.87 (d, 1H, J ) 7.9). 13C NMR (101 MHz): δ 22.1, 23.4,
24.7, 40.2, 50.0, 61.8, 122.5, 124.4, 126.2, 126.3, 127.0, 127.1,
127.5, 127.6, 127.8, 128.4, 128.6, 129.9, 130.1, 131.1, 131.2,
132.5, 134.1, 136.2, 136.9, 141.8, 168.5, 170.1. Anal. Calcd
for C22H26N2O2: C, 83.30; H, 6.52; N, 6.48, found C, 83.29; H,
6.72; N, 6.25.
Ben zod ia zep in e 20d . This compound was made from
piperonyloyl chloride, L-tyrosine (O-t-Bu) acid fluoride, and 3,5-
dimethylbenzyl bromide. Cleavage of 129 mg (0.0568 mmol)
of resin and then column chromatography (35:65 ethyl
acetate/hexanes): gave 16.0 mg of the desired product (47%)
as pale yellow crystals: mp 110-112 °C. TLC (50:50 ethyl
acetate/hexanes) Rf ) 0.35. IR: 3400 (b), 1673, 1443, 1261,
Ben zod ia zep in e 19d . This compound was made from
4-methoxybenzoyl chloride, L-phenylalanine acid fluoride, and
ethyl iodide. Cleavage of 176 mg (0.0828 mmol) of resin and
then column chromatography (25:75 ethyl acetate/hexanes)
gave 18.4 mg (58%) of the desired product as pale yellow
crystals: mp 174-176 °C. TLC: Rf ) 0.21. IR: 1673, 1600,
1238 cm-1 1H NMR (400 MHz): δ 2.07 (s, 6H), 3.50 (dd, 1H,
.
J ) 7.7, 13.9), 3.59 (dd, 1H, J ) 6.1, 13.9), 3.76 (dd, 1H, J )
6.1, 7.7), 4.59 (d, 1H, J ) 15.5), 4.76 (s, broad, 1H), 5.63 (d,
1H, J ) 15.5), 6.02 (d, 1H, J ) 1.3), 6.03 (d, 1H, J ) 1.3), 6.54
(s, 2H), 6.74-6.78 (m, 5H), 6.98 (s, 1H), 7.21-7.25 (m, 3H),
7.31 (d, 1H, J ) 2,4), 7.51 (dd, 1H, J ) 2.4, 8.8). 13C NMR
(101 MHz): δ 20.9, 37.0, 49.6, 65.3, 101.4, 107.8, 109.3, 115.0,
117.2, 124.0, 124.7, 124.9, 129.0, 130.9, 131.2, 132.2, 132.3,
132.4, 133.9, 136.1, 138.1, 140.7, 148.0, 149.6, 153.9, 166.5,
169.4. Anal. Calcd for C31H27BrN2O4‚H2O: C, 63.90; H, 4.86;
N, 4.66, found C, 63.54; H, 5.22; N, 4.34.
1510, 1251 cm-1 1H NMR (400 MHz): δ 1.10 (t, 3H, J ) 7.1),
.
3.57-3.61 (m, 2H), 3.70-3.78 (m, 2H), 4.33 (dq, 1H, J ) 14.0,
7.0), 6.90 (d, 2H, J ) 8.9), 7.14-7.21 (m, 2H), 7.25-7.30 (m,
3H), 7.34-7.40 (m, 3H), 7.49-7.54 (m, 3H). 13C NMR (101
MHz): δ 13.4, 38.0, 42.6, 55.4, 65.0, 113.6, 122.2, 124.1, 126.0,
128.1, 129.9, 130.3, 130.7, 131.0, 131.1, 131.7, 139.7, 142.2,
161.4, 167.4, 169.3. Anal. Calcd for C25H24N2O2: C, 78.10;
H, 6.29; N, 7.29, found C, 77.93; H, 6.40; N, 6.97.
Br om in e Clea va ge for Ger m a n iu m -Lin k ed Ben zod i-
a zep in es. To the dried resin were added 15 mL of CH2Cl2
and 4 equiv of Br2. The mixture was stirred for 5 min, then
the cleavage solution was removed by filtration cannula, and
the resin was rinsed with CH2Cl2 (3×) and methanol (3×). The
procedure was repeated once to ensure that all material had
been cleaved from the support. Concentration of the combined
filtrates and column chromatography gave the final products
20a -d .
Ack n ow led gm en t. We thank David Oare (Genen-
tech) and Prof. Peter Vollhardt for helpful discussions.
We gratefully acknowledge the gift of an HF system
from Paul Hoeprich (Argonaut). The NIH (GM50353)
and the Burroughs Wellcome Fund provided support for
this research. Reaction database searches were per-
formed with ISIS (from MDL, Inc., 14600 Catalina St.,
San Leandro, CA 94577).
Ben zod ia zep in e 20a . This compound was made from 3,4-
dimethoxybenzoyl chloride, L-leucine acid fluoride, and benzyl
bromide. Cleavage of 146 mg (0.0671 mmol) of resin and then
column chromatography (25:75 ethyl acetate/hexanes) gave
19.1 mg (55%) of the desired product as pale yellow crystals:
Su p p or tin g In for m a tion Ava ila ble: Procedures for the
determination of resin amine substitution level and yields of
cyanomethyl ester acylations to give 6 (3 pages). This material
is contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
mp 73-74 °C. TLC: Rf ) 0.18. IR: 1683, 1269 cm-1 1H NMR
.
(400 MHz): δ 0.87 (d, 3H, J ) 6.6), 1.02 (d, 3H, J ) 6.6), 1.94
(m, 1H), 2.07 (ddd, 1H, J ) 5.3, 8.0, 13.7), 2.33 (ddd, 1H, J )
5.9, 8.7, 13.7), 3.68 (dd, 1H, J ) 5.3, 8.7), 3.85 (s, 3H), 3.93 (s,
3H), 4.69 (d, 1H, J ) 15.4), 5.70 (d, 1H, J ) 15.4), 6.75 (dd,
1H, J ) 1.9, 8.4), 6.80 (d, 1H, J ) 8.4), 6.97-7.01 (m, 2H),
7.07-7.16 (m, 4H), 7.30 (d, 1H, J ) 8.8), 7.37 (d, 1H, J ) 2.3),
J O961889Y