New solid phase triorganogermanium hydrides for radical synthesis

Article abstract of DOI:10.1055/s-2004-825585

New solid phase triorganogermanium hydrides have been synthesized by the addition of a simple triorganogermanium hydride unit onto Quadragel and Merrifield resins. The new solid phase germanium hydrides have been used to mediate a range of synthetic radical reactions.

Full text of DOI:10.1055/s-2004-825585

LETTER
1215
New Solid Phase Triorganogermanium Hydrides for Radical Synthesis
New
S
olidPhase Triorg
.
R
m
H
ydrides for
R
u
adical Synth
s
esis sell Bowman,*a Sussie L. Krintel,a Mark B. Schillingb
a
Department of Chemistry, Loughborough University, Loughborough, Leics. LE11 3TU, UK
Fax +44(1509)223925; E-mail: w.r.bowman@lboro.ac.uk
b
Process Development, GlaxoSmithKline, Gunnels Wood Road, Stevenage, Herts SG1 2NY, UK
Received 16 December 2003
GeMe2X
GeMe2H
Abstract: New solid phase triorganogermanium hydrides have
been synthesized by the addition of a simple triorganogermanium
hydride unit onto QuadragelTM and Merrifield resins. The new solid
phase germanium hydrides have been used to mediate a range of
synthetic radical reactions.
(b)
HO
(a)
RO
(c)
1, X = Me
2, X = Cl
3, R = H
4, R = Bn
Key words: solid-phase synthesis, triorganogermanium hydride,
Scheme 1 Synthesis of the organogermanium moiety for solid
phase usage. Reagents and conditions: (a) SnCl4, MeNO2, 50 °C, 18
h, not isolated. (b) NaBH4 (2.0 equiv), MeOH, r.t., 7 h, 67% over two
steps. (c) NaH, BnBr, THF, reflux, 3 h, 87%.
radicals, radical reactions, triorganogermanium radicals
As part of our studies towards developing the use of trior-
ganogermanium hydrides to replace the toxic tributyltin
hydride (Bu3SnH) as radical generating reagents, we re-
port our initial studies of robust and easy to prepare solid
phase triorganogermanium hydrides. The application of
solid phase to radical reactions is a new and growing area
of interest.1 The use of a solid phase reagent to replace
Bu3SnH has a number of important advantages. Firstly,
the radical reagent is held on the resin, thereby facilitating
easy purification of products, a notorious problem for
Bu3SnH-mediated radical reactions. Secondly, although
triorganogermanium compounds have low toxicity,2 the
attachment to solid phase will help obviate organoger-
manium impurities in the products of radical reactions.
was stable with good shelf life and proved useful as a
radical generating species in its own right.
Initially, several standard radical reactions were tested
with the germanium hydrides 3 and 4 to determine wheth-
er these moieties would act as radical mediators, so that
once attached to the resins, we could be certain of steric
and reactivity factors (Table 1). The hydride 4 was pre-
pared as a more accurate model for the solid phase re-
agent. Both germanium hydrides 3 and 4 gave good yields
of radical deoxygenation of the thiocarbonyl-imidazolide
ester of glucofuranose 10 in a Barton–McCombie
reaction6 and moderate yields in the cyclisation of the a-
chloroamide 14.7 The yields of the 3-deoxy glucofuranose
11 (94% and 91% respectively) were comparable to the
yield (87%) we obtained using Bu3GeH.8 Similarly, the
yields of the g-lactam 16 (47% for both reactions) were
also comparable to our earlier studies with Bu3GeH.8
These results confirmed our prediction that our germani-
um hydride moiety would show comparable activity to
Bu3GeH on the solid phase resins.
Several solid phase triorganotin hydride resins have been
developed1,3 but only one example of a solid phase germa-
nium hydride has been reported.4 Diethylgermanium
dichloride was used for adding the germanium moiety.
The resins were used synthetically for reducing alkyl and
aryl halides and showed good recyclability.
We considered that the best protocol would involve at-
taching an easily prepared triorganogermanium moiety to
the solid phase resin so that once attached, the resin would
be ready for carrying out radical reactions without further
modification.
The germanium hydride moiety 3 was loaded onto Merri-
field resin 5 and QuadragelTM 7 (Scheme 2) to deliver
good loadings of GeH. The resins were ‘swollen’ in tolu-
ene and the germanium hydride moiety 3 added by stan-
dard methodology. The QuadragelTM was first converted
into the mesylate 8 prior to loading. The resins were
To this end, we adapted the ‘traceless’ germanium linker
developed by Spivey.5 The germanium hydride moiety 3
was synthesized in high yielding steps (Scheme 1). The
trimethylgermane 1 was prepared in three steps using the
procedure of Spivey.5 The conversion of the trimethylger-
mane 1 to the dimethylgermanium chloride 2 gave impure
products. Difficulty in purification of 2 was obviated by
not isolating the product and reducing directly to yield the
required germanium hydride 3. The germanium hydride 3
1
analyzed by FTIR and H and 13C NMR spectroscopy.
The Ge–H stretching frequency was observed at nmax 2030
cm–1 for 6 and 2028 cm–1 for 9. The value for Bu3GeH is
2006 cm–1.8 The germanium hydrogen which appears as a
1
septet in the H spectrum at d = 3.68 ppm for Bu3GeH
could not be observed. The loadings were calculated by
weight difference after drying and should be regarded as
approximate due to the inaccuracy in the use of small
amounts of resins (ca. 5 g). The resins did not appear to
lose efficacy with storage.
SYNLETT 2004, No. 7, pp 1215–1218
0
3.
0
6.
2
0
0
4
Advanced online publication: 10.05.2004
DOI: 10.1055/s-2004-825585; Art ID: D29503ST
© Georg Thieme Verlag Stuttgart · New York

1216
W. R. Bowman et al.
LETTER
GeMe2H
3 (5.0 equiv)
Cl
NaH (7.5 equiv)
THF, r.t., 17 h
O
5
Merrifield
loading 3.8 mmol Cl–/g
6, 1.04 mmol GeH/g
MsCl (5.0 equiv)
pyridine, r.t., 2 h
O
H
O
Ms
O
O
4
4
7
8
3 (2.0 equiv), KI (2.0 equiv),
NaH (2.0 equiv)
QuadragelTM
DMF, 60 °C, 24 h
loading 2.1 mmol OH/g
GeMe2H
O
O
4
9, 2.2 mmol GeH/g
Scheme 2 Synthesis of solid phase germanium hydrides
We carried out representative radical reactions to show bond and rates of abstractions are slower allowing more
the efficacy of solid phase germanium hydrides, i.e. gen- time for cyclisation. Significantly, the solid phase germa-
eration of alkyl, vinyl and aryl radicals in cyclisation reac- nium hydrides yield no uncyclised reduced product 17 in-
tions and deoxygenation, using radical addition to the dicating that the polymer supported germanium hydrides
sulfur of the thiocarbonyl group of thiocarbonyl imida- exhibit a dilution effect. This factor could be important in
zolides, in the Barton–McCombie reaction. The results future use of solid phase germanium hydrides for radical
are shown in Table 1.
cyclisations.
We initially carried out the reactions by rotation of the re- Lastly, in order to test the applicability of our reagents for
action mixture at 85 °C [with the a-bromoamide 15 and 2- generating vinyl radicals, we studied the vinyl bromide
iodo-1-(prop-2-enyloxy)benzene (18)] to avoid damage to 20.8,9 The 6-membered ring product 22 results largely
the resin beads. This temperature was not high enough to from rearrangement of the 5-exo radical intermediate.
facilitate the radical reactions and almost no reaction took This results from the slow rate of H-abstraction from the
place. The latter compound was reacted at a higher tem- germanium hydride by the 5-exo intermediate allowing
perature (95 °C rising to 110 °C) which gave a 26% yield time for the rearrangement.
of the cyclised product. Finally, at reflux a high yield
For the cyclisation of the vinyl radical generated from 20,
we checked the competition between the 5-exo and 6-endo
(78%) was obtained which compares with the reaction
using Bu3GeH (91%).8 Gentle heating under reflux did
products by using polarity reversal catalysis (PRC) with
not damage the beads.
phenylthiol.8,10 The PhSH rapidly intercepts the 5-exo
The Barton–McCombie deoxygenation of the thiocarbon- radical to give only the 5-exo product 21 (30% with no 6-
yl imidazolide ester cholesterol 12 gave a good yield endo product) before 21 is able to undergo a neophyl rear-
(68%) with the QuadragelTM-germanium hydride 9. In rangement to 22.8,9 The resulting phenylthiyl radical
contrast, the deoxygenation of the glucopyranose with the (PhS·) is electrophilic and undergoes rapid reaction with
Merrifield-germanium hydride 6 the yield (36%) was the nucleophilic germanium hydride to complete the chain
lower, especially as compared to the ‘unattached’ germa- reaction. The moderate yields are possibly caused by ad-
nium hydrides 3 and 4.
dition of the intermediate germanium radicals to the gem-
disubstituted alkenes of the products to give triorganoger-
manium adducts via stable tertiary radical intermediates.
With Bu3GeH this causes considerable problems of puri-
fication. This is obviously a drawback for radical reac-
tions, which yield disubstituted alkenes as products.
However, with the QuadragelTM-germanium hydride and
PRC, the adducts remain attached to the resin and only
one pure product 21 is formed, albeit in 30% yield. These
problems need to be further addressed in future.
In the cyclisation of the a-bromoamide 15, the best yield
was obtained with the QuadragelTM-germanium hydride
9. In general, the QuadragelTM-germanium hydride gave
better results than the Merrifield-germanium hydride. The
results again compare with those of Bu3GeH and
Bu3SnH.8 The best results for this reaction have been
obtained with syringe pump addition of Bu3GeH and
Bu3SnH.7
When Bu3GeH and Bu3SnH were added only at the begin-
ning of the reaction, more cyclisation relative to reduction
to 17 was obtained for Bu3GeH than Bu3SnH. This is ex-
pected because the Ge–H bond is stronger than the Sn–H
Our initial results indicate that solid phase germanium hy-
drides show promise for carrying out radical reactions and
can facilitate a clean and simple methodology. Further
Synlett 2004, No. 7, 1215–1218 © Thieme Stuttgart · New York

LETTER
New Solid Phase Triorganogermanium Hydrides for Radical Synthesis
1217
Table 1 Radical Reactions Using Triorgano-Germanium Hydride
Precursor, Germanium-hydride
Products, Yields
Conditionsa
O
O
O
O
O
O
Im
O
O
O
O
O
11
S
10
3 (3.0 equiv)
4 (3.3 equiv)
6 (1.3 equiv)
10 (0%), 11 (94%)
10 (0%), 11 (91%)
10 (0%), 11 (36%)
ACCN,b reflux, 2 h
ACCN, reflux, 8 h
ACCN, reflux, 7 h
H
H
S
H
H
Im
O
13
12
9 (4.4 equiv)
12 (0%), 13 (68%)
ACCN, reflux, 17 h
X
O
O
N
O
N
N
PMB
PMB
17
PMB
16
14, X = Cl
15, X =Br
14; 3 (1.2 equiv)
14; 4 (1.1 equiv)
15; 6 (0.5 equiv)
15; 6 (0.5 equiv)
15; 9 (2.5 equiv)
15; 9 (2.6 equiv)
14 (18%), 16 (47%), 17 (trace)
14 (trace), 16 (47%), 17 (3%)
15 (66%), 16 (30%),c 17 (0%)
15 (55%), 16 (35%),c 17 (0%)
15 (ca. 100%), 16 (trace), 17 (0%)
15 (0%), 16 (57%), 17 (0%)
ACCN, reflux, 8 h
ACCN, reflux, 8 h
ACCN, reflux, 9 h
ACCN, reflux, 24 h
AIBN, 85 °C, 17 h
AIBN, reflux, 19 h
.
I
O
O
18
19
9 (3.0 equiv)
9 (3.5 equiv)
9 (3.6 equiv)
Br
18 (ca. 100%), 19 (trace)
AIBN, 85 °C, 4 h
18 (28%), 19 (26%)
18 (0%), 19 (78%)
AIBN, 95–110 °C, 20 h
AIBN, reflux, 16 h
MeO2C
CO2Me
M
M
M
M
20
21
22
9 (4.0 equiv)
9 (4.0 equiv)
20 (25%), 21 (19%), 22 (14%)
20 (0%), 21 (30%), 22 (0%)
AIBN, reflux, 19 h
AIBN, reflux, 4 h, PhSH (0.1 equiv)
a Toluene was used for swelling the resin and for the reaction.
b ACCN = 1,1¢-Azobis(cyclohexanecarbonitrile).
c The % yield was calculated based on 6.
Synlett 2004, No. 7, 1215–1218 © Thieme Stuttgart · New York

1218
W. R. Bowman et al.
LETTER
studies are underway to optimize the conditions for use of
the solid phase Ge species as radical mediators and in
particular to investigate recycling and shelf life in order to
lower the cost.
Acknowledgment
We thank GlaxoSmithKline and Loughborough University for a
postgraduate studentship (SLK), GlaxoSmithKline for generous
support, Drs Alan Spivey and Ratnosothy Srikaran for helpful ad-
vice and the EPSRC Mass Spectrometry Unit, Swansea University,
Wales for mass spectra.
QuadragelTM-Germanium Hydride 19: QuadragelTM mesylate 8
(5.0 g, ca. 11 mmol) was swollen in toluene (70 mL). 4-(2-Dimeth-
ylgermylethyl)phenol (3, 5.02 g, 22.3 mmol), sodium hydride (0.90
g, 22.4 mmol) and KI (3.72 g, 22.4 mmol) were added and the sus-
pension rotated at 60 °C for 24 h. The reaction was cooled to r.t.,
quenched with aq THF and the resin removed by filtration and
washed with aq THF, THF and MeOH. The resin was dried in vacuo
overnight to yield a dark brown polymer 9 (6.97 g, equivalent to a
loading of 2.2 mmol GeH/g). IR (KBr): 3379, 2929, 2028, 1658,
1606, 1196, 1056 cm–1. 1H NMR (250 MHz): d = 0.17–0.36 (br, Me,
CH2Ge), 1.31–1.54 (br s, PS-CHn), 2.48 (br s, CH2Ar), 3.42–4.20
(br s, OCH2CH2O), 6.56-7.04 (br s, Ph-H, Ar-H). 13C NMR (100
MHz): d = 4.7 (Me), 16.7 (CH2Ge), 31.5 (CH2Ar), 40.0-44.8 (Ph-
CHn), 67.8 (CH2O), 70.2–71.0 (OCH2), 115.0 (Ph-2,6-C, Ar-2,6-
H), 126.0–127.9 (PS-PhCH), 129.1 (Ar 3,5-C), 137.3 (Ar-C) and
157.2 (Ar-C).
References
(1) Ganesan, A. In Radicals in Organic Synthesis, Vol. 2;
Renaud, P.; Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001,
Chap. 1.5, 81.
(2) (a) Craig, P. J.; Van Elteren, J. T. In The Chemistry of
Organic Germanium, Tin and Lead Compounds; Patai, S.,
Ed.; Wiley: Chichester, 1995, Chap. 16, 843. (b) Lukevics,
E.; Ignatovich, M. In The Chemistry of Organic Germanium,
Tin and Lead Compounds; Patai, S., Ed.; Wiley: Chichester,
1995, Chap. 17, 857.
(3) (a) Enholm, E. J.; Schutte, J. P. Org. Lett. 1999, 1, 1275.
(b) DuMartin, G.; Pourcel, M.; Delmond, B.; Donard, O.;
Pereyre, M. Tetrahedron Lett. 1998, 39, 4663.
(4) Mochida, K.; Sugimoto, H.; Yokoyama, Y. Polyhedron
1997, 16, 1767.
(5) (a) Spivey, A. C.; Turner, D. J.; Turner, M. L.; Yeates, S.
Synlett 2004, 111. (b) Spivey, A. C.; Srikaran, R.; Diaper, C.
M.; Turner, D. J. Org. Biomol. Chem. 2003, 1, 1638.
(c) Spivey, A. C.; Turner, D. J.; Turner, M. L.; Yeates, S.
Org. Lett. 2002, 4, 1899. (d) Spivey, A. C.; Diaper, C. M.;
Adams, H. J. Org. Chem. 2000, 65, 5253. (e) Spivey, A. C.;
Diaper, C. M.; Rudge, A. J. Chem. Commun. 1999, 835.
(6) (a) Motherwell, W. B.; Crich, D. Free Radical Reactions in
Organic Synthesis; Academic Press: London, 1991.
(b) Barton, D. H. R.; McCombie, S. W. J. Chem. Soc.,
Perkin Trans. 1 1975, 1574.
Cyclisation of 2-Iodo-1-(propen-2-yloxy)benzene (18): Quad-
ragelTM-germanium hydride 9 (0.34 g, 0.75 mmol) was swollen in
anhyd toluene (12 mL); 2-iodo-1-(prop-2-enyloxy)benzene (54.0
mg, 0.21 mmol) and AIBN (0.12 g, 0.73 mmol) were added and the
mixture was heated under reflux for 16 h. After cooling to r.t., the
resin was removed by filtration and washed with CH2Cl2 several
times. Evaporation of the filtrate gave the crude cyclic product as a
yellow oil. The yield of 3-methyl-2,3-dihydrobenzofuran (19) was
1
22 mg (78%). H NMR (250 MHz): d = 1.33 (d, J = 7.4 Hz, 3 H,
Me), 3.47–3.62 (m, 1 H, 3-H), 4.07 (dd, J = 7.4, 8.8 Hz, 1 H, 2-Ha),
4.68 (dd, J = 8.8, 8.8 Hz, 1 H, 2-Hb), 6.77–6.89 (m, 2 H, Ar-H),
7.08–7.17 (m, 2 H, Ar-H). 13C NMR (62.5 MHz): d = 20.1 (Me),
37.2 (3-C), 31.5 (CH2Ar), 79.1 (2-C), 110.1 (Ar-CH), 121.1 (Ar-
CH), 124.4 (Ar-CH), 128.6 (Ar-CH), 132.9 (3a-C), 160.3 (7a-C).
MS (EI): m/z (%) = 134 (84), 119 (95), 91 (100). The data were
identical to those of the authentic material.11
(7) Gilbert, B. C.; Kalz, W.; Lindsay, C. I.; McGrail, P. T.;
Parsons, A. F.; Whittaker, D. T. E. J. Chem. Soc., Perkin
Trans. 1 2000, 1187.
(8) Bowman, W. R.; Krintel, S. L.; Schilling, M. B. Org.
Biomol. Chem. 2004, 585.
(9) Gómez, A. M.; Company, M. D.; Uriel, C.; Valverde, S.;
López, C. J. Tetrahedron Lett. 2002, 43, 4997; and
references therein.
(10) Roberts, B. P. Chem. Soc. Rev. 1999, 28, 25.
(11) Bhandal, H.; Patel, V. H.; Pattenden, G.; Russell, J. J. J.
Chem. Soc., Perkin Trans. 1 1990, 2691.
Synlett 2004, No. 7, 1215–1218 © Thieme Stuttgart · New York