Biomimetic polyorganosiloxanes: Model compounds for new materials

Article abstract of DOI:10.1039/c4dt00554f

The chemistry of N-organosilylalkyl-substituted heterocyclic bases (thymine, adenine and cytosine) is described, covering the structures of model compounds, the synthesis of substituted oligo-siloxanes and a preliminary report of the synthesis of a poly(organosiloxane) with pendant N-alkyl(heterocycle) functionalities. N-Alkenylthymines CH₂CH(CH₂)_nT (T = thymine, n = 1 (1), 2 (2), 3 (3)) have been prepared and 2 hydrosilylated to form PhMe₂Si(CH₂)₄T (5). Alternatively, 5 was prepared by reaction of PhMe₂Si(CH₂)₄Br (6) with (O,O-SiMe₃)₂T, a method which has also been used to prepare PhMe₂Si(CH₂)₄A (7) and PhMe ₂Si(CH₂)₄C (8) (A = adenine, C = cytosine). Model di- and tri-siloxanes [Br(CH₂)₄(Me) ₂Si]₂O (10), Me₃SiOSi(Me)₂(CH ₂)₄Br (11), PhMe₂SiOSi(Me)₂(CH ₂)₄Br (12) and (Me₃SiO)₂(Me) Si(CH₂)₄Br (13) have been prepared by hydrosilylation of H₂CC(H)(CH₂)₄Br with an appropriate hydrosiloxane and used to prepare Me₃SiO(Me)₂Si(CH ₂)₄T (14), Me₃SiO(Me)₂Si(CH ₂)₄A (15) (both from 11), and (Me₃SiO) ₂(Me)Si(CH₂)₄T (16), (Me₃SiO) ₂(Me)Si(CH₂)₄A (17) (both from 13). 10 reacts with thymine to give a mixture of the pyrimidocyclophane cyclo-T-N,N-[(CH ₂)₄(Me)₂Si]₂O (19) and [T(CH ₂)₄Si(Me)₂]₂O (20), while cytosine reacts similarly to form cyclo-C-N,N-[(CH₂)₄(Me) ₂Si]₂O (21; as an imine) and [C(CH₂) ₄Si(Me)₂]₂O (22); adenine only generates [A(CH₂)₄Si(Me)₂]₂O (18) in an analogous synthesis. Using a related protocol, polymeric {[MeSi(O)(CH ₂)₄Br]₂[Me₂SiO]₉₈} _n (23) has been converted to {[MeSi(O)(CH₂) ₄T]₂[Me₂SiO]₉₈}_n (24) and {[MeSi(O)(CH₂)₄A]₂[Me₂SiO] ₉₈}_n (25). The structures of 4, 5, 8, 19 and 21, along with a 2:1 adduct of 5 with Ni(dithiobiuret)₂ (9) are reported. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4dt00554f

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Biomimetic polyorganosiloxanes: model
compounds for new materials†
Cite this: Dalton Trans., 2014, 43,
7734
Gabriele Kociok-Köhn,a Mary F. Mahon,a Kieran C. Molloy,*a Gareth J. Price,*a
Timothy J. Priorb and Douglas R. G. Smitha
The chemistry of N-organosilylalkyl-substituted heterocyclic bases (thymine, adenine and cytosine) is
described, covering the structures of model compounds, the synthesis of substituted oligo-siloxanes and
a preliminary report of the synthesis of a poly(organosiloxane) with pendant N-alkyl(heterocycle)
functionalities. N-Alkenylthymines CH2vCH(CH2)nT (T = thymine, n = 1 (1), 2 (2), 3 (3)) have been
prepared and 2 hydrosilylated to form PhMe2Si(CH2)4T (5). Alternatively, 5 was prepared by reaction of
PhMe2Si(CH2)4Br (6) with (O,O-SiMe3)2T, a method which has also been used to prepare PhMe2Si(CH2)4A (7)
and PhMe2Si(CH2)4C (8) (A = adenine, C = cytosine). Model di- and tri-siloxanes [Br(CH2)4(Me)2Si]2O (10),
Me3SiOSi(Me)2(CH2)4Br (11), PhMe2SiOSi(Me)2(CH2)4Br (12) and (Me3SiO)2(Me)Si(CH2)4Br (13) have been
prepared by hydrosilylation of H2CvC(H)(CH2)4Br with an appropriate hydrosiloxane and used to prepare
Me3SiO(Me)2Si(CH2)4T (14), Me3SiO(Me)2Si(CH2)4A (15) (both from 11), and (Me3SiO)2(Me)Si(CH2)4T (16),
(Me3SiO)2(Me)Si(CH2)4A (17) (both from 13). 10 reacts with thymine to give a mixture of the pyrimido-
cyclophane cyclo-T-N,N-[(CH2)4(Me)2Si]2O (19) and [T(CH2)4Si(Me)2]2O (20), while cytosine reacts
similarly to form cyclo-C-N,N-[(CH2)4(Me)2Si]2O (21; as an imine) and [C(CH2)4Si(Me)2]2O (22); adenine
only generates [A(CH2)4Si(Me)2]2O (18) in an analogous synthesis. Using a related protocol, polymeric
Received 21st February 2014,
Accepted 20th March 2014
{[MeSi(O)(CH2)4Br]2[Me2SiO]98 n
{[MeSi(O)(CH2)4A]2[Me2SiO]98 n
with Ni(dithiobiuret)2 (9) are reported.
}
(23) has been converted to {[MeSi(O)(CH2)4T]2[Me2SiO]98 n
}
(24) and
DOI: 10.1039/c4dt00554f
}
(25). The structures of 4, 5, 8, 19 and 21, along with a 2 : 1 adduct of 5
www.rsc.org/dalton
logous to nucleic acids have been made from easily-
synthesised vinyl, polypeptide or methacrylate polymers or oli-
Introduction
The use of hydrogen bonds to direct the self-assembly of novel gomers with biologically functional pendant groups combined
materials is one of the most exploited routes to nanostructures with controlled radical polymerization methods to produce
of defined composition. The most sophisticated of such well-defined structures.2,3 Materials based on poly(ε-caprolac-
systems – DNA – in which complementary purine/pyrimidine tone)4 and polylactides5 have been produced, as well as block
base pairs provide a framework for self-replication is arguably copolymers with, for example, poly(ethylene glycol).6 Exploita-
the most simple, yet most structurally appealing, example of tion of the nucleobase interactions in the block materials led
nature’s work. In recent years, the interplay of T/A and C/G to enhanced aggregation over the individual copolymers. The
base pairs (T = thymine, A = adenine, C = cytosine, G = exploitation of these complementary interactions in forming
guanine) has been used to construct numerous nanomaterials, supramolecular polymers has been reviewed by Sessler et al.7
with applications including nanomechanical devices.1
Such polymers have potential biomedical applications8–10 and
Furthermore, there has been interest in the functionali- have also been proposed as thermoresponsive materials.11
sation of polymer chains with biomimetic substituents, in Rowan and co-workers have used the complementary nucleo-
attempts to produce synthetic polymers with the same pro- base interactions between T/A12 or C/A11 to prepare stable,
perties as their naturally occurring counterparts. Polymers ana- thermotropic liquid crystalline polymers which displayed ther-
moreversible phase behaviour.
However, as far as we are aware, there has only been very
limited attention given to silicon-based materials functiona-
aDepartment of Chemistry, University of Bath, Bath BA2 7AY, UK.
E-mail: G.J.Price@bath.ac.uk
lised by the nucleobases. Déléris et al. carried out the most sig-
bStation 16.2 SMX, CLRC Daresbury Laboratory, Daresbury, Warrington,
nificant studies, with the aim of producing anti-HIV agents,13
Cheshire WA4 4AD, UK
or improving the lipophilicity of oligodeoxynucleotides,14,15
†CCDC 956088–956094. For crystallographic data in CIF or other electronic
format see DOI: 10.1039/c4dt00554f
but none of these studies specifically addressed the formation
7734 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
of siloxane polymers. More recent work has focussed on the growth of the peak at ca. 1750 cm−1 due to the carbonyl
synthesis of nanostructured organosilica materials by hydro- groups on thymine (ca. 72 h). The THF and excess allyl
lysis of (EtO)3Si(CH2)n-subsituted bases,16 and the resulting bromide were then removed under reduced pressure leaving
solids were characterised by solid-state NMR.17 Amino acid- the product as a yellow oil from which the product 1-allylthy-
functionalised siloxanes are, however, better known.18 Meijer mine (1) spontaneously crystallised. Two recrystallizations
and co-workers19,20 prepared siloxanes terminated with quad- from toluene liberated the pure product as white platelets
ruple hydrogen-bonding ureido-pyrimidone units which (5.5 g, 81%, m.p. 109–111 °C, lit. 112 °C,14 96–99 °C26).
assembled into high molecular weight polymer networks with
reversible rheology.
Also prepared by the same method were:
1-(But-3-en-1-yl)-5-methylpyrimidine-2,4(1H,3H)-dione (2).
Our own interest in silicon-based polymers,21 and in par- From bis-(trimethylsilyl)thymine (3.0 g 11 mmol) and 4-
ticular polysiloxanes,22,23 has led us to look more closely at the bromo-1-butene (3.0 cm3, 30 mmol) over 72 h. The resulting
chemistry of the latter polymer backbones functionalised by brown oil was chromatographed on silica with a mixture of
nucleobases, which might ultimately afford novel, self-organis- dichloromethane and methanol (98 : 2), yielding one band
ing materials. While our synthetic methodology builds on the from which the crude product was obtained on solvent evapor-
earlier work of Déléris,13–15 our emphasis is on the synthesis ation. Two recrystallizations from toluene yielded the pure
and structural chemistry of both precursors and model com- product (2), recovered as white platelets (1.17 g, 59%, m.p.
pounds as well as the polymers themselves, with a view to 119–121 °C, lit. 133–135 °C27). 1H NMR (300 MHz, CDCl3):
understanding more fundamentally the nature of any self- 9.68 (bs, 1H, NH), 6.98 (s, 1H, CH), 5.76 (m, 1H, CHvCH2),
assembly processes.
5.10 (m, 2H, CHvCH2), 3.78 (t, J = 7.0 Hz, 2H, NCH2), 2.47 (m,
2H, NCH2CH2), 1.92 (s, 3H, CH3); 13C NMR (300 MHz, CDCl3):
164.6 (CvO), 151.0 (CvO), 140.7 (CH), 133.5 (CHvCH2),
118.5 (CHvCH2), 110.4 (CCH3), 48.0 (NCH2), 33.3 (NCH2CH2),
12.3 (CCH3); analysis, calcd for C9H12N2O2: C 60.0, H 6.71, N
15.6%, found C 60.0, H 6.58, N 15.6%.
Experimental
Experimental procedures
Starting materials adenine, thymine, cytosine, guanine,
5-Methyl-1-(pent-4-en-1-yl)pyrimidine-2,4(1H,3H)-dione (3).
phenyldimethylsilane, (3-aminopropyl)-diethoxymethylsilane, From bis-(trimethylsilyl)thymine (3.0 g 11 mmol) and 4-
chlorodimethylsilane, chloroplatinic acid, allylamine, allyl bromo-1-pentene (2.6 cm3, 22 mmol) over 144 h. Work-up as
bromide, 4-bromo-1-butene, hexamethyldisilazane, 3-bis- for 2 yielded the product (3) as brown needles (0.38 g, 18%,
1
(3-aminopropyl)tetramethyldisiloxane, PDMS (Mw ∼ 50 000) m.p. 134–6 °C, lit. 105–106 °C28). H NMR (300 MHz, CDCl3):
and PDMS (Mw ∼ 25 000) were obtained from Sigma-Aldrich 9.84 (bs, 1H, NH), 6.98 (s, 1H, CH), 5.79 (m, 1H, CHvCH2),
and used without further purification. Pentamethyldisiloxane 5.05 (m, 2H, CHvCH2), 3.71 (t, J = 7.3 Hz, 2H, NCH2), 2.11 (m,
and 1,1,1,3,5,5,5-heptamethyltrisiloxane were obtained from 2H, NCH2CH2), 1.92 (s, 3H, CH3), 1.80 (m, 2H, NCH2CH2CH2);
Gelest. All reactions were carried out under nitrogen unless 13C NMR (300 MHz, CDCl3): 164.7 (CvO), 151.2 (CvO), 140.5
otherwise specified. O,O-bis(trimethylsilyl)thymine, N,N-bis- (CH), 133.7 (CHvCH2), 115.8 (CHvCH2), 110.6 (CCH3), 48.0
(trimethysilyl)adenine and O,N-bis(trimethylsilyl)cytosine (NCH2), 30.4 (NCH2CH2), 28.0 (m, 2H, NCH2CH2CH2), 12.3
were prepared as described elsewhere,24 as was nickel bis- (CCH3); analysis calcd for C10H14N2O2: C 61.8, H 7.27, N
(dithiobiuret).25
14.4%, found C 61.4, H 7.18, N 14.1%.
Infra-red spectra were recorded on NaCl plates using a
1-Allyl-5-methyl-4-((trimethylsilyl)oxy)pyrimidin-2(1H)-one (4).
Nexus Nicolet 510P FT-IR spectrometer in the region A mixture of hexamethyldisilazane (20 cm3, 91 mmol), 1-allyl-
4000–400 cm−1 1H, 13C and 29Si spectra were recorded on 5-methylpyrimidine-2,4(1H,3H)-dione (1) (2.00 g, 12 mmol)
.
Bruker Avance (300 MHz) Fourier transform spectrometer, and a few crystals of ammonium sulphate was refluxed for
using TMS as an internal reference. Elemental analyses were 18 h. After this time, the solid had dissolved and a colourless
performed using a Carbo Erba Strumentazione E.A. model solution remained. The excess disilazane was removed under
1106 analyser. The results were duplicated and the mean of reduced pressure, affording the clear oily product (2.68 g,
the duplicated measurements was the final result. Gel per- 93%). 4 spontaneously crystallised upon cooling under nitro-
meation chromatography (GPC) was carried out in chloroform gen. 1H NMR (300 MHz, CHCl3): 6.98 (q, J = 1.0 Hz, CH),
at 30 °C using a flow rate of 1 ml min−1 through two 30 cm ‘PL 5.90 (m, 1H, CHvCH2), 5.27 (m, 2H, CHvCH2), 4.27 (d,
Gel 10 μm Mixed’ columns. Data was collected from a refrac- J = 6.0 Hz, 2H, NCH2), 1.92 (d, J = 1.0 Hz, 3H, CH3), 0.38 (s,
tive index detector on a Viskotek ‘Trisek 2000’ chromatograph 9H, Si(CH3)3); 13C NMR (75 MHz, CDCl3): 170.3 (COSi),
and analysed using ‘Trisec 3.0’ software. Molecular weights are 156.5 (CvO), 143.9 (CH), 132.6 (CHvCH2), 118.9 (CHvCH2),
reported relative to polystyrene calibration.
106.1 (CH), 51.5 (NCH2), 12.5 (CCH3); analysis, calcd for
1-Allyl-5-methylpyrimidine-2,4(1H,3H)-dione (1). A THF C11H20N2O2Si: C 55.0, H 8.39, N 11.7%, found C 55.1, H 8.29,
(50 cm3) solution of bis-(trimethylsilyl)thymine (11.0 g, N 11.8%.
41 mmol)24 and an excess of allyl bromide (13.0 cm3,
1-(4-(Dimethyl(phenyl)silyl)butyl)-5-methyl-4-((trimethylsilyl)-
150 mmol) were refluxed until the IR spectrum displayed no oxy)pyrimidin-2(1H)-one (5). A mixture of 1-(but-3-en-1-yl)-
further generation of the product, monitored by charting the 5-methylpyrimidine-2,4(1H,3H)-dione (2) (0.50 g, 2.74 mmol)
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 7734–7746 | 7735

View Article Online
Paper
Dalton Transactions
and excess of hexamethyldisilazane and a few crystals of C6H5), 3.35 (t, J = 6.9 Hz, 2H, CH2Br), 1.84 (m, 2H, CH2CH2Br),
ammonium sulphate were refluxed for 18 h, after which the 1.45 (m, 2H, SiCH2CH2), 0.74 (m, 2H, SiCH2), 0.27 (s, 6H, Si
opacity due to solid 2 had cleared. The volatiles were removed (CH3)2); 13C NMR (75 MHz, CDCl3): 139.1 (C6H5), 133.5 (C6H5),
under reduced pressure, and to the remaining colourless oil 133.0 (C6H5), 129.2 (C6H5), 128.9 (C6H5), 127.8 (C6H5), 36.2
(4a) was added THF (10 cm3), phenyldimethylsilane (0.37 g, (CH2Br), 33.4 (CH2CH2Br), 22.5 (SiCH2CH2), 14.8 (SiCH2), −3.1
2.74 mmol) and a catalytic amount of chloroplatinic acid in (Si(CH3)2); 29Si NMR (60 MHz, CDCl3): −3.0; analysis, calcd for
propan-2-ol. The solution was heated to 80 °C for 72 h, after C12H19BrSi: C 55.1, H 7.06%, found C 55.7, H 6.91%.
which the IR spectrum (neat liquid) revealed an absence of
9-(4-(Dimethyl(phenyl)silyl)butyl)-9H-purin-6-amine (7).
starting material. The volatiles were removed, affording a Reaction of N,N-bis(trimethysilyl)adenine (0.56 g, 2.02 mmol),
brown solid, which was chromatographed on silica using a 6 (0.55 g, 2.03 mmol) and TBAF (0.53 g, 2.04 mmol)
mixture of dichloromethane and methanol (98 : 2) as eluent. yielded 7 as fine white needles (0.34 g, 55%). 1H NMR
The chromatography yielded two bands. The first was eluted (300 MHz, CDCl3): 8.36 (s, 1H, CH), 7.69 (s, 1H, CH), 7.45
with the solvent front and afforded unreacted phenyldimethyl- (m, 2H, C6H5), 7.30 (m, 3H, C6H5), 5.65 (bs, 2H, NH2), 4.15 (t,
silane. The second was eluted at r = 0.45 and afforded the J = 7.0 Hz, 2H, NCH2), 1.89 (m, 2H, NCH2CH2), 1.37 (m, 2H,
crude product as a white powder. Recrystallization from a SiCH2CH2), 0.80 (m, 2H, SiCH2), 0.25 (s, 6H, Si(CH3)2); 13C
mixture of chloroform and hexane yielded (4-phenyldimethyl- NMR (75 MHz, CDCl3): 155.3 (CNH2), 152.9 (CH), 140.4 (CH),
silyl)butylthymine (5) as colourless tabular crystals (0.43 g, 133.5 (C6H5), 129.0 (C6H5), 127.8 (C6H5), 43.4 (NCH2), 33.6
1
51%). H NMR (300 MHz, 8CDCl3): 8.92 (bs, 1H NH), 7.49 (m, (NCH2CH2), 21.0 (SiCH2CH2), 15.2 (SiCH2) −3.2 (Si(CH3)2), two
2H, C6H5), 7.32 (m, 3H, C6H5), 3.67 (t, J = 7.0 Hz, 2H, NCH2), NCN too weak to be observed; 29Si NMR (60 MHz, CDCl3):
1.91 (s, 3H, CH3), 1.68 (m, 2H, NCH2CH2), 1.33 (m, 2H, −3.1; analysis, calcd for C17H23N5Si: C 62.7, H 7.12, N 21.5%,
SiCH2CH2), 0.78 (m, 2H, SiCH2), 0.27(s, 6H, Si(CH3)2); 13C found C 61.9, H 7.13, N 21.3%.
NMR (75 MHz, CDCl3): 164.3 (CvO), 150.9 (CvO), 140.4 (CH),
4-Amino-1-(4-(dimethyl(phenyl)silyl)butyl)pyrimidin-2(1H)-
110.5 (CCH3), 47.9 (NCH2), 32.7 (NCH2CH2), 21.0 (SiCH2CH2), one (8). Reaction of O,N-bis(trimethysilyl)cytosine (1.15 g,
15.4 (CCH3), 12.3 (SiCH2), −3.1 (Si(CH3)2); 29Si NMR (60 MHz, 4.05 mmol), 6 (1.22 g, 4.05 mmol) and TBAF (1.55 g,
CDCl3): −3.0; analysis, calcd for C17H24N2O2Si: C 64.5, H 7.64, 4.55 mmol) yielded 8 as tabular crystals (0.61 g, 45%). 1H NMR
N 8.85%, found C 63.9, H 7.64, N 8.61%.
(300 MHz, CDCl3): 8.02 (bs, 2H, NH2), 7.48 (m, 2H, C6H5), 7.35
Alternative synthesis of 5. A solution of (4-bromobutyl)phe- (m, 3H, C6H5), 7.13 (d, J = 1.0 Hz, 1H, NCH), 5.86 (d, J = 1.0 Hz,
nyldimethylsilane (6, 1.00 g, 3.69 mmol), bis-(trimethylsilyl)- 1H, CCH), 3.60 (t, J = 7.0 Hz, 2H, NCH2), 1.59 (m, 2H,
thymine (1.00 g, 3.69 mmol) and tetrabutylammonium fluor- NCH2CH2), 1.35 (m, 2H, SiCH2CH2), 0.76 (m, 2H, SiCH2), 0.26
ide (0.96 g, 3.69 mmol) in THF (50 cm3) was refluxed for 72 h. (s, 6H, Si(CH3)2); 13C NMR (75 MHz, CDCl3): 162.5 (CNH2),
After removal of the volatiles, the solid remaining was purified 145.0 (CvO), 140.0 (CH), 133.4 (C6H5), 128.9 (C6H5), 127.8
by chromatography on silica, using a mixture of dichloro- (C6H5), 99.0 (CH), 49.9 (NCH2), 32.7 (NCH2CH2), 21.0
methane and methanol as eluent (98 : 2). The chromatography (SiCH2CH2), 15.5 (SiCH2), −3.1 (Si(CH3)2); 29Si NMR (60 MHz,
yielded three bands. The first eluted with the solvent front and CDCl3): −3.1; analysis, calcd for C16H23N3OSi: C 63.8, H 7.69,
was found to be unreacted 6. The second band eluted at N 13.9%, found C 63.6, H 7.8, N 13.6%.
approximately r = 0.7 and was found to be TBAF. The third
eluted as a broad smear at between r = 0.75 and 0.45 and was 0.22 mmol) and Ni(dtb)2
Nickel bis(dithiobiuret) adduct (1 : 2) with 5 (9). 5 (0.070 g,
25
(0.074 g, 0.22 mmol) were dis-
found to be the target product. The product was washed with solved in a minimum of warm methanol (25 mL), sealed with
water to extract any further traces of remaining TBAF, and then a subaseal punctured with a syringe needle and left to cool to
recrystallised from a mixture of dichloromethane and hexane, room temperature. Pink crystals of Ni(dtb)2·2(5) (9) grew slowly
yielding the pure product (5) as colourless, tabular crystals over a period of days. Yield 20%, m.p. 224 °C; no further analy-
(0.86 g, 74%).
sis of these crystals was carried out.
(4-Bromobutyl)phenyldimethylsilane (6). A mixture of phe-
1,3-Bis(4-bromobutyl)-1,1,3,3-tetramethyldisiloxane (10). A
nyldimethylsilane (4.3 cm3, 18.3 mmol), 4-bromo-1-butene mixture of 4-bromo-1-butene, chlorodimethylsilane and a cata-
(2.0 cm3, 19.89 mmol) and a catalytic amount of chloroplatinic lytic amount of chloroplatinic acid in propan-2-ol was heated
acid in propan-2-ol was heated to 80 °C for 72 h, after which IR at 110 °C until the neat-liquid IR spectrum revealed an
spectroscopy revealed a lack of starting material by an absence absence of starting material, denoted by lack of Si–H stretch-
of a peak at 2150 cm−1 corresponding to the Si–H stretching ing vibration at 2150 cm−1 (ca. 72 h). The volatiles were
vibration. The excess 4-bromo-1-butene was removed under removed and the remaining grey liquid was hydrolysed,
reduced pressure, and the remaining solution was chromato- washed with water, extracted with chloroform and dried over
graphed using chloroform as eluent. The chromatography MgSO4. Column chromatography on silica with chloroform as
yielded one band, eluting with the solvent front and affording eluent afforded the 1,3-bis-(4-bromobutyl)-disiloxane product
the product (4-bromobutyl)phenyldimethylsilane (6, 3.7 g, as an air stable, colourless oil, recovered with the solvent front.
74%). Variations on the experiment were made by varying the 1H NMR (300 MHz, CDCl3): 3.42 (t, J = 7.0 Hz, 2H, CH2Br), 1.88
time of reaction between 18 h and 96 h but no improvements to (m, 2H, CH2CH2Br), 1.48 (m, 2H, SiCH2CH2), 0.52 (m, 2H,
1
the yield were made. H NMR (300 MHz, CDCl3): 7.42 (m, 5H, SiCH2), 0.15 (s, 6H, Si(CH3)2); 13C NMR (75 MHz, CDCl3): 35.7
7736 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
(CH2Br), 33.3 (CH2CH2Br), 21.6 (SiCH2CH2), 17.0 (SiCH2), 0.7 and 1-(4-bromobutyl)-1,1,3,3,3-pentamethyldisiloxane (11;
(Si(CH3)2); 29Si NMR (60 MHz, CDCl3): 7.0.
0.50 g, 1.77 mmol). Yield: 0.07 g, 12%. 1H NMR (300 MHz,
Also prepared by the same method were:
CDCl3): 8.44 (s, 1H, CH), 7.93 (s, 1H, CH), 7.09 (bs, 2H, NH2),
1-(4-Bromobutyl)-1,1,3,3,3-pentamethyldisiloxane (11). From 4.14 (t, J = 7.0 Hz, 2H, NCH2), 1.81 (m, 2H, NCH2CH2), 1.24 (m,
1,1,3,3,3-pentamethyldisiloxane (2.40 g, 16.0 mmol) and 4- 2H, SiCH2CH2), 0.47 (m, 2H, SiCH2), 0.15 (s, 6H, Si(CH3)2),
bromo-1-butene (2.0 cm3, 19.9 mmol); yield 2.80 g, 61%. 0.07 (s, 9H, SiCH3); 13C NMR (75 MHz, CDCl3): 155.8 (CNH2),
1H NMR (300 MHz, CDCl3): 3.35 (t, J = 7.0 Hz, 2H, CH2Br), 1.83 152.2 (CH), 149.4 (NCN), 140.6 (CH), 118.6 (NCN), 42.3 (NCH2),
(m, 2H, CH2CH2Br), 1.45 (m, 2H, SiCH2CH2), 0.52 (m, 2H, 32.6 (NCH2CH2), 19.5 (SiCH2CH2), 16.9 (SiCH2), 1.7 (Si(CH3)2),
SiCH2), 0.07 (s, 6H, Si(CH3)2), 0.06 (s, 9H, Si(CH3)3); 13C NMR 0.1 (Si(CH3)3); 29Si NMR (60 MHz, CDCl3): 7.1, 7.6 (OSiMe2,
(75 MHz, CDCl3): 35.3 (CH2Br), 32.8 (CH2CH2Br), 21.1 OSiMe3); analysis, calcd for C14H27N5OSi2: C 49.8, H 8.06, N
(SiCH2CH2), 16.5 (SiCH2), 0.7 (Si(CH3)2), 0.6 (Si(CH3)3); 20.8%, found C 49.3, H 7.93, N 21.7%.
29Si NMR (60 MHz, CDCl3): 7.6 (Me3SiO), 7.0 (OSiMe2); analy-
sis, calcd for C9H23Si2OBr: C 38.2, H 8.18%, found C 38.0, pyrimidine-2,4(1H,3H)-dione (16). From
H 8.01%.
(40 cm3) of thymine (0.46 g, 3.65 mmol), potassium carbonate
1-(4-(1,1,1,3,5,5,5-Heptamethyltrisiloxan-3-yl)butyl)-5-methyl-
DMSO solution
a
1-(4-Bromobutyl)-1,1,3,3-tetramethyl-3-phenylldisiloxane (0.62 g, 4.50 mmol) and 3-(4-bromobutyl)-1,1,1,3,5,5,5-hepta-
(12). From 1,1,3,3,-tetramethyl-3-phenyldisiloxane (3.70 g, methyltrisiloxane (13; 0.50 g, 1.40 mmol). Yield: 0.06 g, 11%.
16.0 mmol) and 4-bromo-1-butene (2.0 cm3, 19.9 mmol); yield 1H NMR (300 MHz, CDCl3): 8.58 (bs, 1H, NH), 6.96 (s, 1H, CH),
1
3.10 g, 56%. H NMR (300 MHz, CDCl3): 7.45 (m, 5H, C6H5), 3.68 (t, J = 7.0 Hz, 2H, NCH2), 1.91 (s, 3H, CH3), 1.70 (m, 2H,
3.35 (t, J = 6.9 Hz, 2H, CH2Br), 1.83 (m, 2H, CH2CH2Br), 1.45 NCH2CH2), 1.35 (m, 2H, SiCH2CH2), 0.54 (m, 2H, SiCH2), 0.10
(m, 2H, SiCH2CH2), 0.52 (m, 2H, SiCH2), 0.33 (s, 6H, Si(CH3)2), (s, 18H, Si(CH3)3), 0.09 (s, 3H, SiCH3); 13C NMR (75 MHz,
0.05 (s, 6H, Si(CH3)3); 13C NMR (75 MHz, CDCl3): 139.0 (C6H5), CDCl3): 162.2 (CvO), 148.8 (CvO), 138.6 (CH), 108.6 (CCH3),
132.2 (C6H5), 132.1 (C6H5), 128.5 (C6H5), 128.4 (C6H5), 46.7 (NCH2), 30.5 (NCH2CH2), 18.3 (SiCH2CH2), 15.3 (CH3),
127.0 (C6H5), 35.2 (CH2Br), 32.6 (CH2CH2Br), 21.0 (SiCH2CH2), 10.5 (SiCH2), −1.9 (SiCH3), −2.13 (Si(CH3)3); 29Si NMR
16.5 (SiCH2), 0.6 (Si(CH3)2), 0.4 (PhSi(CH3)2); 29Si NMR (60 MHz, CDCl3): 7.4 (SiMe3), −22.3 (OSiMe); analysis calcd for
(60 MHz, CDCl3): 8.6 (PhMe2SiO), 7.3 (O-SiMe2CH2); analysis, C16H34N2O4Si3: C 47.7, H 8.51, N 6.96%, found C 46.9, H 8.24,
calcd for C14H25Si2OBr: C 48.7, H 7.30%, found C 49.0, N 6.96%.
H 7.30%.
9-(4-(1,1,1,3,5,5,5-Heptamethyltrisiloxan-3-yl)butyl)-9H-purin-
3-(4-Bromobutyl)-1,1,1,3,5,5,5-heptamethyltrisiloxane (13). 6-amine (17). From a DMSO solution (40 cm3) of adenine
From 1,1,1,3,5,5,5-heptmethyltrsiloxane (3.00 g, 13.5 mmol) (0.40 g, 2.96 mmol), potassium carbonate (0.48 g, 3.52 mmol)
and 4-bromo-1-butene (1.4 cm3, 15.5 mmol); yield 3.80 g, 79%. and 3-(4-bromobutyl)-1,1,1,3,5,5,5-heptamethyltrisiloxane (13;
1H NMR (300 MHz, CDCl3): 3.31 (t, J = 7.0 Hz, 2H, CH2Br), 1.77 0.39 g, 1.09 mmol). Yield: 0.08 g, 11%. 1H NMR (300 MHz,
(m, 2H, CH2CH2Br), 1.38 (m, 2H, SiCH2CH2), 0.38 (m, 2H, CDCl3): 8.30 (s, 1H, CH), 7.74 (s, 1H, CH), 6.66 (br, 2H, NH2),
SiCH2), −0.08 (s, 18H, Si(CH3)3), −0.09 (s, 3H, O2Si(CH3)); 4.14 (t, J = 7.0 Hz, 2H, NCH2), 1.86 (m, 2H, NCH2CH2), 1.32 (m,
13C NMR (75 MHz, CDCl3): 34.1 (CH2Br), 31.6 (CH2CH2Br), 19.9 2H, SiCH2CH2), 0.46 (m, 2H, SiCH2), 0.07 (s, 18H, Si(CH3)3,
(SiCH2CH2), 14.6 (SiCH2), −2.2 (O2Si(CH3)), −2.4 (Si(CH3)3); 0.06 (s, 3H, SiCH3); 13C NMR (75 MHz, CDCl3): 155.1 (CNH2),
29Si NMR (60 MHz, CDCl3): 7.5 (Me3SiO), −22.5 (O2SiMe); 153.0 (CH), 149.2 (NCN), 140.0 (CH), 118.6 (NCN), 41.8 (NCH2),
analysis, calcd for C11H29Si3O2Br: C 37.0, H 8.18%, found 31.5 (NCH2CH2), 18.5 (SiCH2CH2), 15.2 (SiCH2), 1.83 (SiCH3,
C 37.4, H 8.18%.
5-Methyl-1-(4-(1,1,3,3,3-pentamethyldisiloxanyl)butyl)pyr- (OSiMe); analysis, calcd for C16H33N5O2Si3: C 46.7, H 8.08,
imidine-2,4(1H,3H)-dione (14). From DMSO solution N 17.0%, found C 45.5, H 7.92, N 17.0%.
(40 cm3) of thymine (1.20 g, 9.54 mmol), potassium carbonate
9,9′-[(1,1,3,3-Tetramethyldisiloxane-1,3-diyl)dibutane-4,1-
(1.57 g, 11.30 mmol) and 1-(4-bromobutyl)-1,1,3,3,3-penta- diyl]bis(9H-purin-6-amine) (18). From DMSO solution
1.02 (Si(CH3)3; 29Si NMR (60 MHz, CDCl3): 7.3 (Me3SiO), −22.0
a
a
methyldisiloxane (11; 0.50 g, 1.77 mmol). Yield: 0.08 g, 14%. (40 cm3) of adenine (0.91 g, 6.72 mmol), potassium carbonate
1H NMR (300 MHz, CDCl3): 8.62 (bs, 1H, NH), 7.01 (s, 1H, CH), (1.11 g, 7.99 mmol) and 1,3-bis(4-bromobutyl)-1,1,3,3-tetra-
3.68 (t, J = 7.0 Hz, 2H, NCH2), 1.96 (s, 3H, CH3), 1.70 (m, 2H, methyldisiloxane (10; 0.50 g, 1.24 mmol). Yield: 0.10 g, 16%.
NCH2CH2), 1.35 (m, 2H, SiCH2CH2), 0.54 (m, 2H, SiCH2), 0.04 1H NMR (300 MHz, CDCl3): 8.33 (s, 1H, CH), 7.74 (s, 1H, CH),
(s, 6H, SiCH3), −0.01 (s, 9H, SiCH3); 13C NMR (75 MHz, 7.16 (br, 2H, NH2), 4.13 (t, J = 7.0 Hz, 2H, NCH2), 1.79 (m, 2H,
CDCl3): 162.1 (CvO), 148.7 (CvO), 138.4 (CH), 108.5 (CCH3), NCH2CH2), 1.45 (m, 2H, SiCH2CH2), 0.45 (m, 2H, SiCH2),
46.3 (NCH2), 30.6 (NCH2CH2), 18.3 (SiCH2CH2), 15.9 (CH3), −0.08 (s, 6H, Si(CH3)2); 13C NMR (75 MHz, CDCl3): 155.7
10.4 (SiCH2), −1.7 (SiCH3), −2.0 (SiCH3); 29Si NMR (60 MHz, (CNH2), 152.1 (CH), 149.3 (NCN), 140.6 (CH), 118.5 (NCN), 42.2
CDCl3): 7.0, 7.6 (OSiMe2, OSiMe3); analysis, calcd for (NCH2), 32.5 (NCH2CH2), 19.5 (SiCH2CH2), 16.8 (SiCH2), −0.1
C14H28N2O3Si2: C 51.2, H 8.59, N 8.62%, found C 50.7, H 8.32, (Si(CH3)2); 29Si NMR (60 MHz, CDCl3): 7.0, 6.4 (SiMe2); analy-
N 8.62%.
sis, calcd for C22H36N10OSi2: C 51.5, H 7.08, N 27.3%, found
9-(4-(1,1,3,3,3-Pentamethyldisiloxanyl)butyl)-9H-purin-6- C 49.8, H 6.94, N 28.4%.
amine (15). From a DMSO solution (40 cm3) of adenine
6,6,8,8,15-Pentamethyl-7-oxa-1,13-diaza-6,8-disila-bicyclo-
(1.29 g, 9.54 mmol), potassium carbonate (0.48 g, 3.52 mmol) [11.3.1]heptadec-15-ene-14,17-dione (19) and 1,1′-[(1,1,3,3-tetra-
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 7734–7746 | 7737

View Article Online
Paper
Dalton Transactions
methyldisiloxane-1,3-diyl)dibutane-4,1-diyl]bis(5-methyl- (m, 4H, SiCH2), 0.01 (s, 12H, Si(CH3)2); 13C NMR (75 MHz,
pyrimidine-2,4(1H,3H)-dione) (20). A DMSO solution (40 cm3) CDCl3): 165.4 (CvNH), 156.1 (CvO), 145.3 (CH), 93.2 (CH),
of thymine (0.85 g, 6.72 mmol), potassium carbonate (1.11 g, 54.3 (1NCH2), 48.7 (3NCH2), 34.0 (1NCH2CH2) 32.3
7.99 mmol) and 1,3-bis(4-bromobutyl)-1,1,3,3-tetramethyldi- (3NCH2CH2), 21.5 (1NCH2CH2CH2), 20.2 (3NCH2CH2CH2), 16.9
siloxane (10; 0.50 g, 1.44 mmol) was stirred at room temperature (SiCH2), 0.2 (Si(CH3)2); 29Si NMR (60 MHz, CDCl3): 7.2, 6.6
for 72 h, after which a white precipitate of potassium bromide (SiOSi); analysis calcd for C16H31N3O2Si2: C 54.4, H 8.44, N
formed, considerably thickening the solution. The solution 11.9%, found C 54.1, H 9.33, N 11.9%.
was filtered and the residue washed with further DMSO. 19
After removal of 21 by filtration, 22 was isolated in the
crystallised first from solution on standing at room tempera- same manner at 20, yield: 0.05 g, 9%. (22): 1H NMR (300 MHz,
1
ture for 48 h (0.04 g, 8%). H NMR (300 MHz, CDCl3): 6.99 (s, CDCl3): 8.01 (s, 2H, NH2), 7.10 (d, J = 6.9 Hz, 1H, CH), 5.78 (d,
1H, CH), 4.16 (t, J = 7.0 Hz, 4H, NCH2), 1.99 (s, 3H, CH3), 1.74 J = 6.9 Hz, 1H, CH), 4.24 (t, J = 6.9 Hz, 2H, N-CH2), 1.74 (m,
(m, 4H, NCH2CH2), 1.35 (m, 4H, SiCH2CH2), 0.52 (m, 4H, 2H, NCH2CH2), 1.35 (m, 2H, SiCH2CH2), 0.51 (m, 2H, SiCH2),
SiCH2), 0.07 (s, 12H, SiCH3); 13C NMR (75 MHz, CDCl3): 163.1 0.04 (s, 6H, Si(CH3)2); 13C NMR (75 MHz, CDCl3): 165.9
(CvO), 151.9 (CvO), 137.2 (CH), 109.4 (CCH3), 46.7 (1NCH2), (CNH2), 156.6 (CvO), 146.0 (CH), 93.7 (CH), 50.1 (NCH2), 32.6
39.9 (3NCH2), 32.9 (1NCH2CH2), 30.9 (3NCH2CH2), 19.3 (NCH2CH2), 20.4 (SiCH2CH2), 18.0 (SiCH2), 0.6 (Si(CH3)2); 29Si
(1NCH2CH2CH2), 18.8 (3NCH2CH2CH2), 17.4 (CH3), 12.5 NMR (60 MHz, CDCl3): 7.1, 6.5 (SiOSi); analysis calcd for
(SiCH2), 0.0 (Si(CH3)2); 29Si NMR (60 MHz, CDCl3): 7.1, 6.5 C20H36N6O3Si2: C 51.7, H 7.81, N 18.1%, found C 50.8, H 8.49,
(SiOSi); analysis calcd for C17H32N2O3Si2: C 55.4, H 8.75, N N 16.9%.
7.60%, found C 55.4, H 8.82, N 7.67%.
Bromobutyl-functionalised poly(dimethylsiloxane) (23). Tar-
After isolation of 19 by filtration, the DMSO was removed getting a polymer with 0.2% bromobutyl functionality, a
under reduced pressure followed by washing three times with mixture of 3-(4-bromobutyl)heptamethyltrisiloxane (13; 0.1 g)
water (50 cm3). Thin layer chromatography of the resulting and poly(dimethylsiloxane) (PDMS; 20.4 g) with molecular
white powder, using a mixture of dichloromethane and hexane weight ca. 50 000 was stirred at 70 °C with a catalytic amount
(98 : 2), suggested a mixture of three remaining species, at r = of glacial acetic acid and water for 120 h during which time
1.00–0.95, 0.70 and 0.65–0.40. Chromatography on silica using the viscosity increased markedly. The viscous oil was dissolved
the same eluent yielded three bands in each case. The first, in dichloromethane, washed with sodium hydrogencarbonate
eluted with the solvent front, was found to be trace amounts of solution until no further effervescence was observed, dried
unreacted siloxane (10). The second was found to be trace over anhydrous MgSO4 and filtered. Removal of the volatiles
amounts of DMSO, and the third yielded the desired product afforded a viscous, air stable, colourless oil. Yield 84%, Mn
1
20 as a white powder (0.06 g, 8%). A white powder remaining (GPC) 23700, PDI 3.7. PDMS (0.2% bromobutyl) (23): H NMR
uneluted was assumed to be the unreacted thymine. Recrystal- (300 MHz, CDCl3): 3.40 (t, J = 6.7 Hz, 2H, CH2Br), 1.88 (m, 2H,
lization was attempted from dichloromethane, chloroform, CH2CH2Br), 1.39 (m, 2H, SiCH2CH2), 0.40 (m, 2H, SiCH2), 0.07
ethyl acetate and a mixture of dichloromethane and hexane, (s, Si(CH3)2); 29Si NMR (60 MHz, CDCl3): −22.0.
but crystals of the product were not obtained. 1H NMR
Thymine-functionalised PDMS (24). A mixture of 23 (1.0 g)
(300 MHz, CDCl3): 8.57 (br, 1H, NH), 6.98 (s, 1H, CH), 3.70 (t, and 50 mg of each of thymine and potassium carbonate was
J = 7.0 Hz, 2H, NCH2), 1.92 (s, 3H, CH3), 1.70 (m, 2H, stirred at room temperature in DMSO (50 cm3) for 18 h. A
NCH2CH2), 1.34 (m, 2H, SiCH2CH2), 0.54 (m, 2H, SiCH2), 0.03 white, water soluble precipitate formed in the reaction which
(s, 6H, Si(CH3)2); 13C NMR (75 MHz, CDCl3): 163.1 (CvO), was removed by filtration. Removal of volatiles under reduced
151.9 (CvO), 137.2 (CH), 109.4 (CCH3), 46.7 (1NCH2), 39.9 pressure resulted in a viscous, cloudy oil which was twice dis-
(3NCH2), 32.9 (1NCH2CH2), 30.9 (3NCH2CH2), 19.3 solved in chloroform and washed repeatedly with water before
(1NCH2CH2CH2), 18.8 (3NCH2CH2CH2), 17.4 (CH3), 12.5 the organic layers were combined and dried over anhydrous
(SiCH2), −0.03 (Si(CH3)2); 29Si NMR (60 MHz, CDCl3): 7.1, 6.4 magnesium sulphate. Removal of the volatiles afforded a very
(SiOSi); analysis calcd for C22H38N4O5Si2: C 53.4, H, 7.74, N viscous, air stable, colourless oil. Yield 55%. Mn (GPC) 22100,
1
11.3%, found C 52.3, H 7.65, N 10.9%.
Also prepared by the same method were:
PDI 4.1. PDMS (0.2% butylthymine) (24): H NMR (300 MHz,
CDCl3): 6.97 (s, 1H, CH), 3.70 (m, 2H, NCH2), 1.92 (s, 3H,
16-Imino-6,6,8,8-tetramethyl-7-oxa-1,13-diaza-6,8-disila- CCH3), 1.70 (m, 2H, NCH2CH2), 1.39 (m, 2H, SiCH2CH2), 0.88
bicyclo[11.3.1]heptadec-14-en-17-one (21) and 1,1′-[(1,1,3,3- (m, 2H, SiCH2), 0.07 (s, Si(CH3)2); 29Si NMR (60 MHz, CDCl3):
tetramethyldisiloxane-1,3-diyl)dibutane-4,1-diyl]bis(4-aminopyr- −22.0.
imidin-2(1H)-one) (22). From a DMSO solution (40 cm3) of
Adenine-functionalised PDMS (25). A mixture of 23 (1.0 g)
cytosine (0.75 g, 6.72 mmol), potassium carbonate (1.11 g, and 50 mg of each of adenine and potassium carbonate was
7.99 mmol) and 1,3-bis(4-bromobutyl)-1,1,3,3-tetramethyldi- stirred at room temperature in DMSO (50 cm3) for 18 h. A
siloxane (10; 0.50 g, 1.24 mmol). 21 crystallised from solution white, water soluble precipitate was removed by filtration.
over a 48 h period (as with 19), yield: 0.03 g, 7%. 1H NMR Removal of volatiles under reduced pressure resulted in a
(300 MHz, CDCl3): 7.66 (d, J = 7.0 Hz, 1H, CH), 6.45 (br s, 1H, viscous, cloudy oil which was twice dissolved in chloroform
CvNH), 5.92 (d, J = 7.0 Hz, 1H, CH), 4.19 (t, J = 7.0 Hz, 4H, and washed repeatedly with water before the organic layers
NCH2), 1.74 (m, 4H, NCH2CH2), 1.41 (m, 4H, SiCH2CH2), 0.60 were combined and dried over anhydrous magnesium
7738 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Table 1 Crystal data for 4, 5, 8, 9, 19 and 21
Compound reference
4
5
8
9
19
21
Chemical formula
Formula mass
Crystal system
C
11H18N2O2Si
C17H24N2O2Si
316.47
Triclinic
C16H23N3OSi
301.46
Monoclinic
9.5090(15)
C19H28N5Ni0.5O2S3Si1
480.03
Triclinic
C17H31N2O3Si2
367.62
Triclinic
C16H32N3O2Si2
354.62
Triclinic
238.36
Orthorhombic
13.0223(2)
8.0964(4)
8.3500(4)
7.6867(3)
7.1751(2)
a/Å
b/Å
c/Å
α/°
β/°
17.4770(3)
12.0846(2)
8.3765(5)
13.0731(7)
98.284(3)
92.095(3)
99.840(2)
862.75(8)
40.651(6)
8.7418(13)
11.0160(6)
14.6670(8)
103.051(3)
94.631(3)
107.523(2)
1237.12(11)
ˉ
P1
8.6444(4)
17.9501(11)
85.273(2)
79.127(2)
64.283(2)
1055.30(9)
8.9309(3)
15.8082(5)
96.032(1)
96.808(1)
95.324(2)
994.58(5)
98.108(3)
γ/°
Unit cell volume/Å3
Space group
2750.34(8)
Pna21
8
3345.3(9)
P21/c
8
ˉ
ˉ
ˉ
P1
2
P1
2
P1
2
Z
2
No. of refl’ns measured
No. of independent refl’ns
Rint
Final R1 (I > 2σ(I))
Final wR(F2) (I > 2σ(I))
Final R1 (all data)
Final wR(F2) (all data)
Goodness of fit on F2
Flack parameter
CCDC number
46 491
6224
9597
3883
21 914
5856
12 877
5522
11 515
3796
15 784
4534
0.0722
0.0379
0.0819
0.0573
0.0897
1.033
0.0449
0.0465
0.1191
0.0579
0.1288
1.043
0.0641
0.0585
0.1540
0.0885
0.1759
1.052
0.0537
0.0722
0.2032
0.1135
0.2361
1.034
0.0632
0.0593
0.1324
0.0943
0.1528
1.029
0.0463
0.0555
0.1087
0.0781
0.1156
1.107
0.01(6)
956088
956089
956090
956091
956092
956093
sulphate. Removal of the volatiles afforded a viscous, air the refinement, as the disorder proximate to these centres has
stable, colourless oil. Yield 48%. Mn (GPC) 21300, PDI 5.6. resulted in substantial smearing of the electron density. Aniso-
PDMS (0.2% butyladenine) (25): 1H NMR (300 MHz, CDCl3): tropic refinement was confined to the fractional occupancy
8.02 (br, 2H, NH2), 6.96 (s, 1H, CH), 6.91 (s, 1H, CH), 4.11 (t, silicons and C(12)–C(13) in this disordered region. 21: There is
J = 6.0 Hz, 2H, NCH2N), 1.70 (m, 2H, NCH2CH2), 1.34 (m, 2H, disorder in the hetero-aromatic ring over two sites in the ratio
SiCH2CH2), 0.88 (m, 2H, SiCH2), 0.07 (s, Si(CH3)2); 29Si NMR 70 : 30, the major component of which was refined anisotropi-
(60 MHz, CDCl3): −22.0.
cally. Hydrogen atoms on N(2) [and its disordered analogue N(2A)]
and C(8) (which is involved in hydrogen bonding) could not be
located and had to be placed on calculated positions.
Crystallography
Experimental details relating to the single-crystal X-ray crystal-
lographic studies are summarised in Table 1. Data for 4, 5, 9,
19 and 21 were collected on a Nonius Kappa CCD diffracto-
meter at 150(2) K using Mo-Kα radiation (λ = 0.71073 Å), while
that for 8 was collected at 150(2) K using the synchrotron radi-
ation (λ = 0.846 Å) at Daresbury Station 9.8. Structure solution
was followed by full-matrix least squares refinement and was
performed using the WinGX-1.70 suite of programmes.29
Unless stated otherwise, all non-hydrogen atoms were refined
anisotropically, while hydrogen atoms were added in calcu-
lated positions save those involved in hydrogen bonding,
which were located and freely refined. Specific details are as
follows: 4 has two independent molecules in the asymmetric
unit while in 8 the two molecules in the asymmetric unit are
connected to each other via hydrogen bonding. 9: The asym-
metric unit consists of half a nickel complex (in which the
metal centre is located at a crystallographic inversion centre)
and one heavily disordered silylated thymine unit. The flexible
arm of the thymine functionality [C(9)–C(19) and the silicon]
exhibited extensive disorder which is modelled over 3 sites in a
40 : 30 : 30 ratio. Only the phenyl component attached to the
40% occupancy silicon [Si(1)] could be located with any
reliability, and this is included subject to the restraint of being
treated as a rigid hexagon. In the cases of Si(1A) and Si(1B),
only the ipso carbon of the phenyl component is included in
Results and discussion
The synthetic strategy we have followed is summarised in
outline in Scheme 1 for thymine. There are two complemen-
tary routes to the target polymers, each starting from a
Scheme 1
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 7734–7746 | 7739

View Article Online
Paper
Dalton Transactions
trimethylsilyl-protected base, which affords better regiospecifi- ing β-silylation is minimal at best, presumably as a result of
city of the subsequent reactions.14 The first protocol involves the bulky thymine substituent close to the β-carbon of the
N-alkenylation of the base followed by protection and hydro- olefin. Deprotection of the O-SiMe3 group takes place on the
1
silylation, while the alternative is to form an α-silyl-ω-bro- silica column. The H NMR of 5 includes resonances at 3.67,
moalkane by hydrosilysilyation of an appropriate alkene a 1.68, 1.33 and 0.78 ppm corresponding to the saturated
priori, with attachment of the nucleobase as a final step. N(CH2)4Si group and loss of the signals in 2 at 5.76 and
Where possible we have attempted similar reactions on all four 5.10 ppm due to the alkenyl hydrogens. The 29Si NMR of 5 is a
nucleoside bases, but in general the chemistry of guanine (and single peak at −3.0 ppm, typical of a C4Si environment and
to a slightly lesser extent, adenine) has proved the least tract- reflecting the purity of the sample.
able due to solubility considerations, while thymine has given
the broadest range of characterisable products.
Synthesis and structural chemistry of N-1-alkenyl- and
N-1-(organosilyl)alkyl thymines
ð2Þ
ð1Þ
The structures of 4 (Fig. 1) and 5 (Fig. 2) are of interest for
the insights they provide on the self-assembly of the pyrimi-
dine base.
The asymmetric unit of 4 is formed of two independent
molecules held together by a C–H⋯O hydrogen bond. The silyl
group has protected the less sterically hindered of the carbonyl
moieties [Si(1)–O(2) 1.6893(18), O(4)–Si(2) 1.686(2) Å], with the
silicon atom in the plane of the heterocycle. The Si–O bond
length compares well to the complementary group in bis(tri-
methylsilyl)thymine [1.6951(16) Å],24 though it is long in com-
parison to similar bonds in compounds such hexa-
(trimethylsiloxy)benzene [Si–O 1.655(3), 1.673(4) Å].33 The
heterocyclic ring is moderately distorted [C(7)–N(2) 1.305(4),
C(7)–C(5) 1.418(4) Å] due to the steric bulk of the SiMe3 group
The N-1-alkenylthymine derivatives 1–3 have been synthesised
from the reaction of silyl-protected thymine with the appropri-
ate alkenylbromide. An excess of alkenyl bromide was refluxed
with the silylated base until the IR spectrum of the reaction
mixture displayed no further generation of the product,
measured by monitoring the evolution of a peak at 1720 cm−1
due to the carbonyl group in thymine. After removal of the
volatiles, thick liquids remained from which 1 crystallised in
good yield (81%), while 2 and 3 were first separated by chrom-
atography from excess reagent before crystallization; yields fell
markedly with increasing chain length (59, 18% for 2, 3
respectively). 1,14,26 2,27,30 328,30 and analogues with longer
alkenyl chains (n = 9; C11)16,17 have been prepared by others,
usually using NaH as base to deprotonate the thymine, though
there appears to be a wide variation in literature melting
points cited for these products. Furthermore, the use of an
unprotected base leads to a mixture of regio-isomers with sub-
stitution at both N-1 and N-3,31 something that was not seen
in our synthesis where the steric protection offered by the
Me3Si groups directs substitution to N-1 rather than the more
acidic32 (but now more hindered) N-3 site. NMR data for 1 and
3 are similar to those reported by others, while spectral charac-
terization of 2 is reported for the first time.
Fig. 1 The structure of 4 showing the atom labelling scheme; ellipsoids
are at the 60% level. Selected geometric data for the molecule based on
As a representative example, 1 was then re-protected by Si(1) as representative: Si(1)–O(2) 1.6893(18), N(1)–C(4) 1.363(4), N(1)–
C(8) 1.395(3), N(1)–C(3) 1.474(2), N(2)–C(7) 1.305(4), N(2)–C(8) 1.371(4),
O(1)–C(8) 1.237(3), O(2)–C(7) 1.344(3), C(1)–C(2) 1.291(5), C(2)–C(3)
1.501(4), C(4)–C(5) 1.348(4), C(5)–C(7) 1.418(4), C(5)–C(6) 1.503(4) Å;
O(2)–Si(1)–C(10) 110.20(13), O(2)–Si(1)–C(9) 111.27(13), C(10)–Si(1)–C(9)
O-silylation using (Me3Si)2NH/(NH4)2SO4 to yield 4 in almost
quantitative yield; protection occurs only at the unhindered
C-4 carbonyl. Given the success of this reaction, 2, with a
longer chain and more exposed alkenyl group, was then re-
111.87(17), O(2)–Si(1)–C(11) 99.98(13), C(10)–Si(1)–C(11) 111.17(16), C(9)–
protected (4a; n = 2; not isolated) and hydrosilylated using Si(1)–C(11) 111.77(17), C(4)–N(1)–C(8) 121.4 (2), C(7)–N(2)–C(8) 118.7 (2),
C(7)–O(2)–Si(1) 126.31(17), C(1)–C(2)–C(3) 123.0(3), N(1)–C(3)–C(2)
111.5(2), C(5)–C(4)–N(1) 121.4(2), C(4)–C(5)–C(7) 114.6(2), N(2)–C(7)–
C(5) 125.8 (2), N(2)–C(8)–N(1) 117.8(2)°. Hydrogen bond data: H(24)–
O(1) 2.29, H(4’’)–O(3) 2.34 Å, ∠C(24)–H(24)–O(1) 158.8°, ∠C(4)–H(4)–
PhMe2SiH in the presence of a catalytic amount of chloro-
platinic acid to afford the N-phenyldimethylsilylbutyl derivative
of thymine (5) in 51% yield. The reaction was monitored by
the loss of the Si–H stretch at 2160 cm−1 in the IR spectrum.
O(3) 164.0°. Symmetry operations: ’ 3/2 − x, 1/2 + y, z − 1/2; ’’ 3/2 − x,
Chromatography indicates only one product is formed, imply- 1/2 − y, 1/2 + z.
7740 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 2 The structure of 5 showing the atom labelling scheme; ellipsoids
are at the 60% level. Selected geometric data: Si–C(9) 1.8772(18),
O(1)–C(1) 1.2378(19), O(2)–C(4) 1.2198(19), N(1)–C(1) 1.381(2), N(1)–C(4)
1.3862(19), N(2)–C(3) 1.374(2), N(2)–C(4) 1.374(2), N(2)–C(6) 1.4778(19),
C(1)–C(2) 1.445(2), C(2)–C(3) 1.353(2), C(2)–C(5) 1.499(2) Å; C(1)–N(1)–
C(4) 126.83(14), C(4)–N(2)–C(3) 121.48(13), N(1)–C(1)–C(2) 115.64(13),
C(3)–C(2)–C(1) 117.59(14), C(2)–C(3)–N(2) 123.80(15), N(2)–C(4)–N(1)
114.63(13)°. Hydrogen bond data: H(1)–O(1’) 1.97(2) Å, ∠N(1)–H(1)–O(1’)
173(2)°. Symmetry operation: 2 − x, 1 − y, 1 − z.
Scheme 2
reaction with the silylated thymine, adenine and cytosine
bases (Scheme 2).
relative to that of thymine [C(7)–N(2) 1.401(5), C(7)–C(5) 1.453(4)
Å].34 In order to accommodate the bulk of the SiMe3
group, the C–O–Si bond angle widens from the tetrahedral
[C(7)–O(2)–Si(1) 126.31(17), C(27)–O(4)–Si(2) 126.77(18)°], to a
marginally greater extent than in bis(trimethylsilyl)thymine
[C–O–Si 125.46(14)°].24 The molecule forms a polymeric chain
in the b-direction, interacting via other CH⋯O hydrogen
bonds [H(4)⋯O(3) 2.34 Å, ∠C(4)–H(4)⋯O(3) 164.0°; H(24)⋯O(1)
2.29 Å, ∠C(24)–H(24)⋯O(1) 158.8°], of comparable length with
those in bis(trimethylsilyl)cytosine [H(1)⋯O(2) 2.26(3) Å],24
though longer than similar examples of C–O⋯H hydrogen
bonds, such as in 1-methylcytosine [H(1)⋯O(2) 2.04(2) Å].35
Hydrosilylation of 4a took place at the exocyclic olefin in
the favoured α-orientation to form a C–Si single bond [C(9)–Si(1)
1.8772(18) Å] in 5. There appears to be no distortion of the
bond distances in the heterocyclic ring [C(4)–N(1) 1.3862(19),
C(1)–N(1) 1.381(2), C(1)–C(2) 1.445(2) Å] relative to thymine [C(4)–
N(1) 1.361(4), C(1)–N(1) 1.401(5), C(1)–C(2) 1.453(4) Å],34
and the N–CH2 bond distance in 5 [N(2)–C(6) 1.4778(19) Å]
compares well with the analogous distance in 4 [1.474(2) Å].
The alkyl moiety does not appear to have caused the hetero-
cyclic ring to distort [C(3)–N(2)–C(4) 121.48(13), N(1)–C(4)–N(2)
114.63(13), C(1)–N(1)–C(4) 126.83(14)°] relative to the same
angles in thymine [C(3)–N(2)–C(4) 122.5(3), N(1)–C(4)–N(2)
115.5(2), C(1)–N(1)–C(4) 126.3(3)°].34 In contrast to 4, loss of
the O-silyl protection in 5 allows the bases to form pairs
through more conventional NH⋯OvC hydrogen bond pairs.
[H(1)⋯O(1) 1.97(2) Å, ∠N(1)–H(1)⋯O(1′) 173(2)°], which are
short compared to the analogous bonds in similar com-
pounds, such as 1-methylcytosine [H(1)⋯O(2) 2.04(2) Å]35 and
bis-(trimethylsilyl)cytosine (2) [2.26(3) Å].24
Compound 5 is prepared in better yield (74%) by this route
than via hydrosilylation of N-butenylthymine (2; 51%). Similar
reactions with silyl-protected adenine and cytosine afforded 7
and 8 in modest yield (55, 45%, respectively), with single-site
substitution at the adenine N-9 and cytosine N-1 positions.32,37
The NMR data for 7, 8 are unexceptional; both show similar
29Si chemical shifts (−3.1 ppm) to 5, 6 (3.0 ppm).
The asymmetric unit of 8 is formed of two independent
molecules of 8 (Fig. 3). The substitution is confirmed as
having taken place N-1, to form an N-alkyl moiety [N(1)–C(5)
1.477(3), N(4)–C(21) 1.474(3) Å] with the N–C bond comparable
to those of 4 [1.474(2) Å] and 5 [1.4778(19) Å], as is the H2C–Si
bond distance [8: C(1)–Si(1) 1.881(3), C(24)–Si(2) 1.870(3); 5:
C(9)–Si 1.8772(18) Å]. The bond angles around the N-1 position
[C(3)–N(1)–C(4) 120.1(2), C(3)–N(1)–C(5) 120.5(2), C(4)–N(1)–
C(5) 119.3(2), C(19)–N(4)–C(20) 120.4(2), C(19)–N(4)–C(21)
120.2(2), C(20)–N(4)–C(21) 119.6(2)°] are very close to 120°, and
Fig. 3 The structure of 8 showing the atom labelling scheme; ellipsoids
are at the 65% level. Selected geometric data based on the molecule
containing Si(1) as representative: Si(1)–C(8) 1.881(3), O(1)–C(4) 1.242(3),
O(2)–C(20) 1.239(3), N(1)–C(3) 1.358(3), N(1)–C(4) 1.393(3), N(1)–C(5)
1.477(3), N(2)–C(1) 1.333(3), N(2)–C(4) 1.362(3), N(3)–C(1) 1.340(3), C(1)–
C(2) 1.423(3), C(2)–C(3) 1.350(3) Å; C(3)–N(1)–C(4) 120.1(2), C(1)–N(2)–
C(4) 119.7(2), N(2)–C(1)–C(2) 122.1(2), C(3)–C(2)–C(1) 117.0(2), C(2)–
C(3)–N(1) 121.5(2), N(2)–C(4)–N(1) 119.5(2)°. Hydrogen bond data:
H(3A)⋯N(5) 2.102(17) Å, ∠N(3)–H(3a)⋯N(5) 178(2)°; H(6A)⋯N(2) 2.112(17) Å,
∠N(6)–H(6a)⋯N(2) 174(2)°; H(3b)⋯O(2’’) 2.026(18) Å, ∠N(3)–H(3b)⋯O(2’’)
164(3)°; H(6B)⋯O(1’) 2.022(18) Å, ∠N(3)–H(3b)⋯O(2’) 163(3)°. Symmetry
operations: ’ 1 − x, −y, −z − 1; ’’ 2 − x, −y, −z − 2.
The alternative route to silylalkyl-substituted bases
(Scheme 1) involves formation of R3Si(CH2)n+2Br by a priori
hydrosilylation of H2CvC(H)(CH2)nBr followed by alkylation of
the protected bases in a manner analogous to eqn (1), using a
stoichiometric quantity of tetrabutylammonium fluoride as co-
reagent. Accordingly, PhMe2Si(CH2)4Br (6) was prepared in
74% yield by chloroplatinic acid-catalysed hydrosilylation of
H2CvC(H)CH2CH2Br. 6, which is a known compound,36 was
characterised by NMR, including δ(29Si) at −3.0 ppm, before
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 7734–7746 | 7741

View Article Online
Paper
Dalton Transactions
are less distorted than those in 5. The molecule forms a dimer others in the complex between dithiobiuret and uracil,38
held together by two NH⋯N hydrogen bonds [H(3A)⋯N(5) despite the general preference for thymine to form just two
2.102(17) Å, ∠N(3)–H(3a)⋯N(5) 178(2)°; H(6A)⋯N(2) 2.112(17) hydrogen bond interactions with its partner.40,41 It also follows
Å, ∠N(6)–H(6a)⋯N(2) 174(2)°], considerably shorter than those a similar pattern to other complexes in which thymine forms a
in bis-(trimethylsilyl)adenine [2.38(1) Å],24 but of comparable triple ADA : DAD hydrogen bond pair (D = donor, A = accep-
length with those in bis(trimethylsilyl)cytosine [2.26(3), 2.24(3) tor).42 There is a further network of intermolecular NH⋯O and
Å].24 Furthermore, additional H-bond donors allow dimers to notably weaker NH⋯S interactions which generate a chain of
further aggregate into a laddered structure via two symmetry- hydrogen-bonded donor : acceptor complexes propagating
related N–H⋯O hydrogen bonds [H(3b)⋯O(2″) 2.026(1) Å, along b. This is in contrast to the Ni(biuret)2·2uracil·2H2O
∠N(3)–H(3b)⋯O(2″) 164(3)°; H(6B)⋯O(1′) 2.022(18) Å, ∠N(3)– which forms a sheet structure via hydrogen bonding between
H(3b)⋯O(2′) 163(3)°], though these are longer than those in 5 N(5)–H (the hydrogen replaced by the alkylsilylgroup), O(1)
[1.97(2) Å].
and water (numbering as in Fig. 4).
The ability of compounds such as 5 and 8 to form more
For completeness, we have also determined the structure of
complex hydrogen-bonded networks was explored by co-crys- the related complex of nickel dithiobiuret with two equivalents
tallising the thymine derivative 5 with nickel dithiobiuret.25 of thymine (9a),43 which adopts a similar sheet structure to
5
was added to Ni(S2C2N3H4)2 in warm MeOH in 1 : 1 ratio, from that of hydrated Ni(biuret)2·2uracil but is far more compact
which Ni(S2C2N3H4)2·2(5) [9; 20% yield, m.p. 224 °C, pink crys- due to direct intermolecular hydrogen bonding between
tals] crystallised, despite the reagent stoichiometry. The struc- thymine units (rather than via water) and a concomitant puck-
ture of 9 is shown in Fig. 4.
ering of the sheet to accommodate the methyl group not
The structure of 9 shows the three-fold hydrogen bonding present in uracil; no further comment is necessary here.
between the complex and the base with the silylated alkyl
chains at the periphery of the molecule on either side. The
metal is itself square planar with Ni–S distances [Ni–S(1)
Synthesis of model siloxanes and their compounds with
pendant purine/pyrimidine bases
2.1466(9), Ni–S(2) 2.1671(9) Å] consistent with those in the two
In order to gain insights into the attachment and resulting
other nickel dithiobiurets reported previously [2.150–2.168 Å].38,39
structures of purine and pyrimidine bases linked to siloxane
The two biuret ligands are not coplanar but form a flattened
chains, we have synthesised several model siloxanes (10–13;
Scheme 3) for further elaboration.
chair conformation about the metal. The pattern of hydrogen
bonds between biuret and thymine shows two strong NH⋯O
4-Bromobut-1-ene, the silane and a catalytic amount of
and a weaker NH⋯N interaction,32 as has been observed by
H2PtCl6 were heated in propan-2-ol at 110 °C until υ(Si–H) at
ca. 2050 cm−1 had disappeared from the IR spectrum. After
hydrolysis, the mixture was extracted into CHCl3, dried and
purified by column chromatography; the product eluted with
the solvent front as a single isomer (α-addition) of 10–13 in
59–79% yield.
Thymine, adenine and cytosine were, in various combi-
nations, reacted with 10, 11 and 13 in DMSO in the presence
of K2CO3 as base, since the TBAF used in the coupling reac-
tions to form 10–13 would cause fission of the siloxane com-
ponent. The reactions carried out are summarised in Schemes
4 and 5.
Fig. 4 The structure of 9 showing the atom labelling scheme; ellipsoids
are at the 60% level. Only the major occupancy component of the three
disordered (CH2)3SiMe2Ph moieties is shown for clarity; labels associated
with the carbon atoms of the phenyl ring [C(14)–C(19)] have also been
omitted for the same reason. Selected geometric data: Ni–S(1) 2.1464(11),
Ni–S(2) 2.1665(11) Å, S(1)–Ni–S(1’) 180.0, S(1)–Ni–S(2) 94.71(4);
S(1)–Ni–S(2’) 85.29(4)°. Hydrogen bond data: H(1a)⋯O(1) 1.927(4) Å,
∠N(1)–H(1a)⋯O(1) 170.2(2)°; H(1b)⋯O(2’’’) 2.142(2) Å, ∠N(1)–H(1b)⋯O(2’’’)
165.9(2)°; H(3a)⋯O(2) 2.069(3) Å, ∠N(3)–H(3a)⋯O(2) 169.2(2)°: H(3b’’’)⋯S(2’)
2.921(1) Å, ∠N(3)–H(3b’’’)⋯S(2’) 159.37(4)°; H(4)⋯N(2) 2.207(3) Å, ∠N(4)–
H(4)⋯N(2) 175.5(2)°. Symmetry operations: ’ 1 − x, 1 − y, 1 − z; ’’ −x,
1 − y, 1 − z; ’’’ x − 1, y, z; ’’’’ x − 2, y, z; ’’’’’ 2 − x, 1 − y, 1 − z.
Scheme 3
7742 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
lene group is consistent with this being at the expected N-9
position [δ(1H) = 4.14, δ(13C) 42.3, 41.8 ppm, respectively; cf.
4.15, 43.4 ppm in N-9 substituted 7]. In the case of the two
thymine-containing products (14, 16) however, the 1H/13C
NMR shifts of the CH2N methylene group give evidence for
both dominant N-1 and secondary N-3 substitution in a ratio
ca. 4 : 1. The latter substitution is indicated by δ(1H) at ca.
3.91 ppm (ca. 3.70 ppm for N-1) and δ(13C) at ca. 39 ppm (ca.
46 ppm for N-1). For comparison, the 1H NMR spectrum of
1,3-dibutylthymine has CH2N signals at 3.66 and 3.90 ppm for
N-1 and N-3 butyl groups, respectively.44 The lack of a single
product in these cases is in contrast to the observed reactions
using silyl-protected bases in these reactions, as discussed
earlier.
The reaction chemistry involving 10, which bears two 4-bro-
mobutyl substituents, is more complex. In the case of adenine,
a single product (18) is formed, with N-9 attachment of two
bases at the termini of the two butyl chains. The CH2N NMR
signals [δ(1H) = 4.13; δ(13C) = 42.2 ppm] are very similar to
those in 15 and 17, while data for the compound analogous to
18 but embodying two propyl, rather than butyl, chains pre-
pared by others has NMR chemical shifts of 4.09, 45.7 ppm,
respectively.13
For both thymine and cytosine, two quite distinct products
are formed in each case. First to crystallise from solution are
the pyrimidinocyclophanes 19 and 21, respectively, in which a
single base has been substituted at both N-1 and N-3 positions
by the two alkyl chains of the disiloxane (10). In the case of 19,
the –(CH2)3– chains can clearly be distinguished in the 13C
NMR spectrum between N-1 bound [δ(13C): 46.7, 32.9.
19.3 ppm] and N-3 bound [δ(13C): 39.9, 30.9. 18.8 ppm] for the
three carbon centres closest to the heterocycle, though in the
1H NMR spectrum the associated multiplets overlap, are
poorly resolved and appear as a single signals at 4.16, 1.74 and
1.35 ppm. Similarly, the signals for 21 appear in the 13C NMR
at 54.3, 34.0, 21.5 ppm (N-1 bound) and 48.7, 32.3, 20.2 (N-3
bound) for the two –(CH2)3– chains. For 21, the second depro-
tonation of cytosine required for formation results in gen-
eration of an exocyclic imine, evidenced by the short N(2)–C(7)
distance [1.272(4) Å].
Scheme 4
After crystallization of the two pyrimidocyclophanes, two
further products, 20 and 22, can be isolated, in which pyrimi-
dine bases have attached at the ends of both butyl chains, in
an analogous manner to the formation of 18. Although
different regio-isomers could not be separated chromato-
graphically, NMR evidence points to a major N-1,N-1-substi-
tuted product, with a secondary N-1,N-3 isomer and traces of a
1
third, presumably N-3,N-3 substituted product; H NMR inte-
Scheme 5
grals suggest a 25 : 7 : 1 ratio of these products. The different
isomers are again most easily identified by differences in the
For siloxanes 11 and 13 which each contain only one butyl 1H/13C chemical shifts of the NCH2 group [N-1, δ(1H) 3.70;
side-chain, both thymine and adenine form one product in δ(13C) 46.7 ppm; N-3, δ(1H) 3.96; δ(13C) 40.0 ppm]. However, for
each case though the isolated yields of purified product are 22, only one set of resonances is visible, consistent with domi-
low (ca. 10%). The two products incorporating adenine (15, 17) nant N-1,N-1 substitution. In both 20 and 22, two 29Si reso-
have clean NMR spectra suggesting only one regio-isomer has nances could be seen for the SiOSi residue, similar to those
1
been formed, and the H/13C NMR shifts of the CH2N methy- seen in 18.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 7734–7746 | 7743

View Article Online
Paper
Dalton Transactions
In all the reactions shown in Schemes 4 and 5, the yields of
products 14–22 are low (thymine: 8–14%; adenine: 11–16%;
cytosine: 7–9%).
The structures of 19 and 21 are shown in Fig. 5 and 6,
respectively. The metrical parameters for each are unexcep-
tional, save that the Si–O–Si angle [19: 150.53(15); 21 150.38(10)°]
reflects the wide conformational flexibility associated with
this connectivity.45 In addition, the short C(7)–N(2) bond
in 21 [1.272(4) Å] confirms the amine to imine rearrangement
required to allow cyclophane formation. Both structures incor-
porate a 14-membered macrocycle which includes the siloxane
and spans the N-1 and N-3 positions on the base. There are,
however, differences between the supramolecular structures of
the two species.
In the case of 19, arising from the lack of any suitably
acidic hydrogens to promote hydrogen bonding, two sym-
metry-related molecules stack in a slightly offset π–π manner
[centroid–centroid distance between proximate rings 3.34 Å]
Fig. 6 The structure of 21 showing the atom labelling scheme; ellip-
soids are at the 60% level. Only the major component of the disordered
wherein the thymine rings face each other in a head-to-tail heterocyclic ring is shown for clarity. C(10) is hidden and lies behind C(6).
Selected geometric data: Si(1)–O(1) 1.6321(15), Si(2)–O(1) 1.6375(15),
N(1)–C(10) 1.408(6), N(1)–C(7) 1.421(5), N(3)–C(10) 1.358(8), N(3)–C(11)
1.371(7), N(3)–C(9) 1.372(6), C(7)–C(8) 1.423(5), C(8)–C(9) 1.337(5), N(2)–
C(7) 1.272(4) Å; Si(1)–O(1)–Si(2) 150.42(10), C(10)–N(1)–C(7) 124.8(4),
orientation. Similarly, the cytosine derivative 21 also aggre-
gates through π–π stacking [centroid–centroid distance
between proximate rings 3.33 Å], but these pairs further inter-
act via O(2)⋯H(8) hydrogen bonds to form parallel chains
C(10)–N(3)–C(9) 121.6(5), N(1)–C(7)–C(8) 115.0(3), C(9)–C(8)–C(7)
running approximately in the c-direction. Surprisingly, the 119.4(3), C(8)–C(9)–N(3) 123.8(4), N(3)–C(10)–N(1) 115.3(5)°. Hydrogen
bond data: O(2)⋯H(8’) 2.33 Å, ∠O(2)⋯H(8’)–C(8’) 154.8°. Symmetry
operation: 1 + x, y, z.
exocyclic NH2 group plays no part in the supramolecular
assembly.
The structures of other pyrimidocyclophanes incorporating
a single thymine have been reported,46 largely with macro-
cycle-,47 aza-47–49 and thio-methylene47 bridges between points
on the pyrimidine base. To the best of our knowledge, 21 is
the first such species incorporating cytosine.
Polymer chemistry
As a preliminary exploration of the potential to prepare nucleo-
base-substituted polysiloxane chains, we have prepared silox-
ane polymers carrying pendant thymine and adenine groups
using methodology adapted from that shown in Scheme 4.
Reaction of 3-(4-bromo)butylheptamethyltrisiloxane (13)
with a poly(dimethylsiloxane) with known molecular weight
yielded polysiloxanes with 0.2% of pendant bromobutyl func-
tionality (Scheme 6). The single peak in the 29Si spectrum of
23 at −22.0 ppm, which compares well with the 29Si NMR
chemical shift for octamethyltrisiloxane [−20.8 ppm],50 and
the lack of an observable 29Si resonance at ca. 7.5 ppm due to
trimethylsiloxy end groups of 13 indicate that the latter has re-
equilibrated with the PDMS to form a new polysiloxane con-
Fig. 5 The structure of 19 showing the atom labelling scheme; ellip-
soids are at the 40% level. Selected geometric data: Si(1)–O(1) 1.623(2),
taining pendant bromobutyl groups. The peaks in the 1H NMR
Si(2)–O(1) 1.629(2), O(2)–C(9) 1.219(3), O(3)–C(10) 1.254(5), N(1)–C(9) spectra compare well with those of the trisiloxane starting
1.385(3), N(1)–C(10) 1.391(4), N(1)–C(8) 1.475(4), N(2)–C(9) 1.376(4),
N(2)–C(13) 1.393(4), N(2)–C(14) 1.479(4), C(10)–C(11) 1.386(5), C(11)–
C(13) 1.386(4) Å; Si(1)–O(1)–Si(2) 150.53(15), C(9)–N(1)–C(10) 123.6(2),
C(9)–N(2)–C(13) 123.5(2), N(2)–C(9)–N(1) 115.2(2), C(11)–C(10)–
material (13).
Reaction with thymine and adenine (selected as the two
DNA bases that had shown the most consistent results in the
small molecule chemistry) under similar conditions to those
described above gave rise to the nucleobase-substituted poly-
N(1) 119.2(3), C(13)–C(11)–C(10) 118.9(3), C(11)–C(13)–N(2) 119.4(3)°.
Symmetry operation: 1 − x, 1 − y, 1 − z.
7744 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
in these materials. It may be that higher loadings of base pairs
are needed to achieve strong interactions in solution, particu-
larly with the A–T combination which has only the two hydrogen
bonding sites; using complementary bases with three or four
sites has been used previously20 to provide siloxanes with useful
material properties. To date, however, we have not achieved
complete conversion to pure base-functionalised polymers from
starting bromobutyl-substituted polymers with higher than
0.2% functionalised loadings to test this assertion.
Acknowledgements
We thank Dow Corning for financial support in the form of a
CASE studentship (to D. R. G. S). We also acknowledge the
EPSRC funded service at RAPRA Technology for the GPC
results.
Scheme 6
siloxanes 24 (thymine) and 25 (adenine) as viscous, air stable,
colourless oils in good yield.
References
The resonances due to the methylene protons α to the nitro-
gen in the H NMR spectra of 24 compared with those in 23
1 C. M. Niemeyer and M. Adler, Angew. Chem., Int. Ed., 2002,
41, 3779.
1
show a change to a more deshielded environment, as expected
from substitution with the secondary amine. All the reson-
ances compare very favourably with the analogous ones in 13
which indicates that the same reaction has taken place
between bromoalkyl-functionalised siloxanes and thymine as
in the small molecule chemistry. Substitution occurred exclu-
sively at the more basic 1-position37 assigned by comparison
with the peaks due to N-1 and N-3 addition in the 1H NMR
spectrum of small molecule analogues. The relative integrals
of the peaks due to the methylene protons and the siloxane
methyl protons suggest the molecular weight of the polymer
was not affected during the reaction. The small changes in
values measured by GPC are likely due to the effect of the
added base on the conformation and hence hydrodynamic
volume in solution, the values being reported relative to poly-
styrene calibration.
2 H. J. Spijker, F. L. van Delft and J. C. M. van Hest, Macro-
molecules, 2007, 40, 12.
3 R. McHale and R. K. O’Reilly, Macromolecules, 2012, 45,
7665.
4 I.-H. Lin, C. Cheng, Y. C. Yen and F. C. Chang, Macromole-
cules, 2010, 43, 1245.
5 A. S. Karikari, B. D. Mather and T. E. Long, Biomacro-
molecules, 2007, 38, 302.
6 H. J. Spijker, A. J. Dirks and J. C. M. van Hest, J. Polym. Sci.,
Part A: Polym. Chem., 2006, 44, 4242.
7 M. Fathalla, C. M. Lawrence, N. Zhang, J. L. Sessler and
J. Jayawickramarajah, Chem. Soc. Rev., 2009, 38, 1608.
8 J. A. Pavlisko and C. G. Overberger, in Biomedical and
Dental Applications of Polymers, ed. C. G. Gebelein and
F. F. Koblitz, Plenum, New York, 1981.
9 Y. Imanishi, J. Polym. Sci., Macromol. Rev., 1979, 14, 1.
The NMR spectra for the adenine-substituted polymer 25 10 Y. Imanishi, in Bioactive Polymeric Systems, ed.
also shows sufficient similarity to the small molecule com-
C. G. Gebelein and E. Carraher Jr, New York, 1985.
pounds for us to be confident of the structure of the polymers, 11 S. Sivakova, D. A. Bohnsack, M. E. Mackay, P. Suwanmala
with comparison of the position of the N9-CH2 peak indicating
and S. J. Rowan, J. Am. Chem. Soc., 2005, 127, 18202.
that substitution took place at the most basic 9-position.37 The 12 S. Sivakova, J. Wu, C. J. Campo, P. T. Mather and
change in Mn for the polymer 21 250 to 23 700 is not
significant.
In an initial attempt to investigate the potential interactions
between the polymer chains via hydrogen bonding, equal
S. J. Rowan, Chem – Eur. J., 2006, 12, 446.
13 B. Delord, M. C. Guillorit, J. Lafay, M. L. Andréola,
D. Tharaud, L. Tarrago-Litvak, H. J. A. Fleury and
G. Déléris, Eur. J. Med. Chem., 1996, 31, 111.
amounts of these 0.2% functionalised polymers (24, 25) were 14 J. Thibon, L. Latxague and G. Déléris, J. Org. Chem., 1997,
1
mixed and analysed by H NMR and GPC. Unfortunately, the
1H NMR spectra of the mixture gave no indication of the 15 L. Latxague, J. Thibon and G. Déléris, Tetrahedron Lett.,
potential bonding while the molecular weight distributions
1998, 39, 4025.
correspond to a simple mixture with no interaction between 16 J. J. E. Moreau, B. P. Pichon, G. Arrachart, M. Wong Chi
components. This suggests that little or no hydrogen bonding
Man and C. Bied, New J. Chem., 2005, 29, 653.
interaction takes place in solution between the thymine and 17 G. Arrachart, C. Carcel, J. J. E. Moreau, G. Hartmeyer,
62, 4635.
adenine functionalised siloxane polymers at the low loadings
B. Alonso, D. Massiot, G. Creff, J.-L. Bantignies, P. Dieudonne,
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 7734–7746 | 7745

View Article Online
Paper
Dalton Transactions
M. Wong Chi Man, G. Althoff, F. Babonneau and 37 B. Lippert, Coord. Chem. Rev., 2000, 200–202, 487.
C. Bonhomme, J. Mater. Chem., 2008, 18, 392.
18 J. G. Matisons and A. Provatas, ACS Symposium Series,
38 A. Houlton, D. M. P. Mingos and D. J. Williams, J. Chem.
Soc., Chem. Commun., 1994, 503.
(Silicones and Silicone-Modified Materials), 2000, 729, 128. 39 A. Pignedoli, G. Peyronel and L. Antolini, Acta Crystallogr.,
19 R. F. M. Lange, M. Van Gurp and E. W. Meijer, J. Polym.
Sci., Part A: Polym. Chem., 1999, 37, 3657.
20 R. P. Sijbesma and E. W. Meijer, Chem. Commun., 2003, 1.
Sect. B: Struct. Crystallogr. Cryst. Chem., 1973, 29, 1490.
40 J. Pranata, S. C. Wierschke and W. L. Jorgenson, J. Am.
Chem. Soc., 1991, 113, 2810.
21 N. Devylder, M. Hill, K. C. Molloy and G. J. Price, J. Chem. 41 D. Voet and A. Rich, Prog. Nucleic Acid Res. Mol. Biol., 1970,
Soc., Chem. Commun., 1996, 711. 10, 183.
22 G. J. Price, M. P. Hearn, E. Wallace and A. M. Patel, 42 A. D. Hamilton and D. van Engen, J. Am. Chem. Soc., 1987,
Polymer, 1996, 37, 2303.
109, 5035.
23 G. J. Price, Polym. Int., 2009, 58, 290.
43 Crystal data for 9a: C14H20N10NiO4S4, formula weight
ˉ
24 G. Kociok-Köhn, K. C. Molloy, G. J. Price and
D. R. G. Smith, Supramol. Chem., 2008, 20, 697.
25 K. K. Chatterjee, Z. Phys. Chem., 1974, 89, 88.
26 L. Sacarescu, I. Atudosie, M. Simionescu, G. Sacarescu and
V. Harabagiu, Chem. Heterocycl. Compd., 2011, 47, 602.
27 B. R. Baker and T. J. Schwa, J. Med. Chem., 1996, 9, 73.
579.35, triclinic, P1, a = 7.5968(5), b = 8.2841(5), c = 9.0845(6)
Å, α 104.418(3), β = 98.826(3), γ = 95.465(3)°, V 541.80(6) Å3,
ρcalc1.776 Mg m−3, μ(Mo-kα) 1.328 mm−1, F(000) 298, good-
ness-of-fit on F2 1.093, final R1, wR2 indices [I > 2σ(I)]
0.0450,0.1099, R1, wR2 (all data) 0.0594, 0.1186, max, min
peak and hole 0.495, −0.755 e Å3. CCDC no. 956094.
28 C. Schou, E. B. Pedersen and C. Nielsen, Acta Chem. Scand., 44 L. Galiullina, A. Nikolaev, V. Semenov, V. Reznik and
1993, 47, 889.
S. Latypov, Tetrahedron, 2006, 62, 7021.
29 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837.
30 C. I. Ahn, H. Y. Choi and B. S. Hahn, Heterocycles, 1990, 31,
1737.
31 D. G. Kim and L. V. Gavrilova, Chem. Heterocycl. Compd.,
1997, 33, 1382.
32 W. Saenger, Principles of Nuclei Acid Structure, Springer
Verlag, New York, 1984.
33 S. Aldridge, R. J. Calder, A. Rossin, A. A. Dickinson,
45 J. E. Mark, H. R. Allcock and R. West, Inorganic Polymers,
Prentice Hall, New Jersy, 2nd edn, 2004.
46 A. Kumar, G. Hundal, D. Paul, M. S. Hundal and H. Singh,
J. Chem. Soc., Perkin Trans. 1, 2000, 2295.
47 V. E. Semenov, A. E. Nikolaev, A. V. Kozlov,
S. N. Pod’yachev, O. A. Lodochnikova, O. N. Kataeva,
S. K. Latypov and V. S. Reznik, Russ. J. Org. Chem., 2008, 44,
899.
D. J. Willock, C. Jones, D. J. Evans, J. W. Steed, M. E. Light, 48 K. Doyama, F. Hama, Y. Sakata and S. Misumi, Tetrahedron
S. J. Coles and M. B. Hursthouse, J. Chem. Soc., Dalton
Trans., 2002, 2020.
34 G. Portalone, L. Bencivenni, M. Coleapietro, A. Pieretti and
F. Ramondo, Acta Chem. Scand., 1999, 53, 57.
35 C. J. Abshire and L. Berlinguet, Can. J. Chem., 1964, 42,
1599.
Lett., 1981, 22, 4101.
49 V. E. Semenov, A. E. Nikolaev, L. F. Galiullina,
O. A. Lodochnikova, I. A. Litvinov, A. P. Timosheva,
V. E. Kataev, D. R. Sharafutdinova, Y. Y. Efremov,
A. V. Chernova, S. K. Latypov and V. S. Reznik, Russ. Chem.
Bull., 2006, 539.
36 J. W. Wilt, W. K. Chwang, C. F. Dockus and N. M. Tomiuk, 50 P. Lux, F. Brunet, H. Desvaux and J. Virlet, Magn. Reson.
J. Am. Chem. Soc., 1978, 100, 5534. Chem., 1993, 31, 623.
7746 | Dalton Trans., 2014, 43, 7734–7746
This journal is © The Royal Society of Chemistry 2014