Synthesis, structure, photoluminescence and photoreactivity of 2,3-diphenyl-4-neopentyl-l-silacyclobut-2-enes

FULL PAPER

DOI: 10.1002/chem.200901323

Synthesis, Structure, Photoluminescence and Photoreactivity of

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

Duanchao Yan,^[a]Mark D. Thomson,^[b]Michael Backer,^[a]Michael Bolte,^[a]

Robert Hahn,^[b]Robert Berger,^[c]Wunshain Fann,^{[d, e]}Hartmut G. Roskos,^[b]and

Norbert Auner*^[a]

Dedicated to Professor Yitzhak Apeloig on the occasion of his 65th birthday

Abstract: A series of six 2,3-diphenyl-

4-neopentyl-1-silacyclobut-2-enes with

different 1,1-substituents has been pre-

pared and characterized by single-crys-

tal X-ray crystallography. These com-

pounds possess a cis-stilbene-like chro-

mophore involving also the four-mem-

bered ring, and exhibit a photophysical

behavior similar to that of previously

is degraded in a poor (aqueous) sol-

vent, the formation of nanoscale aggre-

gates leads to a significant enhance-

ment factor in the emission intensity,

due to the suppression of the photo-

reactivity in the more rigid molecular

environment, although the quantum

yield still remains below a few percent.

In the solid-state, however, photoreac-

tivity is completely suppressed leading

to fluorescence quantum efficiencies of

8–23% depending on the 1,1-substitu-

ents, which demonstrates these com-

poundsꢀ potential as chromophores for

condensed-phase luminescence applica-

tions. Two dominant competing photo-

chemical reactions have been identified

in solution (for excitation in the

lowest-energy

absorption

band,

>260 nm), which are analogous to re-

lated (sila-)cyclobutenes and stilbe-

noids. The first involves ring-opening

ꢀ

due to cleavage of the 1,4-Si C bond

to form metastable silabutadienes,

which was confirmed by isolating the

stereospecifically formed allylsilane

which results from a secondary reac-

tion with trapping agents such as meth-

anol or water. The second photochemi-

cal reaction involves ring closure of the

2,3-diphenyl substructure to form a di-

hydrophenanthrene analogue, which

was confirmed by isolating the phenan-

threne derivative that results following

subsequent hydrogen abstraction in the

presence of oxygen. Measurements of

the silacyclobutenes in doped-polymer

thin films reveal a spectroscopic behav-

ior ranging from that in solution to the

nano-aggregate case as the silacyclobu-

tene dopant concentration is increased.

reported

1,2-diphenyl-cycloalkenes.

This chromophore system is confirmed

by a theoretical investigation of the

electronic structure and excitation

spectra. The absorption and photolumi-

nescence of selected derivatives were

studied in solution, as solid powder

samples, and in doped-polymer thin

films. In well-dissolved solution, the si-

lacyclobutenes show only very weak

fluorescence emission (quantum yield

~0.1%), due to competition with pho-

tochemical and non-radiative photo-

physical relaxation. When the solubility

Keywords: aggregation · lumines-

cence · photochemistry · silanes ·

stilbene

[a] Dr. D. Yan, Dr. M. Backer, Dr. M. Bolte, Prof. Dr. N. Auner

Institut fꢁr Anorganische Chemie

[d] Prof. Dr. W. Fann

Institute of Atomic and Molecular Sciences

Academia Sinica, Taipei (Taiwan)

Johann-Wolfgang-Goethe-Universitꢂt, Max-von-Laue-Str. 7

60439 Frankfurt am Main (Germany)

Fax : (+49)69-798-29188

[e] Prof. Dr. W. Fann

We are deeply saddened that our colleague Wunshain Fann passed

away in 2008.

E-mail: auner@chemie.uni-frankfurt.de

[b] Dr. M. D. Thomson, R. Hahn, Prof. Dr. H. G. Roskos

Physikalisches Institut

Supporting information for this article is available on the WWW

under http://dx.doi.org/10.1002/chem.200901323.

Johann-Wolfgang-Goethe-Universitꢂt

60054 Frankfurt am Main (Germany)

[c] Dr. R. Berger

Frankfurt Institute for Advanced Studies (FIAS)

Johann Wolfgang Goethe-Universitꢂt

60438 Frankfurt am Main (Germany)

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

8625

Introduction

The investigation of new fluorescent organic and organome-

tallic chromophores continues to be an active research di-

rection, in particular to provide improved materials for ap-

plications such as organic electroluminescent (EL) devi-

ces.^[1–4]Important requirements for such chromophores are

i) that their emissive properties are not quenched in the

solid-state, ii) that their emission wavelength can be tuned

by variation of substituents, and iii) that they possess favora-

ble chemical functionality to allow flexible synthesis of tail-

ored derivatives and to be incorporated in macromolecular

systems (e.g., polymers). With respect to these criteria, ma-

terials based on organosilicon chromophores, such as

phenyl-substituted siloles,^[5–10]have demonstrated extremely

promising properties. In such compounds, the replacement

of C with Si groups can significantly modify the electronic

properties, and can be used to tune the excitation energy of

fluorescent p–p* excited states. In many cases, the s–p inter-

actions are enhanced leading to a reduced p–p* energy gap

in comparison to their pure C analogues, often red-shifting

the fluorescence spectra (further) into the visible range.^[5,11]

On the other hand, the Si substitution can lead to significant

changes in relative bond strengths and large resultant

changes in the molecular conformation and photochemical

behavior, especially in strained cyclic systems.

Here we present the synthesis, crystal structures, and pho-

tophysical/photochemical behavior of a set of organosilicon

compounds, 1,1-R₂-2,3-diphenyl-4-neopentyl-1-silacyclobut-

2-enes (1-R, see Figure 1 for structures).^[12–14]The relatively

small number of previous reports on 1-silacyclobut-2-enes

(SCBs)^[15–23]have been chiefly devoted to the 1,1-dimethyl

and 1,1-dimethyl-2-phenyl derivatives (DMSCB, Ph-

DMSCB) and the ring-opening photoreaction that occurs

upon excitation of the lowest-energy (p–p*) absorption

band. This band lies at relatively short wavelengths (i.e.,

below 230 nm),^[22]and the photophysical and photochemical

properties are similar to those of (alkyl-substituted) cyclobu-

tene derivatives (such as the 3,4-dimethyl derivative,

DMCB).^[24,25]In contrast, the SCB compounds presented

here incorporate a cis-stilbene moiety (i.e., the 2,3-diphenyl-

2-alkene substructure, as indicated in Figure 1a), and hence

are also closely related to stilbenoids, in particular 1,2-di-

phenylcycloalkenes (DPCs) such as 1,2-diphenylcyclobut-1-

ene (DPCB, Figure 1).^[26–32]The p–p* absorption band peak

for such stilbenoid chromophores typically lies near 300 nm,

with a corresponding fluorescence emission in the range of

350–500 nm, both of which can be further shifted towards

the visible region via a range of appropriate substitutions at

the doubly bonded carbon atoms.^[33,34]Hence the derivatives

1-R are potential candidates for photo- and electrolumines-

cence applications.

Figure 1. a) Title compounds 1,1-R₂-2,3-diphenyl-4-neopentyl-1-silacyclo-

but-2-enes 1-R, and related literature compounds, b) 1,1-dimethyl-2-R’-si-

lacyclobutene ACTHNUGRTNEUNG(Ph-)DMSCB, c) cis-stilbene and d) 1,2-diphenylcyclobu-

tene DPCB.

tation energies, which confirm the assignment of a stilbenoid

chromophore and indicate that the presence of the 1-Si only

weakly perturbs the HOMO/LUMO energies (from a com-

parison of the experimental and theoretical ionization po-

tentials). The small variation of the absorption/fluorescence

wavelengths in solution with varying 1,1-substituents R is

shown to be due to a variation in the Stokes shift (as op-

posed to the 0–0 transition energy), whilst the small fluores-

cence quantum efficiencies (F_f~0.1%) indicates the pres-

ence of efficient competing photochemical/photophysical

non-radiative relaxation. The fluorescence intensity is ob-

served to increase significantly when nanoscale clusters are

formed in degraded EtOH solutions (by the addition of

H₂O), accompanied by a reduction of the excitation energy

by ~1000 cm^ꢀ1, although the absolute F_fvalues still remain

around only a few percent. We then present measurements

of the absorption and fluorescence for solid-state (doped

KBr powder) samples, where F_factually increases to values

appropriate for solid-state luminescence applications (with

F_fin the range 8–23%, depending on R). Whilst prelimina-

ry reports on these compounds have indicated that an ap-

preciable solid-state fluorescence is observed,^[14,35]this is the

first detailed quantitative analysis including quantum effi-

ciency estimates. The variation of F_fand the fluorescence

band maxima (~30 nm) versus R are interpreted in light of

the different packing observed in the crystal structures.

In the second part of this paper we present a study of the

photoreactivity of 1-R in solution and doped-polymer thin

films. In terms of related photochemical compounds, the de-

rivatives 1-R possess similar structural characteristics to: i)

the previously reported compounds (Ph-)DMSCB which ex-

hibit an efficient electrocyclic ring-opening reaction due to

ꢀ

excited-state cleavage of the strained 1,4-Si C bond to yield

a short-lived silabutadiene (SBD) intermediate;^[15–23]and ii)

stilbenoids^[28,36](including DPCs^[29]) where electrocyclic ring

closure between the Ph groups forms the corresponding di-

hydrophenanthrene (DHP) analogue. We show that both

these photoreactions are observed for 1-R in solution, by

characterization of the isolated secondary products yielded

We have performed an analysis of the absorption and

fluorescence properties of a set of 1-R derivatives, in solu-

tion, as nano-aggregates (in degraded H₂O/EtOH solutions),

and in solid powder samples. The results are compared with

calculations of the (gas-phase) electronic structure and exci-

8626

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

from subsequent trapping reactions. For the case of the SBD

photoproduct (ring opening), the trapping agent contains a

hydroxyl group (e.g., MeOH, H₂O) which leads to the ste-

reospecific formation of the corresponding (stable) allylsi-

lane, whereas for the DHP photoproduct, irreversible for-

mation of the phenanthrene analogue occurs via hydrogen

abstraction by oxygen.^[27,30]Due to the significant differences

in electronic and steric factors between 1-R and the previ-

ously characterized SCBs/stilbenoids, the presence of both

competing photoreactions for 1-R was not a clearly foresee-

able result - for example, neither photoreaction is observed

for the stilbenoid DPCB.^[27]For each photoreaction, we in-

vestigate the net rate of trapped secondary photoproduct

formation from time-dependent measurements of the ab-

sorption or fluorescence spectra during photolysis. We con-

clude by discussing the effect of condensed-phase environ-

ment (solution, doped polymer thin films, nano-aggregates

and crystalline powder) on the observed degree of photo-

chemical DHP formation.

Scheme 1. Synthesis of 1-R from 1-Cl (corresponding yields indicated in

brackets).

ported previously^[39](C1: d=157.10 ppm; C2: d=

143.60 ppm), because the carbon connected to silicon atom

should have a lower d value (see Figure 2 for the atomic

numbering convention employed here). The detection of ²⁹Si

satellite signals of the peak at d=143.60 ppm owing to

²⁹Si–¹³C spin–spin coupling (¹J_SiC=69.7 Hz) confirms this as-

signment. While the 1,1-substituents have very small effect

on the -CMe₃groups, they do have some influence to C1,

C2, C3 and CH₂. The chemical shift values d²⁹Si are depen-

dent on the 1,1-substituents. While the 1,1-dialkyl com-

pounds 1-Me and 1-Et have positive d values, the com-

pounds 1-Vi and 1-Ph have negative d values originating

from the shielding effect of vinyl and phenyl groups. The

data for 1-H are strongly shifted to high field (d=

ꢀ33.2 ppm) with similar d value (ꢀ33.9 ppm) reported for a

closely related 1-hydrogen silacyclobutene derivative.^[37]

Results and Discussion

Synthesis: The derivatives, 1,1-R₂-2,3-diphenyl-4-neopentyl-

1-silacyclobut-2-enes, were prepared starting from 1,1-di-

chlorosilacyclobutene (1-Cl), which was synthesized accord-

ing to previously published procedures^[37,38]from trichlorovi-

nylsilane, tBuLi and tolane in pentane solution. Hydrogena-

tion of 1-Cl with LiAlH₄in Et₂O (at 08C) resulted in the

precipitation of LiAlCl₄and the formation of dihydrosilacy-

clobutene (1-H) in 68% yield. A pure white waxy solid of 1-

H was obtained by distillation at 1028C at 10^ꢀ2mbar, and

colorless single crystals were grown from cold pentane solu-

tion (ꢀ358C). The silacyclobutenes 1-Me, 1-Et, and 1-Vi

were synthesized by nucleophilic substitution reactions of 1-

Cl and the corresponding Grignard reagents in THF at

room temperature with yields >90%. Tetraphenylsilacyclo-

butene (1-Ph) was prepared from reaction of 1-Cl with

PhMgBr in boiling THF solution for 6 h, giving a white solid

product with 92% yield. The syntheses are summarized in

Scheme 1. The compounds generally have a high tempera-

ture stability (up to >2008C).

The silacyclobutenes were characterized by the usual

¹H NMR, ¹³C NMR and ²⁹Si NMR standard procedures (see

Experimental Section). The data of 1-Cl are obtained from

a previous report.^[39]The proton chemical shift values (d) of

CH₂consisted of two multiplets in a ratio of 1:1 due to the

diastereotopic methylene protons. Similarly, the ¹H NMR

spectrum also reveals that the two 1-substituents on the sili-

con atom are diastereotopic, for example, the two 1-methyl

substituents in 1-Me giving rise to two chemical shift values

at d = 0.45 and 0.34 ppm, respectively. As expected, the in-

fluence of 1-substituents to backbones, 2,3-diphenyl-4-neo-

pentyl-1-silacyclobut-2-enes, is decreased with increasing

number of intermediate bonds (Si > CH > CH₂> CH₃).

For 1-Cl we have revised the assignment of C1 (d=

143.60 ppm) and C2 (d = 157.10 ppm) compared to that re-

Single-crystal structures: All silacyclobutene compounds

1-R gave colorless single crystals from various organic sol-

vents such as Et₂O, pentane, and ethanol. The ORTEP dia-

grams are shown in Figure 2, whilst selected crystal data,

analysis parameters and bond lengths/angles are summar-

ized in Tables 1 and 2. All derivatives 1-R are chiral. Both

enantiomers are included in the unit cells and related to

each other through symmetry centers or mirror planes,

except for 1-Et that forms twins in the chiral space group

P2₁2₁2₁. 1-H is monoclinic with the Pn space group. Each

unit cell includes four complete molecular structures in two

independent sets. The structures of the two independent

molecules display similar bond lengths and angles (Table 2).

Only the angles between the 2,3-phenyl groups and olefinic

double bonds (C2-C1-C11 and C1-C2-C21) differ by about

2–38. Compared to the precursor 1-Cl, the substitution on

silicon in going from electronegative chlorine substituents to

ꢀ

hydrogen causes a small increase of the endocyclic silicon

carbon bond lengths and a small contraction of the inner C-

Si-C angle. Due to the almost planar structure of the four-

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8627

N. Auner et al.

although we include the struc-

ture and details here for com-

pleteness. 1-Et forms colorless

cubic crystals with the ortho-

rhombic space group P2₁2₁2₁,

also with four equivalent mole-

cules in each unit cell. 1-Vi

gives monoclinic crystals (space

group P2₁/n), again with four

equivalent molecules in each

cell. In contrast, 1-Ph when re-

crystallized from an ethanol/

chloroform (1:1) mixture forms

triclinic crystals (space group

¯

P1) with three non-equivalent

molecules in each unit cell. All

the diorgano compounds 1-R

possess similar structures with

only small differences in bond

lengths and angles. For the

compounds studied here, there

are no obvious trends concern-

ing the influence of the 1,1-di-

organo substituents on the sila-

cyclobutene ring structure. In

the four-membered ring, the

bonds of silicon to sp²-hybrid-

ꢀ

ized carbon (Si1 C1) are short-

er by about 5 pm compared to

3

ꢀ

the silicon sp -carbon bond

ꢀ

(Si1 C3), which indicates that

the latter bonds are also

weaker. This is consistent with

the observed photochemical

ring-opening

occurs during irradiation of the

lowest-energy (spin-allowed)

reaction

that

absorption band (see Section

on Photochemistry in solution

and doped-polymer thin films),

which involves cleavage of the

ꢀ

Si1 C3 bond. Other corre-

sponding endocyclic and exocy-

ꢀ

clic Si C and C C bond lengths

have comparable values.

For DPCs, the photophysical

and photochemical properties

are strongly influenced by the

phenyl-phenyl conformation of

the stilbene subunit.^[26,29]In

Table 3 we present some rele-

vant geometric parameters of

the stilbene subunit for selected

Figure 2. Molecular structures of 1-R. The numbering of atoms is adjusted in accordance with each other for

comparison. Previously published structure for 1-Me^[40]included for comparison. Bottom panel: Schematic and

labeling of selected stilbene subunit geometry used in Table 3.

membered ring, the other inner angles also change accord-

ingly. 1-Me recrystallized in the monoclinic space group P2₁/

c. There are four equivalent molecules in each unit cell. The

structure of this derivative has been published elsewhere,^[40]

derivatives 1-R, in comparison to those of DPCB and 1,2-di-

phenylcyclopentene (DPCP) (geometric labels as shown at

the bottom of Figure 2). In a previous report,^[26]the comput-

ed gas-phase structures for DPCB/DPCP were reasonably

8628

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

Table 1. Crystal data, experimental conditions, and summary of structure refinement for 1-R (including selected data previously published for 1-Me).^[40]

1-Cl

1-H

1-Me

1-Et

1-Vi

1-Ph

empirical formula

formula weight

l [ꢄ]

C₂₀H₂₂Cl₂Si

361.37

0.71073

C₂₀H₂₄Si

292.48

0.71073

173(2)

C₂₂H₂₈Si

320.54

–

C₂₄H₃₂Si

348.59

0.71073

173(2)

C₂₄H₂₈Si

344.55

0.71073

C₃₂H₃₂Si

444.67

0.71073

173(2)

T [K]

173(2)

cryst. syst.

space group

a [ꢄ]

b [ꢄ]

c [ꢄ]

monoclinic

P2₁/c

17.4873(11)

6.1101(5)

18.7176(13)

90

104.481(5)

90

1936.4(2)

4

1.240

0.394

760

0.43ꢅ0.32ꢅ0.22

2.25–26.46

ꢀ21ꢁhꢁ21

ꢀ7ꢁkꢁ7

ꢀ23ꢁlꢁ23

21063

monoclinic

Pn

6.8880(14)

18.237(4)

14.187(3)

90

92.30(3)

90

1780.7(6)

4

1.091

0.125

632

0.24ꢅ0.14ꢅ0.12

3.64–25.92

ꢀ8ꢁhꢁ8

ꢀ22ꢁkꢁ21

ꢀ17ꢁlꢁ16

11524

monoclinic orthorhombic

monoclinic

P2₁/n

9.3154(12)

15.0030(17)

14.923(2)

90

98.757(11)

90

2061.3(4)

4

1.110

0.117

744

0.31ꢅ0.11ꢅ0.09

3.50–25.87

ꢀ11ꢁhꢁ10

ꢀ18ꢁkꢁ18

ꢀ18ꢁlꢁ18

12240

triclinic

P1

¯

P2₁/c

P2₁2₁2₁

17.577(2)

6.116(1)

18.830(1)

90

9.4981(6)

9.7166(10)

23.1675(16)

90

10.7601(11)

15.5893(17)

23.340(2)

90.422(9)

94.506(9)

96.737(9)

3875.5(7)

6

1.143

0.108

1428

0.27ꢅ0.22ꢅ0.18

2.17–25.78

ꢀ13ꢁhꢁ13

ꢀ18ꢁkꢁ18

ꢀ28ꢁlꢁ28

26816

a [8]

b [8]

104.95

90

g [8]

90

V [ꢄ³]

1955.72

4

2138.1(3)

4

Z

1_calcd[Mgm^ꢀ3

]

1.087

1.083

m [mm^ꢀ1

]

–

0.113

760

F₀₀₀

cryst size [mm³]

q range [8]

0.52ꢅ0.33ꢅ0.32

2.27–27.90

ꢀ12ꢁhꢁ12

ꢀ12ꢁkꢁ11

ꢀ30ꢁlꢁ30

21240

index ranges

rfln.s collected

indep. rflns, [R

–

(int)]

3969, [0.0355]

6132, [0.0548]

6132/2/395

5094, [0.0498]

5094/0/227

3954, [0.0750]

3954/0/226

13783, [0.0957]

13783/0/892

data/restraints/parame- 3969/0/208

ters

goodness-of-fit on F²

final R indices

[I>2s(I)]

1.016

R1=0.0315,

wR2=0.0857

0.831

–

0.981

0.920

1.070

R1=0.0435,

wR2=0.0784

0.140/ꢀ0.140

R1=0.0336,

wR2=0.0866

0.244/ꢀ0.137

R1=0.0554,

wR2=0.1076

0.256/ꢀ0.185

R1=0.0835,

wR2=0.2476

0.635/ꢀ0.404

largest diff. peak/hole, 0.373/ꢀ0.205

–

eꢄ^ꢀ3

Table 2. Selected bond lengths (in ꢄ), angles, torsion angles, and dihedral angles (in degrees) of 1-R derivatives (including selected data previously pub-

lished for 1-Me^[40]).

1-Cl

1-H (a,b)^[a]

1-Me

1-Et

1-Vi

1-Ph (a,b,c)^[a]

ꢀ

Si1 C1

1.8291

1.8832

2.0449 Si1 Cl1

2.0357 Si1 Cl2

1.354

1.5484

1.473

1.480

1.856, 1.857

1.896, 1.893

1.858

1.909

1.862

1.859

1.356

1.539

1.468

1.481

77.05

108.3

1.8711

1.9058

1.8685

1.8774

1.3573

1.5317

1.475

1.484

76.56

109.54

1.851

1.903

1.857

1.865

1.354

1.541

1.490

1.492

77.15

110.40

1.887, 1.881, 1.872

1.925, 1.940, 1.917

1.855, 1.881, 1.892

1.886, 1.899, 1.888

1.315, 1.396, 1.359

1.577, 1.537, 1.547

1.499, 1.456, 1.481

1.496, 1.543, 1.501

75.67, 77.79, 76.80

109.54, 110.35, 115.13

ꢀ

Si1 C3

ꢀ

Si1 C21

ꢀ

Si1 C22

C1=C2

1.358, 1.366

1.540, 1.538

1.475, 1.467

1.494, 1.494

77.29, 77.57

ꢀ

C2 C3

ꢀ

C1 C9

ꢀ

C2 C15

C1-Si1-C3

79.45

C21-Si1-C22

106.05

Cl1-Si1-Cl2

89.13

109.50

81.72

130.25

127.79

3.94

Si1-C1-C2

C1-C2-C3

C2-C3-Si1

C9-C1-C2

C1-C2-C15

Si1-C1-C2-C3

90.8, 90.31

90.85

108.12

83.69

129.12

128.49

ꢀ4.76

38.24

56.03

62.63

90.84

108.02

84.49

129.69

128.63

2.54

63.10

30.31

66.44

91.22

107.7

83.82

133.0

130.4

2.6

23.19

43.07

51.20

92.29, 88.46, 90.34

107.58, 108.67, 108.99

83.30, 83.26, 82.57

132.16, 131.34, 133.80

132.14, 130.34, 130.03

ꢀ9.49, 11.89, 10.19

21.3, 15.9, 23.2

107.8, 107.9

84.03, 84.02

132.86, 131.89

132.45, 131.25

ꢀ2.2, ꢀ3.99

14.33, 20.79

47.62, 47.52

51.32, 56.47

angle of plane Si1, C1, C2, C3 with phenyl ring C9–14

angle of plane Si1, C1, C2, C3 with phenyl ring C15–20

angle between phenyl groups

38.27

55.98

63.07

44.2, 44.3, 34.6

54.2, 54.6, 49.0

[a] Compounds 1-H and 1-Ph have two and three non-equivalent molecules in each unit cell, respectively.

close to the experimental crystal structures, indicating that

the crystal data is also a useful indicator for gas-/solution-

phase structure, despite the presence of some packing ef-

fects.

The ethylene bond lengths, r₁, for 1-R except for 1-H vary

only slightly (in the range r₁=1.354–1.357 ꢄ) and are highly

consistent with those for DPCB/DPCP (r₁=1.35 ꢄ/1.34 ꢄ).

The two non-equivalent ethylene-phenyl bond lengths r₂for

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8629

N. Auner et al.

Table 3. Comparison of stilbene subunit geometry (single crystal) for se-

lected 1-R, as well as literature data for 1,2-diphenylcyclobutene (DPCB)

and 1,2-diphenylcyclopentene (DPCP). Bond lengths, angles and phenyl-

phenyl distance labels as indicated in Figure 2. Twist angle q defined as

angle between mean planes of silacyclobutene ring and each 2,3-Ph

group. Multiple rows for 1-H and 1-Ph for each non-equivalent molecule.

Values for theoretical gas-phase structures (where available) in square

brackets.

r₁[ꢄ] r₂[ꢄ]

F [8]

W [8] q [8]

d_2ꢀ2’

[ꢄ]

1-H

1.365

1.358

1.468/1.495 131.9/131.2 7.43

1.474/1.495 132.8/132.5 4.69

20.79/

47.52

14.3/

3.34

3.31

47.62

A

[132.2/

129.7]

1.475/1.484 130/129

G

ACHTUNGTRNE[NUNG 3.39]

Figure 3. Correlation plot of selected geometric parameters (ethylene-Ph

bond angle F, and phenyl twist angle q, single crystal) of the stilbene sub-

structure in compounds 1-R, in comparison with DPCB and DPCP.^[26]

Average values taken for the asymmetric phenyl groups (and over the

three nonequivalent molecules for 1-Ph). Data from both experimental

single-crystal (circles) and theoretical gas-phase (squares, labels in square

brackets) structures.

1-Et

1.357

1.356

3.5

ACHTUNGTRENNUNG

1-Me^[a]

(S₀)

1.468/1.481 129.1/128.5 7.5

A

[131.4/

129.5]

[128.9/

125.9]

1.480]

1-Me (S₁)

A

ACHTUNGTRENNUNG

1.419]

1-Vi

1-Ph

1.354

1.357

ACHTUNGTRENNUNG[1.361] [1.468/

1.490/1.492 133/130.4

5.03

10.17/ 21.34/

44.2

13.39/ 15.87/

44.3

10.28 23.23/

34.52

23.2/43.1 3.238

1.48/1.51

132/130.8

[131.6/

3.24

3.30

spanned by DPCB and DPCP. The fact that the Ph ring-

closing reaction (to form DHP) occurs for DPCP (F_rc

~50%), and not for DPCB (F_rc~0%), has been attributed

in part to this difference in the ground-state (Franck-

Condon) structure.^[29]Hence, in the case of 1-R, one might

expect that the DHP photoreaction in solution (see Section

on Stilbene-subunit ring-closing photoreaction) can occur

with reasonable efficiency, at least in some of the derivatives

(R=Me, Et).

In order to assess which features of the crystal structures

are due to intramolecular versus packing effects, we calcu-

lated the gas-phase ground-state structures (using B3LYP/

def-TZVP, see Quantum chemical calculations, Experimen-

tal Section) for R=H, Me, and Ph, as well as that for

DPCB. Selected structure parameters are listed in Table 3

(in square brackets), whilst the corresponding q, f values

are also included in Figure 3. The key features of the crystal

structure are reproduced, including the asymmetric confor-

mation of the 2- and 3-Ph groups (notwithstanding certain

complications for 1-H and 1-Ph with non-equivalent mole-

cules in the unit cell). The theoretical q, F data in Figure 3

demonstrates that the trend described above is expected to

be retained in the gas-phase (and quite possibly, in solution),

and that the stilbene subunit conformation remains inter-

mediate between that of DPCB and DPCP. One observes,

however, that the general position of each 1-R derivative

(R=H, Me, Ph) on this plot is altered, suggesting that the

gas-phase conformation is altered to a lesser extent than in

the solid-state. In conclusion, a key result of this analysis is

that due to the distortion of the four-membered ring (via

1.480]

129.5]

A

ACHTUNGTRNE[NUNG 3.385]

42.5]

16/26

A

DPCB^[b]

DPCP^[b]

1.35

[1.360]

1.34

1.47

136.5

10.3

[8.5]

7.5

3.4

[3.429]

3.4

U

A

U

G

ACHTUNGTRENNUNG

[a] For 1-Me, computed values are given for both the energy-optimized

electronic ground- (S₀) and lowest-energy excited (S₁) singlet state struc-

tures. [b] Experimental data taken from refs. [100] (DPCB) and [26]

(DPCP), respectively.

1-R vary over a somewhat larger range (1.468–1.510 ꢄ), en-

compassing the values for DPCB/DPCP (r₁=1.47 ꢄ/

1.48 ꢄ). A key difference between the reported structure of

DPCB and DPCP^[26]involves the co-planarity of the Ph

groups. The out-of-plane angles of the phenyl groups for

DPCB (q=16/268, the asymmetry due to packing effects)

are significantly smaller than for DPCP (q=44/488). How-

ever, the resulting distance between the 2-C atoms of the

phenyl groups remain essentially the same (d_2–2’=3.4 ꢄ),

ꢀ

due to a larger C=C C bond angle for DPCB (f=136.58)

compared with DPCP (f=128.68) (and, to a lesser degree,

the larger ethylene torsion W for DPCB). In comparison,

the values for q, f for 1-R vary over a relatively wide range

between those for DPCB and DPCP, with resulting phenyl–

phenyl separations of d_2–2’= 3.2–3.5 ꢄ. To illustrate the rel-

ative phenyl arrangements more clearly, we show in Figure 3

a correlation plot of q and f for 1-R and DPCB/DPCP. As

could be expected, a clear correlation exists in that as the

bond angle f is reduced, the phenyl rings twist further out

of plane (i.e., q) to alleviate the repulsion between the 2,2’-

position hydrogen atoms. The structural parameters for 1-

Me and 1-Et are particularly close to that of DPCP, whereas

for 1-Ph and 1-Vi they fall near the center of the range

ꢀ

elongation of the 1,4-Si C bond relative to the correspond-

ꢀ

ing 3,4-C C bond in DPCB) the stilbenoid structure is

driven toward that of the five-membered-ring carbon-ana-

logue DPCP.

Photophysics in solution: The UV/Vis absorption spectra of

1-R are shown in Figure 4a (for R=Me, Et, Vi and Ph, in

8630

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

yielded systematic deviations to the asymmetric experimen-

tal band-shapes, we employed a more detailed model corre-

sponding to an unresolved Franck-Condon progression with

a single dominant displaced harmonic vibration (with n˜ =

1600 cm^ꢀ1; details given in the Experimental Section). The

resultant fitted band, and residual absorption for 1-Ph are

included in Figure 4a, which indicates the close agreement

between the low-energy absorption tail and the fitted model

band. Inspection of the residual absorption curve indicates

the presence of an additional absorption band between the

two resolvable maxima (i.e., the shoulder in the residual ab-

sorption in the range ~250–280 nm). In the case of cis-stil-

bene and DPCs, there is a well-defined minimum between

bands A and B,^[26]indicating that the replacement of the 1-C

with Si introduces an additional near-UV transition in these

compounds, which may well possess some s_Si–p* character.

The experimental and fitted absorption bands for the 1-R

derivatives (R=Me, Et, Vi, Ph) are shown in the left por-

tion of Figure 4b. The fitted absorption peak and oscillator

strengths^[42]are listed in Table 4 and vary with R from l=

283 nm/f=0.22 (for R=Vi) to l=288 nm/f=0.34 (for R=

Ph).

Excitation of this lowest-energy p–p* band in fresh solu-

tions of 1-R results in a weak fluorescence emission band,

observable in relatively high concentration solutions (10^ꢀ4m)

before the onset of the dominant emission from photoprod-

ucts (see Section on Photochemistry in solution and doped-

polymer thin films). In the right portion of Figure 4b we

show the corresponding fluorescence spectra (l_ex=286 nm

for R=Me, Et, Vi, l_ex=300 nm for R=Ph). Qualitatively,

these spectra exhibit a clear reflection symmetry with re-

spect to their absorption bands about a relatively constant

transition origin, indicating that the emissive state should be

fairly well described by a displaced excited-state potential of

similar shape to that of the ground state.^[42]The fluorescence

band peak wavelengths extracted from these spectra (ob-

tained by fitting quadratic functions about the peak region),

as well as estimates of the transition origins and Stokesꢀ

shift (using both absorption/fluorescence spectra) are includ-

ed in Table 4. From the data, one sees that while the sub-

stituents R do have a moderate effect on the Stokes shift

(with an increase of ~1130 cm^ꢀ1in going from R=Ph to

R=Vi), the transition origin is essentially constant (the var-

Figure 4. a) UV/Vis absorption spectra of selected 1-R derivatives in

EtOH (R as indicated, 2.5ꢅ10^ꢀ5m). Partially resolvable band regions in-

dicated labeled ’A’ and ’B’. Fitted model absorption band and residual

for 1-Ph included as dashed lines. (b) Low-energy absorption band

region and corresponding fluorescence spectra for 1-R (1.0ꢅ10^ꢀ4m).

Fitted absorption bands shown as dashed curves (single displaced har-

monic mode model as described in text). Inset c) shows the fluorescence

quantum efficiencies vs Stokes shift for each derivative.

fresh EtOH solutions). Two partially resolved bands are evi-

dent (whose regions are labeled by ’A’ and ’B’), with peaks

lying near ~280 and ~230 nm, respectively. Both of these

features are highly consistent in position and intensity with

the p–p* bands observed for cis-stilbene and related DPCs

(such as DPCB),^[26,41]for example, for cis-stilbene in EtOH

the reported peak positions for the two lowest-lying bands

are at 277 nm (band A) and 224 nm (band B), respective-

ly.^[41]Moreover, we note that for the previously reported

SCB derivatives, DMSCB^[22]and Ph-DMSCB (the latter

bearing a single Ph-group at the 2-position),^[23]no such ab-

sorption bands with comparable

intensity were observed. Hence

Table 4. Absorption and fluorescence parameters for selected 1-R in (molecularly-dissolved) EtOH solution.

Computed gas-phase results (using B3LYP/def-TZVP, see text) for the lowest-energy (singlet) transition given

in square brackets.

we assign the key absorption

features for 1-R to analogous

p–p* transitions predominantly

localized on the stilbene subu-

nit.

In order to better resolve the

absorption features and gain

parameters for the lowest-

energy band (in region ’A’), we

fit the resolvable low-energy

region (i.e., l>290 nm). As the

Abs. max. l_abs[nm]^[a]Fl. max.

l_em[nm]

Transition

Stokes shift [cm^ꢀ1

]

Osc.

Fl. quantum

eff. [%]

origin [nm]^[b]

strength^[a]

f

1-Me 285.9

[301.4, 293.5^[c]

1-Et 285.0

399.6

334.7

9950

0.34

0.13

]

[456.6, 446.2^[c]

402.4

]

[367.9, 361.9^[c]

334.8

]

[11280, 11654^[c]

10240

]

[0.22,0.31^[c]

0.29

]

0.096

0.084

0.14

1-Vi 283.0

1-Ph 287.9

400.0

391.7

333.6

333.3

10340

9200

0.22

0.33

[303.9, 295.8^[c]

]

[0.28,0.30^[c]

[a] Peak positions and oscillator strengths obtained from fitted model absorption bands, as described in text.

[b] Determined from reflection analysis of absorption and fluorescence bands. [c] Using PBE0/def-TZVP

use of a single Gaussian band single-point and vertical transition energies.

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8631

N. Auner et al.

iation over the range 333.3–334.8 nm amounting to only

~130 cm^ꢀ1which is close to the margin of uncertainty of the

present analysis). This shows

that the 1,1-substituents R do

not lead to a discernible change

of the relative p/p* electronic

energies of the stilbene-like 2,3-

phenyl-2-ene chromophore, but

do influence the conformational

properties in solution.

structure and little purely electronic influence from the 1,1-

substituents is expected.

In order to quantify the weak

fluorescence, we performed an

analysis of the fluorescence

quantum efficiencies (F_f)^[43]for

each derivative in Figure 4,

using the compound 2,5-diphe-

nyloxazole (PPO) in EtOH as a

standard (whose absorption and

emission properties are very

Figure 5. Calculated (Kohn–Sham) frontier molecular orbitals for cis-stilbene, DPCB, 1-H and 1-Ph (generated

in Molden^[99]).

similar to those of 1-R, see Solution fluorescence quantum

yield determination, Experimental Section for details). The

resultant F_fvalues are listed in Table 4, varying from 0.084

(R=Vi) to 0.14% (R=Ph). As discussed in the Section on

the Photochemistry in solution and doped-polymer thin

films, these low values of F_fare a result of two competing

photochemical reactions (involving fairly large-scale confor-

mational changes), as well as non-radiative photophysical

relaxation. A trend emerges if we compare F_fversus the

Stokes shift for each derivative (Figure 4c), where one ob-

serves that F_fdecreases significantly with increasing Stokes

shift. Such a trend is reasonable, considering the competi-

tion with photochemical relaxation: As the Stokes shift (dis-

placement between Franck-Condon and emissive conforma-

tions) increases, a correspondingly larger fraction of the ex-

cited molecules migrate into the non-radiative decay chan-

nels, hence reducing F_f.

The calculated excitation spectra are shown in Figure 6

(using the functional/basis-set combination PBE0/def-TZVP,

see Quantum chemical calculations, Experimental Section).

The lowest energy (vertical singlet) p–p* transition is well

separated from the higher energy transitions and occurs at

294.7 nm (cis-stilbene), 312.7 nm (DPCB), 295.2 nm (1-H),

293.5 nm (1-Me), and 295.8 nm (1-Ph), respectively. For

comparison, we include the experimental absorption band

peak wavelengths in EtOH solution: l_abs= 277 nm (cis-stil-

bene^[41]), 297 nm (DPCB^[44]), 285.9 nm (1-Me), and 287.9 nm

(1-Ph), which are consistently at higher energy than the cal-

culated (gas-phase) values by 2170, 1690, 905, and 928 cm^ꢀ1,

respectively (note that the energy order is nevertheless pre-

served). The calculated transition energies using B3LYP/def-

TZVP were even further red-shifted, by ~1000 cm^ꢀ1(see

Supporting Information), which is consistent with previously

observed trends for the performance of the various function-

als.^[46,47]

The photophysical data indicate that the p–p* excitation

energy in 1-R (l_abs=283–288 nm) is very similar to that of

cis-stilbene (277 nm^[41]) and DPCB (297 nm^[44]) (all values

for room-temperature EtOH solution). In order to gain fur-

ther understanding into the nature of the chromophore and

electronic structure in 1-R, we carried out a set of theoreti-

cal calculations of the gas-phase excitation energies (using

TD-DFT, see Quantum chemical calculations, Experimental

Section). The calculated frontier (Kohn–Sham) orbitals

(HOMO, LUMO) are shown in Figure 5 for cis-stilbene,

DPCB, 1-H and 1-Ph. From visual comparison, the calculat-

ed HOMO/LUMO for cis-stilbene closely resemble those

published previously,^[45]while for the results here the orbi-

tals are much the same across all four compounds, strongly

supporting the notion that the chromophore in 1-R is very

similar to that of stilbenoids. An inspection of the frontier

orbitals for 1-Ph shows no significant delocalization onto

the 1,1-Ph₂substituents, nor is there any clear indication for

distortion of the orbitals between 1-H and 1-Ph. Hence the

electron density changes upon excitation and the transition

densities are expected to be localized on the stilbene sub-

In order to assess the possible role of solvatochromism,^[48]

we computed the lowest-energy singlet transitions for cis-

stilbene, 1-H, 1-Me and 1-Ph including solvent effects in

EtOH using an effective dielectric continuum method (see

Quantum chemical calculations, Experimental Section). This

calculation predicted a moderate bathochromic shift for all

compounds, that is, ꢀ532 (cis-stilbene), ꢀ367 (1-H), ꢀ390

(1-Me), and ꢀ328 cm^ꢀ1(1-Ph). This is a reasonable result

considering that hypsochromic shifts typically result only

when one has both appreciable and anti-parallel dipole mo-

ments in the ground- and excited-states in a polar solvent.^[48]

Hence electrostatic solvent effects cannot be invoked to ex-

plain this deviation between experiment and theory, which

is presumably due to (modest) systematic deficiencies in the

theoretical description (although shifts due to specific sol-

vent-chromophore interactions cannot be fully ruled out).

Nevertheless, the experimental and predicted oscillator

strengths f in Figure 6a,b,d,e show good agreement (note

that the oscillator strengths used here for cis-stilbene and

DPCB are for 3-methylpentane solutions^[26]), for example,

8632

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

values^[26]for cis-stilbene (U_ip=8.20 eV) DPCB (U_ip=

7.71 eV) and 3,3,4,4-Me₄-DPCB (U_ip=7.67 eV). Hence the

IPs for 1-Me and DPCB differ by only 0.05 eV, and there is

no clear evidence for a silicon-induced shift of the p-orbital

energy. This is in contrast to the case of siloles^[10]where s_Si–

p/s_Si*–p* conjugation effects are significant. The IPs pre-

dicted from the HF calculations (i.e., applying Koopmansꢀ

theorem) with the def-TZVP basis set were 7.89 (cis-stil-

bene), 7.54 (DPCB), 7.88 (1*-H), 7.89 (1-H), and 7.79 eV

(1-Me), respectively (note that 1*-R denotes a theoretical

derivative of 1-R in the absence of the 4-neopentyl group).

Evidently, the theoretical values for cis-stilbene and DPCB

are systematically lower than the experimental values (by

0.31 and 0.17 eV, respectively), although this error is within

a commonly reasonable range (~0.3 eV) for the application

of Koopmansꢀ theorem to molecules of this size,^[49]whereas

the value for 1-Me (somewhat fortuitously) differs by only

0.03 eV. Note that additional calculations for cis-stilbene

and DPCB using an outer-valence Greenꢀs function treat-

ment with an additionally polarized TZVP basis set (def2-

TZVP) predicted essentially the same ionization potential

as the HF calculations, that is, within ~0.01 eV.

The results for DPCB, 1*-H, and 1-Me are summarized in

the orbital energy diagram in Figure 7. A comparison of the

calculated orbital energies for DPCB and its direct analogue

1*-H would suggest that the HOMO is stabilized by some

0.34 eV upon Si substitution (comparable to that previously

calculated for cyclopentadiene (CPD) versus silole:

0.54 eV^[5]), whereas the LUMO is slightly destabilized

(whereas for CPD vs silole the calculated LUMO is strongly

stabilized by ~1.3 eV). However, the experimental and

theory results for DPCB and 1-Me also demonstrate that

the difference between calculated and theoretical values is

not systematic, and in the absence of an experimental IP of

1*-H it is currently not possible to confirm whether the pre-

dicted HOMO shift of around 0.34 eV indeed occurs. As-

suming the correction terms would be more consistent for

the silicon derivatives 1-H and 1-Me, the predicted destabili-

zation of the HOMO of 1-Me by 0.10 eV (~810 cm^ꢀ1) would

be attributable to the inductive effects of the more electron-

rich Me₂Si group on the stilbenoid chromophore. This is

consistent with the corresponding experimental gas-phase IP

values, which shows a shift of 0.16 eV (see Supporting Infor-

mation) relative to 1-Me. In concluding this section, we note

that future studies of 1-R in frozen solvent glasses—al-

though not directly applicable to practical luminescence ap-

plications–-would be interesting in order to assess how the

medium rigidity (and reduced thermal activity of low-fre-

quency vibrations) affect the fluorescence yield. Indeed, in

the case of cis-stilbene in solution F_fgoes from <1% at

room temperature to 75–79% upon cooling to 77 K due to

the suppression of photochemical/photophysical relaxa-

tion^[50,51](with the remaining fraction of non-radiative decay

presumably due to residual internal conversion).

Figure 6. Calculated excitation energy spectra (gas-phase, TD-DFT,

PBE0/def-TZVP) for a) cis-stilbene, b) DPCB, c) 1-H and d) 1-Me, and

e) 1-Ph (solid lines). Also shown in a), b), d), e) are the lowest-energy ex-

perimental band maxima in EtOH solution (dashed line, values for cis-

stilbene, DPCB taken from refs. [26,41,44]).

for 1-Ph f = 0.33 (exptl) and 0.27 (theory). Moreover, the

sequence of higher energy transitions calculated for 1-Ph

correspond well to the higher energy band (band B, l_abs

~230 nm, Figure 4), whilst the appearance of additional in-

termediate transitions compared to cis-stilbene/DPCB may

account for the absence of a well-defined minimum between

bands A and B, as identified earlier.

We also performed additional (gas-phase) TD-DFT calcu-

lations for 1-Me in both the ground-state and excited-state

equilibrium structures (B3LYP/def-TVZP), in order to

obtain the predicted Stokes shift between absorption and

fluorescence spectra. The values obtained were l_abs= 301.4

and l_em= 456.6 nm, which corresponds to a Stokes shift of

11280 cm^ꢀ1. This is in relatively good agreement with the

experimental value for 1-Me in EtOH of 9950 cm^ꢀ1

(Table 4).

Although the p–p* transition energy of 1-R versus cis-stil-

bene/DPCB in solution does not suggest any significant

electronic effect due to the presence of Si, the question re-

mained as to whether a concerted shift of both the p and p*

energies occurs. In order to address this issue, we performed

gas-phase photoelectron spectroscopy (PES) on the com-

pound 1-Me, as well as Hartree–Fock (HF) calculations to

predict the ionization potentials (IP, U_ip).

The experimental IP for 1-Me (estimated from the first

maxima in the PES spectra, see Supporting information)

was U_ip=7.76 eV, which can be compared to literature

Effect of aggregation in solution: Given that the fluores-

cence quantum yields of 1-R are very low in molecularly dis-

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8633

N. Auner et al.

formed an analysis of the band parameters and determina-

tion of the F_fvalues (see above), the results of which are

listed in Table 5. A visual inspection of the bands in Fig-

ure 8c shows that the clear reflection symmetry between ab-

sorption and fluorescence bands observed in pure EtOH so-

lution (Figure 4c) is not preserved in the nano-aggregates,

indicating that the excited state conformation might be dis-

torted somewhat by intermolecular interactions in the latter

case.

The key feature in the data is that the fluorescence inten-

sity increases significantly as a result of aggregation (as indi-

cated by the fluorescence spectra in the inset Figure 8b for

1-Ph), with the F_ffactor relative to pure EtOH solution

varying from 2.8 (1-Me) to 16.4 (1-Ph). For all the deriva-

tives, the aggregation is apparently accompanied by a reduc-

tion in the p–p* energy gap, with red-shifts in the transition

origin ranging from 825 cm^ꢀ1(1-Me) to 1201 cm^ꢀ1(1-Ph),

and corresponding red-shifts of the absorption and fluores-

cence bands (Table 5).

Figure 7. Energy diagram for the calculated (Hartree–Fock, solid lines)

frontier orbitals of DPCB, 1*-H (1-H in the absence of the 4-neopentyl

group) and 1-Me. Also included for DPCB and 1-Me are the HOMO en-

ergies implied by the experimental ionization potentials (dashed lines).

solved solution due to competition with internal conversion

and photochemical relaxation, we wished to explore the

effect of moving towards the solid-state by inducing aggre-

gation in solutions with degraded solvation properties.

Indeed, the formation of nanoscale molecular aggregates in

solution has been reported to greatly enhance the fluores-

cence yield of certain compounds. Such is the case for

phenyl-substituted siloles (e.g., for 1-Me-1,2,3,4,5-pentaphe-

nylsilole, F_fgoes from 0.06% in EtOH solution to 21%

upon aggregation^[9]), although there the effect of aggrega-

tion was to reduce the photophysical, as opposed to photo-

chemical, non-radiative efficiency. As per ref. [8] we chose a

10:1 mixture of H₂O/EtOH as the solvent, which allowed

the preparation of samples without significant precipitation

of the solute.

The rate of photoproduct formation for the nano-aggre-

gate samples was significantly lower than for the pure EtOH

solutions (by at least an order of magnitude), indicating that

the more rigid aggregate environment hinders the associated

large-amplitude motion (see Section on Photochemistry in

In Figure 8a we show the absorption spectrum of 1-Ph in

such a H₂O/EtOH solution (2.5ꢅ10^ꢀ5m). As can be seen, a

long diffuse low-energy tail emerges in the spectrum of the

H₂O/EtOH solution, which we attribute to scattering effects

and indicates the formation of aggregates in this sample.

Such a smooth and non-fluctuating scattering signature indi-

cates that the aggregates possess a size on the nanometer

scale. In order to model the nano-aggregate absorption

bands and extract estimates of the band parameters (as de-

scribed in the last section), we first analyzed this scattering

tail (in the range l=375–500 nm), which could be fit very

well using a function of the form l^ꢀp, where the fitted values

of p varied between 4.0 (1-Me) to 6.2 (1-Vi). We then fitted

the net absorption spectra (i.e., after subtraction of the scat-

tering contribution) with the same absorption band model

as for the pure EtOH solution spectra in the last section

(using the data in the range 295–400 nm). The results of this

fitting procedure are illustrated for 1-Ph in Figure 8a, whilst

the net absorption spectra and fitted bands for each 1-R de-

rivative are shown in Figure 8c. The model parameters are

listed in Table 5, including the changes relative to pure

EtOH solution. We also measured the fluorescence spectra

of H₂O/EtOH solutions (1.0ꢅ10^ꢀ4m, Figure 8c) and per-

Figure 8. Absorption/fluorescence spectra of 1-R as nano-aggregates in

H₂O/EtOH (10:1) solution. a) Absorption and b) fluorescence for 1-Ph.

In a) components of spectral fitting and residual as listed in legend

(model described in text). In b) the fluorescence spectrum of molecular-

ly-dissolved EtOH solution is included to indicate the change upon ag-

gregation. c) Absorption and fluorescence spectra vs R. The fitted scat-

tering contributions have been subtracted from the absorption spectra in

c). Fluorescence spectrum for 1-Ph scaled down by a factor of 5 for

visual comparison. (Absorption: 2.5ꢅ10^ꢀ5m, fluorescence: 1.0ꢅ10^ꢀ4m,

l_ex=286 nm, except for 1-Ph: l_ex=300 nm).

8634

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

Table 5. Absorption and fluorescence parameters for 1-R as nano-aggregates in H₂O/EtOH (10:1) solution, including changes relative to molecularly-dis-

solved EtOH solution data in Table 4.

Abs. l_max[nm]^[a]

(shift rel. to EtOH

sol. [cm^ꢀ1])

Fl. l_max[nm]

Transition origin

[nm] (shift rel. to EtOH

sol. [cm^ꢀ1])

Stokes shift [cm^ꢀ1

(change rel. to

]

Osc. strength^[a]

(ratio to EtOH sol.)

Fl. quantum

eff. [%] (ratio to

EtOH sol.)

(shift rel. to EtOH

sol. [cm^ꢀ1])

EtOH sol. [cm^ꢀ1])

1-Me

1-Et

1-Vi

1-Ph

290.5 (ꢀ550)

290.0 (ꢀ610)

287.6 (ꢀ570)

294.0 (ꢀ720)

410.3 (ꢀ650)

419.4 (ꢀ1010)

413.7 (ꢀ830)

411.3 (ꢀ1220)

344.2 (ꢀ830)

347.9 (ꢀ1130)

344.4 (ꢀ940)

347.2 (ꢀ1200)

10050 (+100)

10640 (+400)

10600 (+260)

9700 (+500)

0.28 (ꢅ0.82)

0.25 (ꢅ0.86)

0.20 (ꢅ0.91)

0.28 (ꢅ0.85)

0.37 (ꢅ2.8)

0.38 (ꢅ4.0)

0.36 (ꢅ4.3)

2.3 (ꢅ16.4)

solution and doped-polymer thin films). Whilst this reduc-

tion in photoreactivity may well be the main reason for the

relative increase in F_f, we note that the absolute F_fvalues

are still small (i.e., ranging between 0.36–2.3%). Hence the

dominant fate of the excited state is still non-radiative, and

most of the potential fluorescence gain is countered by an

increase in non-radiative photophysical decay. In terms of

molecular exciton theory,^[52–54]the bathochromic shift of the

transition origin with respect to (molecularly dissolved)

EtOH solutions for all derivatives 1-R here suggests that the

aggregates are composed of J-type chromophores (i.e.,

head-to-tail transition dipole moments, m₁₂), as opposed to

H-type (parallel m₁₂). In that case, molecular exciton

quenching (which occurs for H-type aggregates^[53]) can not

be invoked to explain such an increase in the non-radiative

decay rate, and the mechanism presumably involves inter-

molecular vibrational coupling. We note that the combina-

tion of both a red-shift and fluorescence quenching has

indeed been observed in other reports (e.g., refs. [55,56]). In

any case, the present situation is qualitatively different from

that for phenylated siloles,^[7,8]where the aggregate environ-

ment suppresses non-radiative photophysical decay resulting

in an enhanced F_f.

The fluorescence intensity for 1-Ph is significantly higher

than for the other 1-R derivatives: this result was also repro-

duced in test measurements using acetone/H₂O as the sol-

vent. Note that this higher value is not due to the different

excitation wavelengths used for the data here (l_ex= 286,

300 nm), as evidenced by measurements of the fluorescence

excitation spectrum, which is essentially flat over this l_ex

range. One possible factor concerning this singular behavior

is that the crystallization of 1-Ph during synthesis has been

observed to be particularly efficient. In this case, one can

anticipate that the tendency to form closely packed aggre-

gates may also be higher for 1-Ph relative to the other deriv-

atives reported here.

and emission strength. Here we employ measurements on

both pure powder samples and dilutely doped KBr powder

samples to allow i) determination of the absorbance spectra

(and hence transition origin) in the solid state and ii) quanti-

tative estimates of the fluorescence quantum yields.

The weakly doped KBr powder samples were prepared by

manual grinding of the powder mixtures (with weight con-

centrations of 1-R in the range 0.3–3%, R = Me, Et, Vi,

Ph), such that resulting samples had a surface texture close

to that of an undoped KBr reference sample. The diffuse re-

flectance spectra of the samples were measured using the

photoluminescence spectrometer (see Solid-powder fluores-

cence absorption, Experimental Section) whilst synchro-

nously scanning l_ex= l_em. The angle of detection differs

from the reflection angle of the incident excitation beam,

thus strongly suppressing the specular reflection component

(as was confirmed by comparing the detected signal level

with an aluminum mirror in the sample position). To esti-

mate absorbance spectra, the diffuse reflectivity ratio was

calculated from the spectra of the pure (reference) and

doped KBr samples, followed by the Kubelka–Munk trans-

formation to yield the absorbance (see Experimental Sec-

tion).^[57,58]Test measurements with the KBr samples doped

with the reference compound sodium salicylate (Na-sal)^[59–61]

confirmed that the spectral shape did not depend significant-

ly on dopant concentration (in the range 0.3–3 wt%). The

resulting absorbance spectra for 1-R (1 wt%) are shown in

the left portion of Figure 9. Qualitatively, one can see that

substituents R significantly influence position of band, as

discussed further below.

The corresponding PL spectra are shown in the right por-

tion of Figure 9 (using l_ex= 340 nm). Note that the data

shown are for pure powder samples of 1-R (in order to

obtain a higher signal-to-noise ratio), although measured

spectra for both pure powder and doped KBr samples were

found to be highly consistent (indicating that the grinding

procedure still preserves the microcrystalline environment).

We assign the observed PL as fluorescence from the initially

excited state, as supported by the following observations: i)

the absorption and PL spectra possess a fair degree of re-

flection symmetry; ii) the Stokes shift is not larger, in fact

somewhat smaller, than that in EtOH solution (see below);

iii) time-gated measurements with a resolution of ~10 ms in-

dicated that the PL signal decays with the same rate as the

scattered excitation light (i.e., within the machine resolu-

tion).

As presented in the next section, the fluorescence spectra

of the nano-aggregates here still differ significantly from

that in the solid state powder samples, suggesting that the

aggregates still have a morphology distinct from that of the

(micro-)crystalline phase.

Solid-state photoluminescence: Previous reports on deriva-

tives of 1-R as solid-state powders^[14,35]indicated an appreci-

able photoluminescence (PL) intensity, as well as a signifi-

cant influence of the Si substituents R on the wavelength

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8635

N. Auner et al.

state is smaller than that for solution/nano-aggregates. This

is consistent with a more strongly constrained conformation

in the crystalline environment, where the displacement be-

tween ground- and excited-state equilibrium geometries is

smaller due to additional energy barriers to intramolecular

motion enforced by the surrounding crystal.^[62]

In order to estimate the PL quantum efficiencies F_f, we

employed the procedure outlined in ref. [61] using Na-sal as

a reference standard (additional details given in the Experi-

mental Section). The absorption and emission spectra of

Na-sal are very similar to those of 1-R, minimizing the need

for complicated corrections of the measured data (e.g., due

to the excitation and emission characteristics of the spec-

trometer). This method assumes that scattered excitation

and emitted PL light are collected at the detector with the

same net efficiency, allowing one to estimate the quantum

yield from calculating number of photons absorbed from the

excitation light by dopants (with the aid of an additional un-

doped KBr reference measurement) relative to number of

photons in the emission spectrum. Tests with Na-sal-doped

KBr powder samples with a variable doping fraction 0.3–

3% (corresponding to absorption fractions of 0.44–0.70)

yielded highly consistent raw F_fvalues. (Note that non-

linear relation between concentration and diffuse reflectivity

is also consistent with the Kubelka–Munk expression^[57].)

The raw value determined for Na-sal with this method and

l_ex=340 nm was F_f= 68.6%, which differs from the previ-

ously reported value of 53%.^[61]We attribute this difference

to a UV roll-off in the PL spectrometer emission sensitivity,

and hence in order to provide conservative estimates of F_f

for 1-R, we scaled down their raw values by the correspond-

ing factor (0.77).

Figure 9. Normalized absorption (left) and fluorescence spectra (right) of

1-R in the solid state (microcrystalline powders, R as indicated). Absorp-

tion spectra estimated from diffuse reflectance spectra using Kubelka–

Munk transformation (see text).

The parameters extracted from the spectra are listed in

Table 6. Note that as the peak wavelengths of the absorb-

ance spectra are not readily obtainable from the data at

hand, the transition origins were estimated by determining

the wavelength about which the absorbance and PL spectra

possessed the best reflection symmetry (against wavenum-

ber), with the Stokes shift then estimated as twice the differ-

ence between transition origin and PL peak positions. The

estimated transition origin (for the derivatives where data is

available) varies from 342.9 (1-Me) to 366.6 nm (1-Ph) (a

range equivalent to 1890 cm^ꢀ1), whilst the Stokes shift is rel-

atively constant, varying from 6690 (1-Me) to 7290 cm^ꢀ1(1-

Ph). Hence the predominant effect in tuning the emission

wavelength is due to intermolecular electronic interactions,

as opposed to Stokes shift (excited-state conformation) var-

iations. Moreover, the estimated transition origin is again

red-shifted significantly relative to that in EtOH solution (as

per the nano-aggregate case in the last Section), with shifts

varying from 714 (1-Me) to 2730 cm^ꢀ1(1-Ph). The largest

red-shift of the transition origin between EtOH solution and

solid powder occurs for the phenyl-rich 1-Ph. This would be

consistent with a stronger intermolecular p-interaction be-

tween phenyl groups of adjacent molecules. A comparison

of Tables 4 and 5 indicates that the Stokes shift in the solid-

The calculated F_fvalues are also given in Table 6, and

range from 7.4 (1-Vi) to 21.1% (1-Et). Hence, in the solid-

state, the fluorescence yield is significantly higher than for

the molecules in (EtOH) solution or as nano-aggregates,

and reaches a magnitude with sufficient potential for lumi-

nescence applications. The general trends, in particular the

significantly stronger emission from 1-Et, were consistently

observed with other excitation wavelengths and samples (in-

cluding pure 1-R powder samples). In comparing the data in

Table 6, we find no clear corre-

Table 6. Absorption and PL spectral data for 1-R in the solid state.

lation between F_fand the other

[a]

Fl. l_em[nm]

Transition origin [nm]

(shift rel. to EtOH

sol. [cm^ꢀ1])

Stokes shift [cm^ꢀ1

]

Fl. quantum

eff. [%]

spectroscopic parameters, nor

with respect to the solution/

nano-aggregate data. Inspection

of the structure data in Table 3

also does not reveal any partic-

ular intramolecular feature of

the stilbene subunit geometry

for 1-Et which would readily

account for the higher value of

F_f. Given that the spectroscopic

properties in EtOH solution are

far less dependent on the sub-

stituents R, we conclude that

both F_fand the transition

(shift rel. to EtOH

(change rel. to EtOH

sol. [cm^ꢀ1])

1-Cl

1-H

1-Me

421.1

420.8

389.6

–

6990

ACHTUNGTRENNUNG

–

10

342.9

A

U

1-Et

1-Vi

1-Ph

21

7

G

N

ACHTUNGTRENNUNG

E

ACHTUNGTRENNUNG

9^[b]

A

U

ACHTUNGTRENNUNG

[a] Due to the absence of absorption maxima in the available data, Stokes shifts are estimated as twice the dif-

ference between transition origin and fluorescence peak position. [b] Reported value for 1-Ph is the average

between measurements for two samples, 8.3 and 9.3%.

8636

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

origin are related in a complex way to the crystal structure

and packing, including the role of intermolecular (lattice) vi-

brations in determining the non-radiative relaxation rate.

In order to look for a trend concerning intermolecular in-

teractions in the derivatives 1-R, we analyzed the separation

of the phenyl rings in neighboring molecules (in terms of

their respective centroids), whose p–p or hydrogen-bonding

interactions may play an important role in fluorescence

quenching.^[62,63]For the derivative 1-Vi with the lowest F_f

lowest-energy p–p* band of 1-R in solution in the absence

of oxygen and hydroxyl-group-containing trapping agents,

only the original starting compound was recovered. Howev-

er, in the case where either i) HO-containing species or ii)

oxygen were present, two main stable photoproducts were

found via chemical analysis of the irradiated solutions. As

presented in this section, these two photoproducts could be

identified as the secondary products resulting from the ini-

tial products of two primary photochemical reactions: i)

ring-opening of the four-membered ring due to cleavage of

value (7%), the smallest inter-centroid distance is D_Ph

=

ꢀ

4.98 ꢄ (with a Ph–Ph planar angle of a_Ph= 48.98). A pack-

ing diagram of 1-Vi extended along the direction of close Ph

contacts is shown in Figure 10, and demonstrates that an ef-

fective intermolecular stacking is present for this derivative.

In comparison, for the derivative 1-Et (with the highest

value F_f= 21%) D_Ph= 5.33 ꢄ (a_Ph= 66.48), whilst for 1-

Me (F_f= 10%) D_Ph= 5.00 ꢄ (a_Ph= 86.08). For the deriv-

ative 1-Ph (F_f= 9%), the situation is somewhat more com-

plex, given the additional 1,1-Ph₂substituents and the pres-

ence of three nonequivalent molecules per unit cell

(Table 2). A systematic analysis of the structure data yields

a value of D_Ph= 4.94 ꢄ (55.88), although this is between a

1-Ph and a 3’-Ph group, that is, the Si substituent and a

neighboring stilbenoid Ph group. The shortest intermolecu-

lar distance between two stilbenoid (2–3’) Ph groups is

the 1,4-Si C bond to form a silabutadiene intermediate, and

ii) ring-closure of the stilbenoid subunit to form a dihydro-

phenathrene (DHP) analogue. We provisionally refer the

reader to the deduced photochemical reaction Scheme

shown in Scheme 2c, and discuss the relevant results for

each photochemical pathway in turn in the following subsec-

tions.

Silacyclobutene ring-opening photoreaction: It is well es-

tablished that a electrocyclic ring opening reaction occurs in

the silacyclobutenes DMSCB and Ph-DMSCB (Figure 1b)

when excited in their lowest-energy absorption band (below

~230 nm),^[15–23]analogous to the case of cyclobutene.^[42]The

resultant photoproduct is a short-lived silabutadiene (silene)

which in the absence of trapping agents undergoes thermal

ring-closure to form the starting compound (e.g., with t

~1.5 ms for DMSCB in hexane^[21]). In the presence of hy-

droxyl-bearing species in the solution however, a 1,2-addi-

tion trapping reaction occurs to produce a stable allylsilane,

as depicted in Scheme 2a. In the case of DPCB (Figure 1d)

however, no such photoreaction occurs for excitation into

the p–p* band, and in dilute solutions its fluorescence yield

approaches unity (with F_fdependent on the solvent^[27,31]),

although photodimerization ([2+2] cycloaddition) can occur

at higher concentrations.^[64](Note that a thermal ring open-

ing reaction can occur for ground-state DPCB at elevated

found to be significantly larger, that is, D_Ph

(22.88).

= 5.44 ꢄ

temperatures.^[65,66]

)

One key difference between (Ph)-DMSCB and DPCB is

that the excited chromophores are distinct in character and

excitation energy, that is, for DMSCB the absorption edge is

at ~230 nm (in pentane^[22]) similar to that of cyclobutenes^[25]

where the excitation is concentrated on the four-membered

ring. In the case of DPCB (and 1-R), the excitation of the

chromophore is delocalized onto the exocyclic phenyl

groups and has a far lower energy band edge (~300 nm).

Moreover, it is conceivable that steric constraints imposed

by the phenyl groups might also play a role in inhibiting the

ring opening. On the other hand, in comparing DPCB and

Figure 10. Crystal packing diagram for 1-Vi indicating intermolecular dis-

tance between stilbenoid-Ph-ring centroids.

Taking these results together, we can tentatively identify a

trend that F_fis reduced for smaller D_Ph(for R = Vi, Me,

Et) due to a stronger interaction between Ph groups of the

stilbenoid chromophores. Following this assertion, in the

case of 1-Ph the reduced value of F_fis due to a close con-

tact between stilbenoid and non-stilbenoid Ph groups, which

would result in a qualitatively different intermolecular

quenching mechanism in as far as the Si substituents are es-

sentially spectators in the excited-state dynamics. Based on

these conclusions, there should be scope to optimize F_ffor

these chromophores via crystal engineering of the intermo-

lecular interactions.

ꢀ

1-R, the strained 1,4-Si C bond (see Section on Single-crys-

tal structures) suggests a higher tendency for excited-state

bond cleavage in the latter case. Hence it was not a priori

clear whether the ring-opening photoreaction would pro-

ceed with non-negligible yield for 1-R, although, as shown

in the following, the reaction does indeed take place with

significant yield, as evidenced by the analysis of secondary

products in solutions containing HO-bearing trapping

agents.

Photochemistry in solution and doped-polymer thin films:

Following irradiation with wavelengths corresponding to the

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8637

N. Auner et al.

could not observe the inter-

mediate species directly, the

close analogy of the trapped

secondary products to that for

(Ph-)DMSCB^[15–23]strongly sug-

gests the presence of the inter-

mediate

silabutadiene

5-R

(Scheme 2c). Moreover, the

fact that only the initial starting

compound was isolated in the

absence of trapping agents indi-

cates that 5-R is metastable and

can return to 1-R though ther-

mal ring-closure.

From irradiated solutions of

1-R (R = Me, Ph) in THF/H₂O

(to form 2-R) or THF/MeOH

(to form 3-R), we could isolate

the corresponding photoprod-

ucts by silica gel column chro-

matography, with corresponding

photoreaction yields of 62.3 (2-

Me), 56.7 (2-Ph), 87.8 (3-Me)

and 41.6% (3-Ph). The isolated

samples were colorless oily liq-

uids, and while the water ad-

ducts 2-R are stable in air the

methanol adducts 3-R gradually

degrade to a complex mixture.

The modestly lower net yield of

2-Ph compared to 2-Me would

be consistent with the larger

bulk of the 1,1-Ph₂substituents

in the former case, in terms of

Scheme 2. a) Photochemical ring-opening reaction of 1,1-Me₂-silacyclobtuene (DMSCB) and subsequent trap-

ping of silene intermediate with -OR group.^[17]b) Photochemical ring-closure of cis-stilbene to form metastable

9,10-dihydrophenanthrene (DHP), followed by irreversible hydrogen abstraction by oxygen to form phenan-

threne.^[30]c) Analogous photochemical/trapping reactions deduced for 1-R derivatives and notation for inter-

mediate and secondary photoproducts.

In the first set of photolysis experiments we aimed at iso-

lating sufficient quantities of the photoproducts for chemical

analysis. Typically, solutions of containing ~1 g of 1-R in

THF and methanol (2:1), or THF and H₂O (2:1), were irra-

diated under nitrogen atmosphere in an immersion well

photoreactor with an air-cooled 5 W Zn lamp. The photore-

action flasks were chosen to be of Duran glass to absorb

light below 300 nm, such that the excitation was essentially

from the 307 nm Zn spectral line. Photolysis with quartz

glassware resulted in the formation of a complex mixture of

photoproducts including decomposition products due to the

additional excitation from the 214 nm Zn line. In all cases, a

significant yield of the trapped ring-opened products (2-R

for THF/H₂O and 3-R for THF/MeOH solutions,

Scheme 2c) was observed, as supported by GCMS, TLC and

¹H NMR analysis of the irradiated solutions (see Experi-

mental Section). Importantly, in a more recent photoreac-

tion involving a new derivative with phenylethynyl substitu-

hindering the approach of the HO-trapping moiety to the

silene group in 5-R.

Additional comparative experiments using the nonpolar

solvents n-pentane or n-hexane, in place of THF, shows that

the net formation rate of 2-R/3-R is reduced by about a

factor of 2 compared to that in THF. This is reasonable,

given that the rate of ring-closure of 5-R back to the initial

1-R should also depend on the stabilization of 5-R due to

electrostatic solvent interactions,^[42]such that the lifetime

(and hence trapping reaction probability) would be higher

in (polar) THF.

Measurements of the ¹H NMR spectra of the isolated

photoproducts 2-R/3-R exhibited a clear set of peaks (see

Experimental Section), indicating the formation of only a

single isomer in the trapped photoproducts. One such spec-

trum is shown in Figure 11 for 2-Ph. The pseudotriplet

around 6 ppm is assigned to the vinyl proton, H1. A

ROESY measurement reveals the interaction between H1

and H2, which confirms that the compound is in the E con-

figuration. The proton-NMR spectra of the other ring-

opened photoproducts (2-Me, 3-Me, 3-Ph) indicates that

ꢂ

ents 1-C C-Ph (and H₂O trapping agent) yielded the prod-

ꢂ

uct 2-C C-Ph which could be isolated and recrystallized

ꢂ

from pentane solution. The single-crystal structure of 2-C

C-Ph (given in ref. [14]) unquestionably confirms the identi-

ty of the secondary photoproducts 2-R/3-R. Although we

they also possess a similar E structure. These results are also

[14]

ꢂ

confirmed from the crystal structure 2-C C-Ph, and pro-

8638

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

vide strong evidence for the position of the double bond.

Hence the ring-opening photoreaction and subsequent trap-

ping reaction proceed both chemo- and stereoselectively.

Figure 11. ¹H NMR spectrum of 2-Ph (in CDCl₃, isolated from irradiated

sample of 1-Ph in THF/H₂O solution).

In order to obtain an estimate of the quantum yield F_roof

the primary ring-opening reaction, we performed time-de-

pendent absorption measurements during photolysis of 1-Ph

in pure MeOH.

As the absorption spectrum from an isolated sample of

3-Ph (Figure 12a) was available, we could analyze the reac-

tion kinetics by a linear regression of the time-dependent

spectra. The measurement was performed with a 3 mL

sample (6.3ꢅ10^ꢀ5m, room temperature) in a 1 cm quartz

cuvette equipped with magnetic-bar stirring, with excitation

at 320 nm with a power of 65 mW for a period of 8.1 h (see

Experimental Section).

The time-dependent spectra (each averaged over ~15 min

intervals) are shown in Figure 12b, where one observes a net

reduction in absorption strength and the formation of a

weak structure. These time-dependent spectra could be well

described as a linear combination of the known spectra of 1-

Ph and 3-Ph, allowing us to estimate the formation of the

trapped ring-opened product 3-Ph vs time, as shown in Fig-

ure 12c. In order to demonstrate the relative irradiation

level, we renormalize the time axis in terms of the number

of photons absorbed per 1-Ph molecule present in the

sample (based on the initial optical density).

A detailed model for the photoproduct kinetics would

need to account for i) the primary quantum efficiency for

the ring-opening reaction (F_ro), ii) the lifetime of the inter-

mediate (which can return to the starting product 1-Ph via

ring re-closure, typically on a fairly short time scale of some

ms^[21]), and iii) the rate of encounter and reaction with the

HO-bearing trapping species. For the present case in pure

MeOH, we make the simplifying assumption that the trap-

ping probability is approximately unity, which allows us to

use a first-order irreversible kinetics model based on a

lower-bound estimate of F_ro. The fitted model curve is also

Figure 12. a) UV/Vis absorption spectra of isolated photoproducts 3-Ph

(a) and 4-Ph (c) isolated from irradiated samples of 1-Ph (in pen-

tane) in the presence of either MeOH or O₂trapping agents, respectively.

b) UV/Vis absorption time series during irradiation of 1-Ph in pure

MeOH solution (l_ex=320 nm), where 3-Ph is formed due to ring-opening

reaction (to form 5-Ph) and subsequent MeOH trapping. c) Correspond-

*

ing fraction of 3-Ph from regression analysis ( ; using known spectra for

1-Ph and 3-Ph), and corresponding fit (c), assuming first-order irrever-

sible kinetics and unity trapping probability (F_ro=7.1%). Time base re-

scaled to correspond to number of photons absorbed/molecule in sample.

quantum yield for these compounds for comparison. Howev-

er, data are available for example, for various 1,2-dimethyl-

cyclobutene derivatives (in hexadecane),^[67]where F_ro

ranged from 5–26% (totals including different isomers),

which encompasses our lower bound estimate here for 1-Ph

in MeOH. From the results here we can state that the ring-

opening photoreaction accounts for at least 7.1% of the

fluorescence quenching in solution, although this implies

that other non-radiative relaxation pathways must also be

present given the low values of F_f~0.1%.

Stilbene-subunit ring-closure photoreaction: In the absence

of HO-containing trapping species and in the presence of O₂

a different stable reaction product was observed upon exci-

tation of the p–p* band of 1-R. The formation of this reac-

tion product was first identified by its strong fluorescence

spectrum (see below), which rapidly inundates the weak

fluorescence from 1-R. Given that this fluorescence spec-

trum was very similar to that of phenanthrene,^[68]and the

necessary presence of oxygen, we anticipated that the pho-

toreaction was analogous to that in cis-stilbene,^[29,30]that is,

i) an electrocyclic ring-closure reaction to form a DHP inter-

mediate followed by ii) hydrogen abstraction by O₂to yield

phenanthrene, as shown in Scheme 2b.

shown in Figure 12c and corresponds to a value of F_ro

=

7.1%.

While the net reaction yields and trapping rates (for vari-

ous trapping agents) have been reported for (Ph-)-

DMSCB,^[20,21]we did not find any estimates of the initial

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8639

N. Auner et al.

In order to isolate sufficient quantities of this photoprod-

uct for chemical analysis, we again began with the prolonged

photolysis of relatively large samples (see Experimental Sec-

tion). From a starting solution of 1-Ph (0.5 g) in toluene

(30 mL) (with a slow bubbling of oxygen), we isolated the

photoproduct 4-Ph from the irradiated solution via silica gel

column chromatography, with a net yield of 56.6%. The

1

identity of 4-Ph was confirmed by a series of 1D and 2D H,

¹³C and ²⁹Si NMR measurements (see Experimental Sec-

tion). Given that no formation of 4-R was observed for irra-

diation of solutions in the absence of O₂, we conclude that

4-R is generated upon hydrogen abstraction from the pri-

mary DHP-analogue photoproduct 6-R, as shown in

Scheme 2c. The absorption spectrum of an isolated sample

of 4-Ph is shown in Figure 12a, and is very similar to that re-

1

ported for phenanthrene,^[68]being composed of a strong A₁^þ

transition near 250 nm, with progressively weaker shoulders

1

at lower energies due to the ¹B₂^þand A₁^ꢀtransitions.^[42]

In order to gain further quantitative information concern-

ing the net formation of 4-Ph, we carried out a set of time-

dependent fluorescence measurements during irradiation of

1-Ph in hexane (1.5ꢅ10^ꢀ6m, with ambient oxygen content

and magnetic-bar stirring). In this case, the sample remained

in the PL spectrometer, and the excitation beam served as

both the photochemical excitation for 1-Ph and fluorescence

excitation for 4-Ph (l_ex=300 nm, P = 22.4 mW).

Figure 13. a) Absorption and time-dependent fluorescence spectra (aris-

ing from photoproduct 4-Ph) during 300 nm irradiation of 1-Ph in

hexane (1.5ꢅ10^ꢀ6m, ambient oxygen content). b) Corresponding forma-

tion of 4-Ph determined from the integrated PL intensity during irradia-

*

tion (exptl , including a fitted curve (c) based on a simplified first-

order kinetics model (with

a net secondary formation efficiency of

0.30%). Absolute fraction of 4-Ph determined from absorption spectrum

fitting in a).

The time-dependent fluorescence spectra are shown in

Figure 13a, right, where the growth of the characteristic

phenanthrene-like spectrum is observed between ~350–

450 nm, which possesses a pronounced vibronic structure

with a peak spacing of ~1300 cm^ꢀ1. In order to calibrate the

population of 4-Ph from the fluorescence spectra, we mea-

sured the absorption spectra before and after irradiation,

also shown in Figure 13a, left. The absorption spectrum

after irradiation could be fit with the known spectra for iso-

lated samples of 1-Ph and 4-Ph (the fit curve is also includ-

ed in Figure 13a), allowing us to deduce the final population

fractions of 46 and 54%, respectively. Using the initial and

final fractions, we used the integrated intensities of each

fluorescence spectrum to determine the kinetics of 4-Ph,

which is shown in Figure 13b (again with the time axis nor-

malized to the relative number of absorbed photons).

We can estimate an effective net quantum yield of forma-

tion F_secof the secondary photoproduct 4-Ph by fitting this

curve assuming irreversible kinetics, which results in the

fitted curve in Figure 13b and a value of F_sec=0.3%. In in-

terpreting this value, one should note that the intermediate

6-R can also most likely undergo an efficient ring-opening

photoreaction back to 1-R via absorption of the excitation

light, as per DHP.^[30]Hence the photochemical quantum

yield of 6-R (F_rc) is expected to be significantly larger than

F_sec, the latter including the competition between the back-

photoreaction to 1-R and reaction with the (diffusion-limit-

ed) O₂-trapping reaction. Further time-dependent studies

with anaerobic solutions would be required to estimate the

initial quantum efficiency F_rc, using absorption measure-

ments to detect the presence of the DHP-analogue 6-Ph

(which should possess a broad absorption band in the visible

range,^[30]although this was not detectable in our measure-

ments here).

Photochemistry in doped-polymer thin films: In evaluating

the potential of the chromophores 1-R for solid-state lumi-

nescence applications, we wished to study their photophysi-

cal/photochemical properties as dopants in polymer thin

films. Given that the polymer host environment is inter-

mediate between solution and solid-state, it was not clear to

what degree photochemical reactions might proceed.

We prepared a series of polymer thin films using polystyr-

ene (PS) as the host with a variable doping of 1-Ph, using

spin-casting from CHCl₃solutions (~50 mgmL^ꢀ1) which

yielded high quality films of some 100 nm thickness. The ab-

sorption and fluorescence spectra of the films are shown in

Figure 14b–d for doping weight fractions of 4, 10 and 46%

(along with the solution, aggregate and solid-power spectra

of the previous sections to allow a more direct comparison).

Note that the films (but not precursor solutions) were mod-

erately exposed to UV light in recording the fluorescence

spectra, such that photoreactivity could take place. The ab-

sorption spectra are relatively clean and consistent with that

in solution phase, demonstrating the optical quality of the

film. The minor contribution from the PS host (main absorp-

tion edge is at ~275 nm) was subtracted using the scaled ab-

sorption of an undoped film.

In the fluorescence spectrum for the 4%-doped sample, a

clear signature from the photoproduct 4-Ph can be seen on

the short-wavelength side (as shown for an irradiated THF

8640

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

green spectral region, with fluorescence quantum efficien-

cies presently reaching F_f=21% for the derivative 1-Et. We

conclude that the dependence of F_f(and l_em) on the 1,1-

substituents is primarily due to crystal packing effects, with

a tentative trend identified in terms of the intermolecular in-

teractions between Ph groups. This suggests that there is

scope to further enhance the fluorescence yield or tune the

emission wavelength by crystal engineering, for example, via

the use of bulkier substituents. The similarity between the

experimental absorption/fluorescence bands and calculated

electronic structure/excitations indicate that the chromo-

phore is essentially of stilbenoid character with only a small

effect from Si on the relevant orbital energies.

However, the inclusion of Si does lead to significant

changes in the photoreactivity in solution compared to

ꢀ

DPCB. The presence of the strained 1,4-Si C bond leads to

the ring-opening photoreaction to form a silabutadiene in-

termediate 5-R (as per other non-stilbenoid SCBs) as well

as a modification of the stilbene subunit conformation such

that the ring-closure of between the 2,3-Ph groups to form a

DHP-analogue intermediate 6-R (as per DPCP) also can

occur. While both intermediates can return to the initial

compound 1-R, the formation of the stable secondary pho-

toproducts 2-R and 4-R due to subsequent trapping reac-

tions in the presence of H₂O or O₂, respectively, suggests

that these compounds could be used for photo-activated

chemical sensing. These compounds would also be interest-

ing candidates for ultra-fast excited-state spectroscopy in

order to study the competition between the two primary

photoreaction pathways (and provide a contrast to cis-stil-

bene, where DHP formation competes instead with cis–trans

photoisomerization).

Besides their use in (crystalline) solid-state applications,

another important synthetic direction is towards the inclu-

sion of the 1-R chromophores in polymers, such as carbo-

and/or polysilanes and silicones (either as pendant groups or

in the main chain), for thin-film luminescence applications.

The use of the Si groups provides synthetic benefits/alterna-

tives to their all-carbon analogues. While the photoreactivity

channels (which appear to be responsible for at least ~10%

of the non-radiative relaxation in solution) could be sup-

pressed by the synthetic addition of intramolecular con-

straints, it is still not clear what the resulting residual frac-

tion of non-radiative photophysical decay (which is domi-

nant in the case of nano-scale aggregates) would be. Further

quantitative studies of the precise photochemical and non-

radiative yields in solutions and thin films are required to

address this issue.

Figure 14. Comparison of UV/Vis absorption and fluorescence spectra of

1-Ph samples (including a variable component of fluorescence from pho-

toproduct 4-Ph) vs environment. For graphs a)–d), that is, THF solution/

doped-PS thin films, absorption spectra are for fresh samples, whilst PL

spectra are recorded following moderate UV exposure. For e)–f) (aggre-

gate in EtOH/H₂O solution and solid powder), no significant photochem-

istry is observed. For the solid powder sample f) the absorption is esti-

mated from the diffuse reflectance spectrum (i.e., as per Figure 9).

solution in Figure 14a), superposed with a fairly smooth dif-

fuse red-shifted component which extends to above 500 nm.

As the dopant concentration goes from 4 to 46%, a clear

trend is evident, in that the contribution from the photo-

product 4-Ph is suppressed and the red-shifted component

begins to dominate. In comparing the fluorescence spectrum

for 46%-doping with that for nano-aggregates in EtOH/

H₂O solution (Figure 14e), one observes that they are

almost indistinguishable. Hence we conclude that aggrega-

tion of the 1-Ph chromophores occurs in the preparation of

the thin films, although the composite precursor solutions

were prepared by mixing well-dissolved CHCl₃solutions of

both 1-Ph and PS followed by rigorous stirring. Such aggrega-

tion behavior in molecularly doped polymer films is well

known.^[69,70]Whether the aggregates here form directly in the

precursor solution or whether this occurs during the evapora-

tion of the cast films is not clear at this stage. However, it is

interesting to note that from the strong similarity of the spec-

tra, it appears that the aggregates formed both in EtOH/H₂O

and in PSACHTUNGTRENNUNG(CHCl₃) blends must have a very similar morphol-

ogy, despite the fact that the chemical environment (viscos-

ity, polarity, etc.) in the two cases is very different.

Experimental Section and Analysis

Synthetic procedures

Conclusions

General: Unless otherwise noted, all starting materials are commercially

available and are used without further purification. All reactions involv-

ing organometallic compounds were carried out under an overpressure of

argon using standard Schlenk techniques. THF and Et₂O were refluxed

We have demonstrated that the compounds 1-R have poten-

tial for use as solid-state fluorophores emitting in the blue-

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8641

N. Auner et al.

and distilled from Na/K alloy and benzophenone under an argon atmos-

phere. Pentane was refluxed and distilled from K metal. All starting ma-

terials were purchased from Fluka. The dichlorosilacyclobutene 1-Cl was

synthesized according to literature procedures.^[38]

13.2 Hz), 1.24 (dd, 1H, CH₂, ³J=12.4 Hz, ²J=13.2 Hz), 0.78 ppm (s, 9H,

CACTHNUTRGNEUNG

(CH₃)₃); ¹³C NMR: d = 158.45 (C_olef, C2), 142.02 (C_olef, C1), 137.79

(C_ar), 137.02 (C_ar), 136.97 (CH_2,olef), 136.13 (CH_2,olef), 133.76 (CH_olef),

133.61 (CH_olef), 128.24 (2ꢅCH_ar), 128.13 (2ꢅCH_ar), 128.03 (2ꢅCH_ar),

127.97 (2ꢅCH_ar), 127.61 (CH_ar), 126.28 (CH_ar), 40.79 (CH₂), 30.83 (C-

¹H and ¹³C NMR spectra were measured on a Bruker AM 250 (250 MHz

for ¹H and 62.9 MHz for ¹³C) spectrometer and ²⁹Si NMR on a Bruker

DPX 250 (49.7 MHz) spectrometer with chloroform-d as solvent at room

temperature. Chemical shifts are reported in d (in ppm) with reference

relative to the CDCl₃peak for ¹³C and TMS peak for ¹H and ²⁹Si. Ele-

ment microanalyses were performed by at the microanalysis laboratory

of Anorganisch-Chemisches Institut der Technischen Universitꢂt Mꢁn-

chen and elemental analysis center at the Institut fꢁr Chemie der Hum-

boldt-Universitꢂt zu Berlin. Melting points were determined in opening

capillary on a STU-SMP1 melting point apparatus.

A

E

d

=

ꢀ14.03 ppm; elemental analysis calcd (%) for C₂₄H₂₈Si: C 83.66, H 8.19;

found: C 82.16, H 7.88.

1,1,2,3-Tetraphenyl-4-neopentyl-1-silacyclobut-2-ene (1-Ph): In a 250 mL

three-necked flask were placed magnesium shavings (3.3 g, 136 mmol)

and dry THF (300 mL). Under intensively stirring, a solution of bromo-

benzene (14.3 mL, 136 mmol) in dry THF (50 mL) was added slowly.

After addition finished, the solution was warm up and refluxed for 2 h.

When the magnesium shavings were completely dissolved, the phenyl-

magnesium bromide solution was cooled down to room temperature.

1,1-Dihydro-2,3-diphenyl-4-neopentyl-1-silacyclobut-2-ene (1-H): In

a

250 mL three-necked flask were placed LiAlH₄(1.05 g, 27.7 mmol) and

dry Et₂O (100 mL). The reaction flask was cooled to 08C by ice bath.

Under stirring, a solution containing 1-Cl (5.0 g, 13.8 mmol) in dry Et₂O

(50 mL) was added dropwise over a period of 1 h. The suspension mix-

ture was warmed up and refluxed for 15 h, and then filtered. The filtrate

was removed the Et₂O by vacuum and the residue was extracted by dry

pentane (100 mL). The pentane solution was contracted by vacuum and

the residue was distilled under high vacuum at around 1028C to obtain a

white waxy solid of 1-H (2.8 g, 9.4 mmol, 68.2%). M.p. 718C; ¹H NMR:

d = 7.41–7.25 (m, 10H, 2ꢅC₆H₅), 5.13 (dd, 1H, SiH, ²J=14.7 Hz, ³J=

A solution of 1-Cl (20 g, 55 mmol) in dry THF (50 mL) was added drop-

wise to above prepared phenylmagnesium bromide solution. After reflux-

ing for 6 h, the reaction mixture was cooled down to room temperature

and treated with ice. The volatiles were removed by evaporation under

reduced pressure and the residue was extracted with Et₂O (100 mL) and

6% HCl (100 mL). The organic phase was collected and washed with

water and brine, then was dried over MgSO₄and evaporated in vacuo to

remove volatiles. Recrystallization from cold n-pentane (ꢀ358C) yielded

white needle crystals of 1-Ph (22.6 g, 51 mmol, 92%). M.p. 768C;

¹H NMR: d

=

8.01–7.23 (m, 20H, 4ꢅC₆H₅), 3.12 (dd, 1H, CH, ³J=

2

2.3 Hz), 4.94 (dd, 1H, SiH, J=14.7 Hz), 2.67 (m, 1H, CH), 1.79 (m, 1H,

3

CH₂), 1.48 (m, 1H, CH₂), 0.91 ppm (s, 9H, CACTHNUTRGNEUNG

(CH₃)₃); ¹³C NMR: d =

12.1 Hz, J=1.9 Hz), 1.74 (m, 1H, CH₂), 1.05 (m, 1H, CH₂), 0.85 ppm (s,

9H, CACHTNUGRTNEUNG

(CH₃)₃); ¹³C NMR: d = 159.06 (C_olef, C2), 142.06 (C_olef, C1), 137.57

158.78 (C_olef, C2), 138.68 (C_olef, C1), 137.39 (C_ar), 136.67 (C_ar), 128.29 (2ꢅ

CH_ar), 128.19 (2ꢅCH_ar), 128.03 (2ꢅCH_ar), 127.83 (CH_ar), 127.73 (2ꢅ

(C_ar), 136.97 (C_ar), 134.42 (C_ar), 133.64 (C_ar), 136.54 (2ꢅCH_ar), 135.10 (2ꢅ

CH_ar), 128.33 (2ꢅCH_ar), 128.26 (2ꢅCH_ar), 128.18 (2ꢅCH_ar), 128.14 (2ꢅ

CH_ar), 128.05 (2ꢅCH_ar), 127.86 (2ꢅCH_ar), 130.07 (CH_ar), 129.89 (CH_ar),

CH_ar), 126.59 (CH_ar), 43.21 (CH₂), 31.42 (CACHTUNTRGENNGU(CH₃)₃), 29.44 (CACHTUNGTRENN(UNG CH₃)₃),

24.56 ppm (CH, ¹J_SiC= 46.4 Hz); ²⁹Si NMR: d = ꢀ33.22; elemental anal-

127.72 (CH_ar), 126.40 (CH_ar), 41.50 (CH₂), 30.71 (C

29.61 ppm (C

(CH₃)₃); ²⁹Si NMR: d

calcd (%) for C₃₂H₃₂Si: C 86.43, H 7.25; found: C 87.53; 7.36.

ACHTUGNTRENUN(GN CH₃)₃), 30.19 (SiCH),

ꢀ7.84 ppm; elemental analysis

ysis calcd (%) for C₂₀H₂₄Si: C 82.13, H 8.27; found: C 81.47, H 8.11.

G

=

1,1-Diorgano-2,3-diphenyl-4-neopentyl-1-silacyclobut-2-enes (1-Me, 1-Et,

1-Vi): 2.2equiv alkylmagnesiumchloride (60.9 mmol) in THF solution was

slowly added to a solution of 1-Cl (10 g, 28 mmol) in dry THF (150 mL)

in a 500 mL three-necked flask. The reaction mixture was stirred at room

temperature for 20 h, and then treated with ice. The volatiles were re-

moved by evaporation under reduced pressure and the residue was ex-

tracted with Et₂O (200 mL) and 6% HCl (200 mL). The organic phase

was collected and washed with water and brine, then was dried over

MgSO₄and evaporated in vacuo to remove volatiles. The residue was re-

crystallized from n-pentane to yield white crystal products 1-Me, 1-Et

and 1-Vi.

Crystal structure determination

X-ray-qualified crystals of 1-H were grown from a concentrated pentane

solution. Crystals of 1-Me, 1-Et, 1-Vi and 1-Ph were obtained by evapo-

rating a mixture solution of ethanol and hexane in ambient conditions. In

each case, the crystal was mounted in inert oil. Data collection was per-

formed at 193(2) K with an area-detector using graphite-monochromated

Mo Ka radiation (0.71073 ꢄ). The structures were solved by direct phase

determination and refined by full-matrix least-squares techniques against

2

F

with the SHELXL-97 program system.^[71–73]Hydrogen atoms were

1-Me: Yield 97%, colorless prime crystals, m.p. 1048C; ¹H NMR: d =

placed in the calculated positions, and all other atoms were refined aniso-

tropically. CCDC 729987 (1-Cl), 729988 (1-H), 729989 (1-Et), 729990 (1-

Vi) and 729991 (1-Ph) contain the supplementary crystallographic data

for this paper. These data can be obtained free of charge from The Cam-

bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_

request/cif. Details on parameters are given in Tables 1 and 2.

3

7.56–6.89 (m, 10H, 2ꢅC₆H₅), 2.23 (dd, 1H, CH, J=12.2 Hz, J=2.2 Hz),

1.41 (dd, 1H, CH₂, ³J=1.9 Hz, ²J=13.6 Hz), 1.18 (dd, 1H, CH₂, ³J=

12.4 Hz, ²J=13.6 Hz), 0.85 (s, 9H,

0.34 ppm (s, 3H, SiCH₃); ¹³C NMR:

CACHTGNUTERNNU(G CH₃)₃), 0.45 (s, 3H, SiCH₃),

2

d

= 155.33 (C_olef, C2, J_SiC=

13.1 Hz), 144.98 (C_olef, C1, ¹J_SiC=50.6 Hz), 138.52 (C_ar), 136.94 (C_ar),

128.14 (2ꢅCH_ar), 127.98 (2ꢅCH_ar), 127.72 (2ꢅCH_ar), 127.30 (CH_ar),

Photolysis and photoproduct analysis

125.99 (CH_ar), 41.14 (CH₂), 30.62 (CACHTNURGTNENUG(CH₃)₃), 29.55 (CAHCUTNGTREN(NUGN CH₃)₃), 29.40

(SiCH), ꢀ1.47 (CH₃, ¹J_SiC=45.1 Hz), ꢀ2.22 ppm (CH₃, ¹J_SiC=47.9 Hz);

²⁹Si NMR: d = 5.49 ppm; elemental analysis calcd (%) for C₂₂H₂₈Si: C

82.43, H 8.80; found: C 81.03, H 8.78.

Ring-opening reaction: Photolysis was performed with an immersion well

apparatus of 35 mL volume made by Duran glass. Runs with THF as the

solvent were conducted at room temperature with a 5 W zinc lamp

(307 nm), which was cooled by a stream of air. All photolysates were

purged with nitrogen for 1 h prior to photolysis.

1

1-Et: Yield 91%, colorless cubic crystals, m.p. 668C; H NMR: d = 7.21–

3

7.00 (m, 10H, 2ꢅC₆H₅), 2.28 (dd, 1H, CH, J=11.7 Hz, J=1.7 Hz), 1.41

(dd, 1H, CH₂, ³J=1.8 Hz, ²J=13.6 Hz), 1.25 (dd, 1H, CH₂, ³J=11.7 Hz,

3-Ph: The starting solution contained 1-Ph (1 g) in a solvent mixture of

THF (20 mL) and MeOH (10 mL). During irradiation, aliquots were

taken at time intervals of two days for TLC analysis. After two months

reaction time, the solvent was removed by vacuum and the resulting mix-

ture was subjected to column chromatography on a silica gel (R_f=0.62,

acetone/hexane 1:10) to give a colorless oil of 3-Ph (0.45 g, 41.6%).

¹H NMR: d = 7.57–6.82 (m, 20H, 4ꢅC₆H₅), 5.96 (pseudo-t, 1H, ³J=

15 Hz, =CH), 3.66 (s, 1H, Si-CH), 3.49 (s, 3H, OCH₃), 1.79 (d, 2H, ²J=

²J=13.6 Hz), 1.13 (t, 3H, CH₃), 0.96 (t, 3H, CH₃), 0.83 (s, 9H, C

ACTHNUGTRNEUGN(CH₃)₃),

1.05–0.72 ppm (m, 4H, 2ꢅCH₂); ¹³C NMR: d = 159.69 (C_olef, C2), 143.98

(C_olef, C1), 138.99 (C_ar), 137.02 (C_ar), 128.12 (2ꢅCH_ar), 128.10 (2ꢅCH_ar),

128.06 (2ꢅCH_ar), 127.90 (2ꢅCH_ar), 127.22 (CH_ar), 125.81 (CH_ar), 40.96

(CH₂), 30.65 (C

ACHTUNGTRENNUNG(CH₃)₃), 29.56 (CACHTUNGTRNE(NUGN CH₃)₃), 27.49 (SiCH), 8.41 (CH₂CH₃),

7.62 (CH₂CH₃), 5.75 (SiCH₂), 5.38 ppm (SiCH₂); ²⁹Si NMR:

d

=

11.22 ppm; elemental analysis calcd (%) for C₂₄H₃₂Si: C 82.69, H 9.25;

found: C 80.37, H 8.86.

7.2 Hz, CH₂), 0.69 ppm (s, 9H, CACHTNUGRTNEUNG

(CH₃)₃); ¹³C NMR: d = 143.67, 140.36,

1-Vi: Yield 91%, colorless cube-like crystals, m.p. 638C. ¹H NMR: d =

7.21–7.00 (m, 10H, 2ꢅC₆H₅), 6.41–5.83 (6H, 2ꢀCH=CH₂), 2.40 (dd, 1H,

CH, ³J=12.1 Hz, ³J=2.3 Hz), 1.39 (dd, 1H, CH₂, ³J=2.3 Hz, ²J=

140.33, 135.54, 135.39, 134.28, 133.61, 129.74, 129.69, 129.60, 129.56,

128.11, 127.80, 127.67, 127.64, 127.44, 126.00, 125.15, 52.01, 46.60, 43.49,

31.17, 29.33 ppm (CACHTUNGTRENNUNG

8642

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

ꢂ

Compounds 3-Me, 2-Ph and 2-Me and 2-C C-Ph were obtained using a

similar procedure using the corresponding reactants 1-R, with THF satu-

transition origin value used in this model should be somewhat blue-shift-

ed compared to the true transition origin. Instead, we chose to fit the ab-

sorption bands using a nominal value of n˜_vib= 1600 cm^ꢀ1, which corre-

sponds well to the dominant displaced high-frequency mode found in

TD-DFT excited-state calculations for 1-Me due to a combination of C=

ꢂ

rated with distilled water as the solvent for 2-Ph and 2-Me and 2-C C-

Ph.

3-Me: Colorless oil, R_f

= 0.73 (acetone/hexane 1:10); 0.97 g, 87.8%;

¹H NMR: d = 7.27–6.96 (m, 10H, 2ꢅC₆H₅), 5.92 (pseudo-t, 1H, ³J=

ꢀ

C and phenyl C C stretch vibrations, (and is also characteristic of the

resonance Raman spectrum for cis-stilbene^[75,76]).

2

15 Hz, =CH), 3.36 (s, 3H, OCH₃), 3.16 (s, 1H, SiCH), 1.93 (d, J=7.6 Hz,

CH₂), 0.90 (s, 9H, CCH₃), 0.18 (s, 3H, SiCH₃), 0.16 ppm (s, 3H, SiCH₃);

Solution fluorescence quantum yield determination: Photoluminescence

(PL) spectra were recorded on a Perkin–Elmer LS 50B luminescence

spectrometer at ambient temperature. For the calculation of fluorescence

quantum efficiencies F_f^[43]of 1-R in solution, we employed the standard

¹³C NMR: d

= 143.67, 141.49, 141.25, 128.84, 128.21, 128.04, 127.74,

127.24, 126.03, 125.00, 50.77, 48.61, 43.31, 31.37, 29.49 (C

G

(SiCH₃), ꢀ2.58 ppm (SiCH₃); ²⁹Si NMR: d = 13.5 ppm.

compound 2,5-diphenyloxazole (PPO) in EtOH (7.8ꢅ10^ꢀ5m, l_ex

=

2-Ph: Colorless oil, R_f=0.13 (acetone/hexane 1:10); 0.59 g, 56.7%;

¹H NMR: d = 7.64–6.85 (m, 20H, 4ꢅC₆H₅), 5.92 (pseudo-t, 1H, ³J=

15 Hz, =CH-), 3.69 (s, 1H, Si-CH), 2.27 (s, 1H, OH), 1.78 (d, 2H, ²J=

300 nm, obtained from Radiant Dyes GmbH and used without further

purification) and assumed the F_fvalue of unity reported for PPO in cy-

clohexane.^[77]Tests with different concentrations and excitation wave-

lengths were performed to ensure consistent results were obtained and

the F_fcalculation includes the necessary correction for the excitation

density corresponding to the right-angle geometry used, whereas correc-

tions due to re-absorption were generally negligible due to the large

Stokes shift for 1-R.

7.6 Hz, CH₂), 0.66 ppm (s, 9H, CH₃); ¹³C NMR: d

= 143.38, 140.52,

139.96, 135.43, 134.65, 134.58, 129.77, 129.69, 129.48, 129.26, 128.11,

128.07, 127.80, 127.75, 127.52, 126.17, 125.36, 47.42, 43.41, 31.09,

29.27 ppm (CACHTUNGTRENNUNG

2-Me: Colorless oil, R_f=0.25 (acetone/hexane 1:10); 0.66 g, 62.3%;

¹H NMR: d = 7.33–7.00 (m, 10H, 2ꢅC₆H₅), 5.87 (pseudo-t, 1H, ³J=

15 Hz, =CH), 3.16 (s, 1H, SiCH), 1.92 (d, ²J=7.7 Hz, CH₂), 1.62 (s, 1H,

OH), 0.89 (s, 9H, CCH₃), 0.22 (s, 3H, SiCH₃), 0.16 ppm (s, 3H, SiCH₃);

Solid-powder fluorescence absorption and quantum yield determination:

In order to establish the absorbance spectra of the powders 1-R, weakly

doped KBr samples were prepared by co-grinding, and the samples

mounted in the PL spectrometer pressed under a quartz window at an

angle away from 458 (such that the detected specular reflection was

strongly suppressed). The diffuse reflection spectra R(l) were measured

by synchronously scanning both l_ex=l_emand taking the ratio of the back-

ground spectrum for pure KBr, and transformed to absorbance A(l)

using the Kubelka–Munk transformation A=(1ꢀR)²/(2R).^[57,58]As a ref-

erence powder substance, we used sodium salicylate (Na-sal) obtained

from Sigma–Aldrich (ꢃ99.5%) and used without further purification.

The absorbance spectra A(l) obtained for Na-sal-doped KBr samples

with doping in the ranges 0.3–3 wt% were highly consistent.

¹³C NMR: d

= 143.47, 141.78, 141.05, 128.74, 128.30, 128.23, 127.91,

126.75, 126.23, 125.20, 49.74, 43.22, 31.35, 29.50 (C

(CH₃)₃), ꢀ0.41

(SiCH₃), ꢀ0.50 ppm (SiCH₃); ²⁹Si NMR: d = 12.54 ppm.

ꢂ

2-C C-Ph: See ref. [14].

Ring-closing reaction: The same photoreaction apparatus as described

above was used. The starting solution contained 1-Ph (0.5 g, 1.12 mmol)

in toluene (30 mL). The toluene was distilled from sodium before use.

Pure oxygen was slowly bubbled through the solution during the reaction

and dry toluene was added periodically to maintain the solution volume.

After one month reaction time, the solvent was removed by vacuum and

the resulting mixture was subjected to silica gel column chromatography

(R_f=0.55, acetone/hexane 1:15) to give a colorless oil of 4-Ph (0.28 g,

56.6%). ¹H NMR: d = 8.75 (m, 2H, phenanthrene), 8.08–7.36 (m, 16H,

Ph, phenanthrene), 3.45 (m, 1H, CH), 2.27 (m, 1H, CH₂), 1.71 (m, 1H,

For the determination of F_ffor 1-R (l_ex=340 nm), we employed the pro-

cedure outlined in ref. [61]. In this procedure, the emission spectrum of

both undoped and doped KBr samples are measured, including the spec-

tral region of the scattered pump light. The raw F_fvalue is determined

from the ratio of the number of photons in the integrated PL spectrum

and the number of photons missing from the scattered pump peak in the

doped sample.

CH₂), 0.98 ppm (s, 9H, CH₃); ¹³C NMR: d

= 154.23 (C_olef), 138.92,

136.63, 135.17, 134.38, 133.60, 133.07, 132.45, 130.45, 130.20, 129.97,

129.62, 128.50, 128.17, 127.85, 127.16, 127.02, 126.45, 126.19, 124.47,

123.82, 123.23, 43.06 (CH₂), 31.49, 29.58, 28.22 ppm; ²⁹Si NMR: d

ꢀ3.36 ppm.

=

This procedure for Na-sal with doping concentrations of 0.3 and 3 wt%

yielded very consistent estimates of F_f. The raw value obtained was F_f=

68.6% (l_ex=340 nm) which differs somewhat from the literature value

F_f=53%, which we attribute to the short-wavelength roll-off in the

emission detection sensitivity. Given that both 1-R and Na-sal have very

similar emission spectra (and hence correction ratio), we scaled down the

raw F_fvalues obtained for 1-R by this same ratio (0.53/0.686=0.77).

UV/Vis spectroscopic data

Solution absorption band modeling: Following the notion that the ob-

served non-Gaussian absorption band-shapes in the Section on Photophy-

sics in solution and Effect of aggregation in solution correspond to an un-

resolved Franck-Condon progression, we employed a band-shape func-

tion corresponding to a single displaced harmonic vibration (using the

general finite-temperature expression, assuming equal displaced vibra-

tional frequencies n˜_vibin both ground and excited states^[74]). To generate

the final smooth model curves during fitting, this discrete Franck-Condon

progression is convolved with a relatively narrow Gaussian function (i.e.,

with a width small compared to the overall Franck-Condon bandwidth)

to represent the effect of additional lower-frequency modes and solvation

broadening.^[74]The fitting parameters for such a model band can be for-

mulated in terms of n˜_vib, Dg (normalized displacement), n₀(effective tran-

sition origin) and the peak absorption strength (which can be expressed

in terms of the transition oscillator strength f). Due to the near co-de-

pendency of these parameters, for the fitting procedure here one requires

at least one of n˜_vib, Dg or n₀to be fixed to an independently determined

value in order to obtain well-defined fitted values for the other parame-

ters. Although we have estimates for the transition origin n₀from a re-

flection comparison of absorption and corresponding fluorescence spec-

tra (see below), this value could only be used in a model which treats all

the significantly displaced modes explicitly. For the model at hand, with a

single effective high-frequency mode, a significant contribution to the

Stokes shift can be expected to arise from additional low- and medium-

frequency modes which are not resolvable in the spectrum and are ab-

sorbed into the width of the Gaussian convolution - hence, the effective

Time-dependent absorption photolysis measurements: The time-depen-

dent absorption measurements during photolysis of 1-Ph in MeOH was

performed with a 3 mL sample (6.3ꢅ10^ꢀ5m) in a 1 mm quartz cuvette

equipped with a small magnetic-bar stirrer, with excitation at 320 nm

from a fiber-coupled Hg high-power arc lamp (Oriel 66942), selected

using a bandpass interference filter and providing 65 mW power at the

sample (measured using a calibrated optical power meter). The absorp-

tion spectra were measured using a fiber-coupled Xe source (Ocean

Optics PX-2) to provide both a probe and reference beam, with the rela-

tive transmission spectra measured using a two-channel CCD spectrome-

ter (Ocean Optics S2000).

Quantum chemistry calculations

Gas-phase molecular structures in the electronic ground state were

energy-minimized with the quantum chemistry program Gaussian03^[78]on

the density functional theory (DFT) level with the three-parameter

hybrid functional B3LYP^[79,80]which employs in the exchange part a

hybrid of Dirac–Slater local exchange^[81,82]extended by Beckeꢀs gradient

correction^[83]and of non-local Fock exchange^[84]as well as in the correla-

tion part the local correlation contribution^[85]together with the Lee–

Yang–Parr generalized gradient correlation term.^[86]The local correlation

contribution that is default in the B3LYP implementation of Gaussian03

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8643

N. Auner et al.

(denoted sometimes as VWN3 as to distinguish it from the recommended

local correlation contribution, which is then referred to as VWN5) was

used. A triple-zeta valence basis set (TZV)^[87]of nucleus-centered con-

tracted Gaussian functions augmented by a single set of polarisation

[4] A. Kraft, A. C. Grimsdate, A. B. Holmes, Angew. Chem. 1998 , 110 ,

41 6 –4 43; Angew. Chem. Int. Ed. 1998 , 37 , 40 2 –4 28.

[5] S. Yamaguchi, K. Tamao, J. Chem. Soc. Dalton Trans. 1998 , 369 3 –

3702.

[6] H. Y. Chen, W. Y. Lam, J. D. Luo, Y. L. Ho, B. Z. Tang, D. B. Zhu,

M. Wong, H. S. Kwok, Appl. Phys. Lett. 2002 , 81 , 57 4 –5 76.

[7] J. Chen, Z. Xie, J. W. Y. Lam, C. C. W. Law, B. Z. Tang, Macromo-

lecules 2003 , 36 , 110 8 –1 117.

[8] J. Luo, Z. Xie, J. W. Y. Lam, L. Cheng, H. Chen, C. Qiu, H. S.

Kwok, X. Zhan, Y. Liu, D. Zhu, B. Z. Tang, Chem. Commun. 2001 ,

174 0 –1 741.

[9] J. Chen, C. C. W. Law, J. W. Y. Lam, Y. Dong, S. M. F. Lo, I. D. Wil-

liams, D. Zhu, B. Z. Tang, Chem. Mater. 2003 , 15 , 153 5 –1 546.

[10] S. Yamaguchi, K. Tamao, Bull. Chem. Soc. Jpn. 1996 , 69 , 232 7 –

2334.

[11] S. Yamaguchi, C. Xu, H. Yamada, A. Wakamiya, J. Organomet.

Chem. 2005 , 690 , 536 5 –5 377.

[12] M. Backer, Ph.D. thesis, Humboldt-Universitꢂt, Berlin (Germany),

1999.

[13] J. Mohseni-Ala, N. Auner, Inorg. Chim. Acta 2006, 359, 4677–4697.

[14] D. Yan, M. Bolte, N. Auner, J. Organomet. Chem. 2008 , 693 , 90 8 –

916.

functions (with exponent coefficient a_p/a₀^ꢀ2= 0.8 on hydrogen, a_d/a₀^ꢀ2

=

0.8 on carbon and a_d/a₀^ꢀ2= 0.35 on silicon) was employed (commonly re-

ferred to as def-TZVP) in these calculations. Tight convergence criteria

were enforced for the structure energy-minimizations with the resulting

maximum absolute value of a component of the energy gradient with re-

spect to the nucleiꢀs Cartesian displacements being below 7 mE_h/a₀and

the predicted change in energy for the next cycle in the structure energy-

minimization being always below 30 nE_h. Harmonic vibrational wave-

numbers were calculated with the help of analytic second derivatives.

Vertical electronic singlet excitation energies at the B3LYP/def-TZVP

equilibrium structures were determined within a time-dependent density

functional theory (TDDFT) framework using besides the B3LYP func-

tional with the def-TZVP basis set also the same functional with the

Pople-type triple-zeta valence basis set^[88,89](with the (12,9)/

ACHTNUTRGNE[NUG 6,5] contrac-

tion on silicon) with one diffuse s-function,^[90]one set of diffuse p-func-

tions^[90]and two sets of polarizing d-functions on non-hydrogen atoms as

well as one set of polarizing p-functions on hydrogen.^{[88, 91]}This basis set

is commonly referred to as 6-311gACTHNUTRGNE(UNG 2d,p). Besides the B3LYP functional

also the Perdew–Burke–Ernzerhof hybrid functional^[92](often denoted as

PBE0), which combines the PBE^{[93, 94]}generalized gradient correction to

the local spin-density exchange-correlation with non-local Fock exchange,

was employed with the two basis sets given above. Solvent-induced ab-

sorption frequency shifts were estimated by computing the transition en-

ergies in the framework of the polarizable continuum model (PCM)^[95]

using the gas-phase ground-state equilibrium structures and the non-equi-

librium solvation model (with solvent parameters for EtOH) which ac-

counts for fast electronic responses from the medium (as opposed to slow

reorientation effects which occur essentially after the absorption pro-

cess).^[96]

[15] H. Gilman, W. H. Atwell, J. Am. Chem. Soc. 1964 , 86 , 558 9 –5 593.

[16] H. Gilman, W. H. Atwell, J. Am. Chem. Soc. 1965 , 87 , 267 8 –2 683.

[17] D. Tzeng, R. H. Fong, H. S. D. Soysa, W. P. Weber, J. Organomet.

Chem. 1981 , 219 , 15 3 –1 61.

[18] R. Okazaki, K.-T. Kang, N. Inamoto, Tetrahedron Lett. 1981 , 22 ,

23 5 –2 38.

[19] G. Bertrand, G. Manuel, P. Mazerolles, G. Trinquier, Tetrahedron

1981 , 37 , 287 5 –2 880.

[20] R. T. Conlin, S. Zhang, M. Namavari, K. L. Bobbitt, M. J. Fink, Or-

ganometallics 1989 , 8 , 57 1 –5 73.

[21] C. Kerst, M. Byloos, W. J. Leigh, Can. J. Chem. 1997 , 75 , 97 5 –9 82.

[22] M. G. Steinmetz, B. S. Udayakumar, M. S. Gordon, Organometal-

lics 1989 , 8 , 53 0 –5 41.

[23] P. B. Valkovich, W. P. Weber, Tetrahedron Lett. 1975 , 16 , 215 3 –

2154.

[24] W. Adam, T. Oppenlꢂnder, G. Zang, J. Am. Chem. Soc. 1985 , 107 ,

392 1 –3 924.

[25] W. J. Leigh, K. Zheng, K. B. Clark, Can. J. Chem. 1990 , 68 , 198 8 –

1997.

[26] G. Hohlneicher, M. Mꢁller, M. Demmer, J. Lex, J. H. Penn, L.

Gan, P. D. Loesel, J. Am. Chem. Soc. 1988 , 110 , 448 3 –4 494.

[27] G. Kaupp, M. Stark, Chem. Ber. 1978 , 111 , 360 8 –3 623.

[28] H. Petek, K. Yoshihara, Y. Fujiwara, Z. Lin, J. H. Penn, J. H. Fred-

erick, J. Phys. Chem. 1990 , 94 , 753 9 –7 543.

[29] J. H. Frederick, Y. Fujiwara, J. H. Penn, K. Yoshihara, H. Petek, J.

Phys. Chem. 1991 , 95 , 284 5 –2 858.

[30] K. A. Muszkat, E. Fischer, J. Chem. Soc. B 1967 , 66 2 –6 78.

[31] C. D. DeBoer, R. H. Schlessinger, J. Am. Chem. Soc. 1968 , 90 ,

80 3 –8 04.

Energy-minimized equilibrium structures of the lowest electronically ex-

cited singlet state were obtained with the program package Turbomole.^[97]

The B3LYP functional with the recommended local correlation contribu-

tion (VWN5) was used for this purpose together with the TZVP basis

set. Grids for the numerical integration that is performed in the calcula-

tion of the exchange correlation contribution involved Lebedevꢀs angular

integration grid with 590 grid points and a radial integration grid employ-

ing Chebyshev polynomials of the second kind with the scaling function

M3 of ref. [98]. Tight convergence criteria were requested in the energy-

minimization of the excited-state structure ensuring a maximum absolute

value of a component of the energy gradient with respect to the nucleiꢀs

Cartesian displacements being less than 10 mE_h/a₀and the norm of the

energy change between two steps being below 10 nE_h. Analytic first de-

rivatives have been used in the structure energy-minimization. Harmonic

force fields were calculated at the stationary points using numerical de-

rivatives of analytical first derivatives of the energy with respect to Carte-

sian displacements of the various nuclei. The step sizes chosen in the cen-

tral numerical differentiation setup (two steps per Cartesian displacement

coordinate, one with positive, one with negative step size increment) was

0.02 a₀.

[32] G. Kaupp, M. Stark, Chem. Ber. 1977 , 110 , 308 4 –3 110.

[33] C. Goedicke, H. Stegemeyer, G. Fischer, E. Fischer, Z. Phys.

Chem. (Muenchen Ger.) 1976, 101, 181–196.

[34] T. Wismonski-Knittel, G. Fischer, E. Fischer, J. Chem. Soc. Perkin

Trans. 2 1974 , 193 0 –1 940.

[35] “Photoluminescence of Organically Modified Cyclosiloxanes”: U.

Pernisz, N. Auner in Organosilicon Chemistry IV: From Molecules

to Materials (Eds.: N. Auner, J. Weis), Wiley, New York, 2000,

p. 505.

[36] D. H. Waldeck, Chem. Rev. 1991 , 91 , 41 5 –4 36.

[37] N. Auner, C. Seidenschwarz, E. Herdtweck, Angew. Chem. 1991 ,

103 , 117 2 –1 173; Angew. Chem. Int. Ed. Engl. 1991 , 30 , 115 1 –1 152.

[38] N. Sewald, W. Ziche, A. Wolff, N. Auner, Organometallics 1993 , 12 ,

412 3 –4 134.

Acknowledgements

We thank the Deutsche Forschungsgemeinschaft (DFG) for financial sup-

port. R.B. acknowledges support from the Volkswagen Foundation and

computer time provided by the Centre for Scientific Computing (CSC)

Frankfurt and the Norddeutscher Verbund fꢁr Hoch- und Hçchstleis-

tungsrechnen (HLRN).

[39] N. Auner, C.-R. Heikenwꢂlder, C. Wagner, Organometallics 1993 ,

12 , 413 5 –4 140.

[40] N.Auner, M. Backer, W. Hiller, private communication, CSD

58056, FIZ Karlsruhe, 1997.

[1] C. W. Tang, S. A. Vanslyke, Appl. Phys. Lett. 1987 , 51 , 91 3 –9 15.

[2] C. W. Tang, S. A. Vanslyke, J. Appl. Phys. 1989 , 65 , 361 0 –3 616.

[3] C. Adachi, T. Tsutsui, S. Saito, Appl. Phys. Lett. 1990 , 57 , 53 1 –5 33.

8644

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 8625 – 8645

2,3-Diphenyl-4-neopentyl-1-silacyclobut-2-enes

FULL PAPER

[41] V. Molina, M. Merchꢆn, B. O. Roos, Spectrochim. Acta Part A

1999, 55, 433–446.

[42] M. Klessinger, J. Michl, Excited States and Photochemistry of Or-

ganic Molecules, VCH, Weinheim, 1995.

[43] J. N. Demas, G. A. Crosby, J. Phys. Chem. 1971, 75, 991–1024.

[44] R. M. Dodson, A. G. Zielske, J. Org. Chem. 1967 , 32 , 2 8 –3 1.

[45] R. Improta, F. Santoro, J. Phys. Chem. A 2005 , 109 , 1005 8 –1 0067.

[46] D. Jacquemin, J. Preat, M. Charlot, V. Wathelet, J.-M. Andrꢇ, E. A.

Perpꢈte, J. Chem. Phys. 2004 , 121 , 173 6 –1 743.

[47] D. Jacquemin, E. A. Perpte, G. E. Scuseria, I. Ciofini, C. Adamo, J.

Chem. Theory Comput. 2008 , 4 , 12 3 –1 35.

[48] P. Suppan, J. Photochem. Photobiol. A 1990 , 50 , 29 3 –3 30.

[49] M. Bendikov, B. Solouki, N. Auner, Y. Apeloig, Organometallics

2002 , 21 , 134 9 –1 354.

[50] S. Sharafy, K. A. Muszkat, J. Am. Chem. Soc. 1971 , 93 , 411 9 –4 125.

[51] G. Fischer, G. Seger, K. A. Muszkat, E. Fischer, J. Chem. Soc.

Perkin Trans. 2 1975 , 156 9 –1 576.

[52] M. Kasha, Radiat. Res. 1963 , 20 , 5 5 –7 0.

[53] M. Kasha, H. R. Rawls, M. A. El-Bayoumi, Pure Appl. Chem.

1965 , 11 , 37 1 –3 92.

[54] U. Rçsch, S. Yao, R. Wortmann, F. Wꢁrthner, Angew. Chem. 2006 ,

118 , 718 4 –7 188; Angew. Chem. Int. Ed. 2006 , 45 , 702 6 –7 030.

[55] W. Y. Huang, H. Yun, H. S. Lin, T. K. Kwei, Y. Okamoto, Macro-

molecules 1999 , 32 , 808 9 –8 093.

[56] J.-H. Hsu, W. Fann, P.-H. Tsao, K.-R. Chuang, S.-A. Chen, J. Phys.

Chem. A 1999 , 103 , 237 5 –2 380.

[57] F. Boroumand, H. van der Bergh, J. E. Moser, Anal. Chem. 1994 ,

66 , 226 0 –2 266.

[58] P. Kubelka, J. Opt. Soc. Am. 1948 , 38 , 44 8 –4 57.

[59] R. Allison, J. Burns, A. J. Tuzzolino, J. Opt. Soc. Am. 1964 , 54 ,

74 7 –7 50.

[60] N. Kristianpoller, J. Opt. Soc. Am. 1964 , 54 , 128 5 –1 286.

[61] M. S. Wrighton, D. S. Ginley, D. L. Morse, J. Phys. Chem. 1974 , 78 ,

222 9 –2 233.

[62] Y. Ooyama, S. Yoshikawa, S. Watanabe, K. Yoshida, Org. Biomol.

Chem. 2007 , 5 , 126 0 –1 269.

[63] H. Langhals, T. Potrawa, H. Hçth, G. Linti, Angew. Chem. 1989 ,

101 , 49 7 –4 99; Angew. Chem. Int. Ed. Engl. 1989 , 28 , 47 8 –4 80.

[64] J. H. Penn, L.-X. Gan, T. A. Eaton, E. Y. Chan, Z. Linti, J. Org.

Chem. 1988 , 53 , 151 9 –1 523.

[65] E. H. White, J. P. Anhalt, Tetrahedron Lett. 1965 , 6 , 393 7 –3 945.

[66] M. A. Battiste, M. E. Burns, Tetrahedron Lett. 1966 , 7 , 52 3 –5 30.

[67] W. J. Leigh, J. A. Postigo, J. Am. Chem. Soc. 1995 , 117 , 168 8 –1 694.

[68] F. Dçrr in Optische Anregung organischer Systeme (Ed.: W. Fçrst),

VCH, Weinheim, 1966, pp. 566–585.

[69] S. Jursenas, N. Kurilcik, R. Karpicz, V. Gulbinas, L. Valkunas, M.

Rutkis, I. Muzikante, Thin Solid Films 2008 , 516 , 890 9 –8 916.

[70] P. Avis, G. Porter, J. Chem. Soc. Faraday Trans. 2 1974 , 70 , 105 7 –

1065.

[71] G. M. Sheldrick, Acta Crystallogr. Sect. A 1990 , 46 , 46 7 –4 73.

[72] SABABS, Program for Absorption Correction of Area Detector

Data, G. M. Sheldrick, University of Gçttingen, Germany, 1996.

[73] SHELXL-97, Program for the Refinement of Crystal Structures,

G. M. Sheldrick, University of Gçttingen, Germany, 1997.

[74] A. M. Kjaer, J. Ulstrup, J. Am. Chem. Soc. 1987 , 109 , 193 4 –1 942.

[75] C. H. Choi, M. Kertesz, J. Phys. Chem. A 1997 , 101 , 382 3 –3 831.

[76] A. B. Myers, R. A. Mathies, J. Chem. Phys. 1984 , 81 , 155 2 –1 558.

[77] Handbook of Fluorescence Spectra of Aromatic Molecules (Ed.:

I. B. Berlman), Academic Press, New York, 1965.

[78] Gaussian 03, Revision D.01, M. J. Frisch, G. W. Trucks, H. B. Schle-

gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgo-

mery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.

Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,

N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.

Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.

Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.

Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gom-

perts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Po-

melli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P.

Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.

Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.

Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S.

Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Pis-

korz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-

Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W.

Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A. Pople,

Gaussian, Inc., Wallingford CT, 2004.

[79] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys.

Chem. 1994 , 98 , 1162 3 –1 1627.

[80] A. D. Becke, J. Chem. Phys. 1993 , 98 , 564 8 –5 652.

[81] P. A. M. Dirac, Proc. R. Soc. London Ser. A 1929, 123, 714–733.

[82] J. C. Slater, Phys. Rev. 1951 , 81 , 38 5 –3 90.

[83] A. D. Becke, Phys. Rev. A 1988 , 38 , 309 8 –3 100.

[84] V. Fock, Z. Phys. 1930, 61, 126–148.

[85] S. H. Vosko, L. Wilk, M. Nuisar, Can. J. Phys. 1980, 58, 1200–1211.

[86] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988 , 37 , 78 5 –7 89.

[87] A. Schꢂfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1993, 100, 5829–

5835.

[88] R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys.

1980 , 72 , 65 0 –6 54.

[89] A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980 , 72 , 563 9 –5 648.

[90] T. Clark, J. Chandrasekhar, G. W. Spitznagel, P. von R. Schleyer, J.

Comput. Chem. 1983 , 4 , 29 4 –3 01.

[91] M. J. Frisch, J. A. Pople, J. S. Binkley, J. Chem. Phys. 1984 , 80 ,

326 5 –3 269.

[92] J. P. Perdew, M. Ernzerhof, K. Burke, J. Chem. Phys. 1996 , 105 ,

998 2 –9 985.

[93] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996 , 77 ,

386 5 –3 868.

[94] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1997 , 78 ,

139 6 –1 396.

[95] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005 , 105 , 299 9 –

3093.

[96] M. Cossi, V. Barone, J. Chem. Phys. 2001 , 115 , 470 8 –4 717.

[97] Turbomole 5.10, R. Ahlrichs, M. Bꢂr, H. Baron, R. Bauernschmitt,

S. Bçcker, N. Crawford, P. Deglmann, M. Ehrig, S. E. K. Eichkorn,

F. Furche, F. Haase, M. Hꢂser, C. Hꢂttig, A. Hellweg, H. Horn, C.

Huber, U. Huniar, M. Kattannek, A. Kçhn, C. Kçlmel, M. Koll-

witz, K. May, P. Nava, C. Ochsenfeld, H. ꢉhm, H. Patzelt, D. Rap-

poport, O. Rubner, A. Schꢂfer, U. Schneider, M. Sierka, O. Treu-

tler, B. Unterreiner, M. Arnim, F. Weigend, P. Weis, H. Weiss,

2008.

[98] O. Treutler, R. Ahlrichs, J. Chem. Phys. 1995 , 102 , 34 6 –3 54.

[99] G. Schaftenaar, J. H. Noordik, J. Comput.-Aided Mol. De. 2000 , 14 ,

12 3 –1 34.

[100] J. Bernstein, Acta Crystallogr. Sect. B 1975 , 31 , 41 8 –4 22.

Received: May 18, 2009

Published online: August 13, 2009

Chem. Eur. J. 2009, 15, 8625 – 8645

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

8645

Article Doi

Synthesis, structure, photoluminescence and photoreactivity of 2,3-diphenyl-4-neopentyl-l-silacyclobut-2-enes

DOI: 10.1002/chem.200901323

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.200901323

Full text of DOI:10.1002/chem.200901323

Products guided by the article

R&D Labs maybe for 277331-71-2

Relevant to this article

Hot Product