A photolabile 'traceless' linker for solid-phase organic synthesis

Article abstract of DOI:10.1016/S0040-4039(00)01644-0

A photolabile linker based on a thiohydroxamic acid has been shown to be efficient 'traceless' linker, revealing an aliphatic CH bond on photolysis at 350 nm. Methods for the synthesis, loading and phtolytic cleavage of the linker are described. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01644-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 9181–9184
A photolabile ‘traceless’ linker for solid-phase organic
synthesis
James R. Horton, Louise M. Stamp and Anne Routledge*
The School of Chemistry, University of Leeds, Leeds LS2 9JT, UK
Received 14 August 2000; accepted 20 September 2000
Abstract
A photolabile linker based on a thiohydroxamic acid has been shown to be an efficient ‘traceless’ linker,
revealing an aliphatic CH bond on photolysis at 350 nm. Methods for the synthesis, loading and
photolytic cleavage of the linker are described. © 2000 Elsevier Science Ltd. All rights reserved.
The synthesis of combinatorial libraries on solid-phase has experienced a rapid expansion in
the last five years and is now routine in drug discovery. Critical to the success of any solid-phase
synthesis is the linker, which attaches the substrate to the support. There are numerous linkers
currently available1 and judicious choice can result in a successful library synthesis. Almost all
linkers are designed to reveal a polar functionality on release of the substrate from the resin. In
contrast, there are few linkers designed to reveal a CH bond on cleavage. The majority of these
so-called ‘traceless’ linkers are designed to create an aromatic CH bond. These are often based
on arylsilane handles that are cleaved via acid-mediated protodesilylation.2,3 There are scant
examples of linkers that reveal aliphatic CH bonds and all have major limitations such as
harsh/unsuitable cleavage reagents or poor yields.4–9 The use of photolabile linkers in solid-
phase synthesis is an attractive strategy as photolysis can provide an orthogonal non-invasive
method of cleavage. To this end we have designed a new linker, which will reveal an aliphatic
CH bond upon photolytic cleavage. Work by Barton has shown that irradiation of the stable
thiohydroxamic ester 1, in the presence of a suitable chain carrier, initiates a rapid decarboxy-
lative radical rearrangement releasing an alkane (Scheme 1).10
To study the utility and scope of this thiohydroxamic acid as a photolabile ‘traceless’ linker
for solid-phase synthesis, the linker was assembled on resin (Scheme 2). Trityl protected
4-hydroxymercaptophenol was attached to chloromethylpolystyrene (0.8 mmol/g,
Novabiochem) by alkylation. The resin was deprotected and immediately reacted with 1,1-thio-
carbonyldiimidazole. Subsequent displacement of imidazole with N-methylhydroxylamine
* Corresponding author. Tel: +44 (0)113 233 6551; fax: +44 (0)113 233 6565; e-mail: anner@chem.leeds.ac.uk
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01644-0

9182
Scheme 1.
Scheme 2. (i) trityl chloride (1 equiv.), pyridine (1 equiv.), CH2Cl2, ambient temp., 4 h, 97%. (ii) chloromethyl-
polystyrene (0.8 mmol/g), NaH (60% dispersion in oil) (3 equiv.), DMF, 60°C, 18 h, 98%.12 (iii)
TFA:CH2Cl2:triethylsilane, (9:10:1), ambient temp., 1 h. (iv) 1,1-thiocarbonyldiimidazole (3 equiv.), CH2Cl2, ambient
temp., 18 h, 78%.13 (v) N-methyl hydroxylamine hydrochloride (3 equiv.), triethylamine (6 equiv.), CH2Cl2, ambient
temp., 18 h, 73%.13
furnished the resin-bound linker 2.11 The loading of the resin was determined to be 0.39 mmol/g
by elemental analysis of nitrogen.
Various coupling methods were investigated in order to optimise the attachment of N-
methylindole-3-acetic acid to linker 2 (Scheme 3). Aliquots of resin were treated with various
coupling reagents and the results shown in Table 1.
Scheme 3.
The optimum coupling was achieved using N,N-diisopropylcarbodiimide (DIC) and 1-hydroxy-
benzotriazole (HOBt) in the presence of dimethylaminopyridine (DMAP).14 The capacity of
resin 3 to release 1,3-dimethylindole was investigated.15 The results are illustrated in Table 2.
The nature of any photoproducts remaining on the resin were not investigated (loss of the
stretch at 1805 cm−1 in the FTIR was noted, indicating complete consumption of the thio-
hydroxamic ester moiety). Photolysis of 3 in the presence of 2-methyl-2-propanethiol gave a low

9183
Table 1
Coupling of N-methylindole-3-acetic acid to resin 2
Method
Yield of 3 (%)b
Acid (2 equiv.),a DIC (2 equiv.), HOBt (2 equiv.), DMAP (0.2 equiv.)
Acid (2 equiv.), DIC (2 equiv.), HOBt (2 equiv.)
85
80
70
73
Acid (3 equiv.), MSNT (3 equiv.), MeIm (2.5 equiv.)
i
Acid (2 equiv.), TBTU (2 equiv.), HOBt (2 equiv.), Pr2EtNH (4 equiv.)
a Based on the loading of resin 2 (0.39 mmol/g).
b The percentage yield is based on the theoretical 100% loading of resin 3 and is determined by elemental analysis
for nitrogen and the amount of N-methylindole-3-acetic acid methyl ester released on treatment with excess NaOMe.
Table 2
Release of 1,3-dimethylindole from resin 3
Conditions
Yield of 1,3-dimethylindole (%)a
THF: MeOH (4:1), 30 min, 350 nm
THF: (TMS)3SiH (40:1), 30 min, 350 nm
THF: Bu3SnH (40:1), 30 min, 350 nm
THF: Bu3SnH (40:1), 60 min, 350 nm
8
44
55
47
31
38
t
THF: BuSH (40:1), 30 min, 350 nm
Toluene: Bu3SnH (40:1), AIBN (1 equiv.),b 18 h, 70°C
a The yield of released 1,3-dimethylindole is determined by HPLC analysis using diode array detection and indole
as an internal standard.
b Based on the loading of resin 3 (0.31 mmol/g).
yield of 1,3-dimethylindole but this reaction had the advantage of facile removal of excess thiol
via evaporation under reduced pressure. Although in all cases the yield of the photolytic release
was slightly disappointing we were gratified to note the purity of the released 1,3-dimethylindole
was >90% (Fig. 1).
Figure 1. HPLC trace (275 nm detection) of released 1,3-dimethylindole with added indole as an internal standard

9184
In this letter we have illustrated the potential of a thiohydroxamic acid as a photolabile
‘traceless’ linker for solid-phase organic synthesis. We are currently optimising photolytic release
from the resin and exploring the chemical stability of the linker.
Acknowledgements
We would like to thank The University of Leeds, The Royal Society (for a research grant),
Pfizer Central Research and GlaxoWellcome (Dr. S. McKeown) for financial support. We also
thank Dr. Maria Fiorini (University of Cambridge) for some preliminary studies.
References
1. James, I. W. Tetrahedron 1999, 55, 4855–4946.
2. Plunkett, M. J.; Ellman, J. A. J. Org. Chem. 1995, 60, 6006–6007.
3. Plunkett, M. J.; Ellman, J. A. J. Org. Chem. 1997, 62, 2885–2893.
4. Ruhland, T.; Andersen, K.; Pedersen, H. J. Org. Chem. 1998, 63, 9204–9211.
5. Hughes, I. Tetrahedron Lett. 1996, 37, 7595–7598.
6. Patchornik, A.; Kraus, M. A. J. Am. Chem. Soc. 1970, 92, 7587–7589.
7. Garibay, P.; Nielson, J.; Hoeg-Jensen, T. Tetrahedron Lett. 1998, 39, 2207–2210.
8. Zaragoza, F. Tetrahedron Lett. 1997, 38, 7291–7294.
9. Sucholeiki, I. Tetrahedron Lett. 1994, 35, 7307–7310.
10. Barton, D. H. R.; Crich, D.; Potier, P. Tetrahedron Lett. 1985, 26, 5943–5946.
11. The signals that could be distinguished from the polystyrene resin backbone were; 13C NMR (CDCl3, ppm)
115.9(ꢀOCC6 Hꢀ), 67.9(ꢀC6 H2Oꢀ).
12. Yield determined by elemental analysis for sulphur.
13. Yield determined by elemental analysis for nitrogen and sulphur.
14. Procedure for loading resin 2 with N-methylindole-3-acetic acid. Resin 2 (0.5 g, 0.195 mmol) was swollen in a
minimum of CH2Cl2. N-methylindole-3-acetic acid (73 mg, 0.39 mmol), HOBt (53 mg, 0.39 mmol) and DMAP
(4.8 mg, 0.039 mmol) were dissolved in anhydrous DMF (2 ml). This solution was added to the resin mixture,
DIC (49 mg, 0.39 mmol) was added dropwise and the reaction was agitated at ambient temperature for 18 h. The
resin was filtered and washed with DMF (3×5 ml), THF (3×5 ml) and CH2Cl2 (3×5 ml). The resin was dried
under vacuum until a constant weight was achieved (72 h). Gel phase FTIR showed a stretch at 1805 cm−1 (CꢁO).
Gel phase 13C NMR (CDCl3, ppm) 168.7(C
6 ꢁO), 68.5(ꢀC6 H2Oꢀ), 33.3(ꢀNC6 H3 of indole), 29.2(ꢀC6 H2COOꢀ).
15. Typical procedure for the photolysis of resin 3. The resin 3 (0.05 g, 0.0155 mmol of loaded N-methylindole-3-acetic
acid) was swollen in THF:XH (40:1) and the solution degassed with nitrogen for 10 min. The solution was
photolysed at 350 nm using a Rayonet photochemical reactor for 30 min. The resin was removed by filtration and
the solution made up to 10 ml. Indole was added and an aliquot of the solution was removed and analysed by
HPLC (diode array detection).
.