C-functionalized, air- and water-stable 9,10-dihydro-9,10- diboraanthracenes: Efficient blue to red emitting luminophores

Article

pubs.acs.org/JACS

C‑Functionalized, Air- and Water-Stable 9,10-Dihydro-9,10-

diboraanthracenes: Eﬃcient Blue to Red Emitting Luminophores

Christian Reus, Sabine Weidlich, Michael Bolte, Hans-Wolfram Lerner, and Matthias Wagner*

Institut fur Anorganische und Analytische Chemie, J. W. Goethe-Universitat Frankfurt, Max-von-Laue-Straße 7,

̈

D-60438 Frankfurt (Main), Germany

S

* Supporting Information

ABSTRACT: 9,10-Dihydro-9,10-diboraanthracene (DBA) pro-

vides a versatile scaﬀold for the development of boron-doped

organic luminophores. Symmetrically C-halogenated DBAs are

obtained through the condensation of 4-bromo-1,2-bis-

(trimethylsilyl)benzene or 4,5-dichloro-1,2-bis(trimethylsilyl)-

benzene with BBr₃in hexane. Unsymmetrically C-halogenated

DBAs are formed via an electrophilic solvent activation reac-

tion if the synthesis is carried out in o-xylene. Mechanistic

insight has been achieved by in situ NMR spectroscopy, which

revealed C-halogenated 1,2-bis(dibromoboryl)benzenes to be

the key intermediates. Treatment of the primary 9,10-dibromo-DBAs with MesMgBr yields air- and water-stable C-halogenated

9,10-dimesityl-DBAs (2-Br-6,7-Me₂-DBA(Mes)₂; 2,6-Br₂-DBA(Mes)₂; 2,3-Cl₂-6,7-Me₂-DBA(Mes)₂; 2,3,6,7-Cl₄-DBA(Mes)₂).

Subsequent Stille-type C−C-coupling reactions give access to corresponding phenyl, 2-thienyl, and p-N,N-diphenylaminophenyl

derivatives, which act as highly emissive donor−acceptor dyads or donor−acceptor−donor triads both in solution and in the

solid state. 2-Thienyl was chosen as a model substituent to show that already a variation of the number and/or the positional

distribution of the donor groups suﬃces to tune the emission wavelength of the resulting benchtop stable compounds from

469 nm (blue) to 540 nm (green). A further shift of the ﬂuorescence maximum to 594 nm (red) can be achieved by switching

from 2-thienyl to p-aminophenyl groups. A comparison of the optoelectronic properties of selected C-substituted DBA(Mes)₂

derivatives with those of the isostructural anthracene analogues unveiled the following: (i) The DBA core is a much better

electron acceptor. (ii) The emission colors of DBAs fall in the visible range of the spectrum (blue to orange), while anthracenes

emit exclusively in the near-ultraviolet to blue wavelength regime. (iii) DBAs show signiﬁcantly higher solid-state quantum yields.

During the past ﬁve years, 9,10-dihydro-9,10-diboraanthra-

cene (DBA) has been established as a promising building block

for organic π-conjugated materials.¹⁴⁻²⁵Because both 9,10-

dibromo-9,10-dihydro-9,10-diboraanthracene (DBA(Br)₂)¹⁹

INTRODUCTION

■

Carbon-based π-conjugated systems nowadays represent an

important class of materials with applications in a wide range of

contemporary technological areas¹such as organic electronics,^2,3

light-emitting devices,⁴or photovoltaics.⁵Low-cost organic

materials exhibit advantageous properties such as being light-

weight, high mechanical ﬂexibility, and solution processability,

which are not found in their silicon or metal counterparts.

However, purely organic compounds with small band gaps and/

or low-lying LUMO energies are rare and often suﬀer from distinct

disadvantages (cf., pentacene is O₂-sensitive⁶and prone to

dimerization via Diels−Alder reactions⁷).⁸The replacement of

selected carbon atoms within the organic π system by main-group

heteroatoms like B,⁹Si,¹⁰N,¹¹P,¹²or S¹³has been used as a tool

to circumvent these shortcomings and, moreover, ever since has

emerged as a powerful method to open exciting new possibilities.

Especially the implementation of Lewis acidic boron atoms has

proven to result in very desirable optoelectronic properties such as

high electron aﬃnities (leading to n-type materials) and intense

luminescence.⁹These eﬀects are rooted in the interaction of the

vacant p orbital at the boron center with the surrounding

π-electron clouds, which is particularly eﬃcient in the molecules’

LUMOs and thereby lowers the energies of these orbitals.

14

and the parent compound DBA(H)₂are readily available, a

variety of 9,10-diorganyl-9,10-dihydro-9,10-diboraanthracene deriva-

tives (DBA(R)₂, A; Figure 1) can be conveniently prepared

through nucleophilic substitution or hydroboration reactions.

Thus, in this approach, the unique chemical nature of the

boron atom is exploited in two ways: Boron ﬁrst serves as an

expedient anchor group to facilitate the introduction of sub-

stituents into the molecule and afterward governs the electronic

structure of the overall π-electron system. Using the 9,10-

positions as the sole sites for derivatization, it has so far been

possible not only to tune the emission characteristics of DBAs

over a wide range, but also to prepare polymeric species^14,15

along with low-molecular weight compounds.

Despite these numerous positive aspects of A-type materials,

two major drawbacks have to be overcome before applications

of DBA ﬂuorophores can be taken into serious consideration.

Received: July 3, 2013

A

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

air- and water-stable donor−acceptor−donor triads and donor−

acceptor dyads. Appropriate choice of substituents allows a

tuning of the emission colors from the blue to the red region of

the visible-light spectrum. In the ﬁnal part of this Article, we

will directly compare selected DBA-containing species with

their all-carbon anthracene analogues and show that boron

doping signiﬁcantly improves the materials’ optoelectronic

properties.

RESULTS AND DISCUSSION

■

Synthesis of Halogenated 9,10-Dimesityl-9,10-dihy-

dro-9,10-diboraanthracenes. A convenient synthesis route

to parent 9,10-dimesityl-9,10-dihydro-9,10-diboraanthracene

(DBA(Mes)₂, R = H; 2) starts from 1,2-bis(trimethylsilyl)-

benzene (R = H; 1)³¹and BBr₃, and is conducted in toluene solu-

tion at 120 °C in a sealed glass ampule (Scheme 1).¹⁹Subsequent

Figure 1. The B- and C-functionalized 9,10-dihydro-9,10-diboraan-

thracenes.

On the one hand, the quantum yields obtained so far rarely

exceed 30%,²³and, even more importantly, most DBA deriva-

tives developed to date proved to be highly sensitive toward air

and moisture. Usually, the weak links are the exocyclic B−C

bonds, whereas the actual DBA framework tends to be rather

stable (we commonly ﬁnd the borinic acid DBA(OH)₂¹⁹as the

product of hydrolytic degradation). The design of more inert

DBAs has therefore to focus on the steric shielding of the

vacant boron p orbital and on the kinetic protection of the

exocyclic B−C bond.

Scheme 1. Synthesis of Parent DBA(Mes)₂(2; R = H)

Along these lines, 9,10-dimesityl-9,10-dihydro-9,10-diboraan-

thracene (DBA(Mes)₂) is a rare example of a long-term

benchtop stable DBA derivative,^18,25which we therefore selected

as the lead structure for further materials development.

DBA(Mes)₂undergoes two reversible one-electron transitions

at comparatively anodic redox potentials of E_1/2= −1.84 V,

−2.73 V (THF; vs FcH/FcH⁺) and is thus electrochemically

well behaved.²⁵In C₆H₆, the compound ﬂuoresces at 460 nm,

yet with a low quantum eﬃciency of only 3%.²⁵In contrast to

A-type materials, the boron atoms of DBA(Mes)₂are no longer

available for further functionalization. As a consequence, any

substituents have now to be introduced into the o-phenylene

rings to create B-type materials (C-substitution; Figure 1). The

successful application of such an approach to the synthesis of

compounds with remarkable and useful optoelectronic proper-

ties has already been demonstrated for phenylene-function-

alized dibenzoboroles²⁶and dibenzo[b,f ]borepins.²⁷In the case

of 9,10-dihydro-9,10-diboraanthracenes, however, correspond-

ing examples are restricted to three frameworks: perﬂuorinated

DBAs,²⁸6,13-dihydro-6,13-diborapentacenes,^29,30and a DBA-

containing graphene ﬂake.²⁴The scarcity of C-substituted DBAs

has to be attributed to the lack of appropriately substituted 1,2-

bis(trimethylsilyl)benzenes, which generally serve as the key

starting materials in DBA syntheses.

Very recently, our group reported viable routes to various

halogenated 1,2-bis(trimethylsilyl)benzenes,^31,32which now

puts us into the position to prepare C-halogenated DBAs. In

this context, we have elucidated key intermediates and mecha-

nistic details of the reaction sequence leading to the assembly of

the DBA(Br)₂framework. These investigations also resulted in

the discovery of a unique C−H activation reaction of aromatic

solvents, which gives access to unsymmetrically C-substituted

DBAs in preparatively useful yields. C-Halogenated DBA(Mes)₂

derivatives provide a universal platform for further transformations,

as the primary halogeno substituents can be replaced at will by

organyl groups through Pd-mediated C−C-coupling reactions.

Herein, we will prove the viability of this concept for the

preparation of highly emissive (quantum yields up to 72%),

treatment of the intermediate DBA(Br)₂with MesMgBr provides 2

in an overall yield of 40%.²⁵Given the comparatively high reac-

tion temperature, the use of a very Lewis acidic boron halide

and a strongly nucleophilic Grignard reagent, the selection of

substituents R that are compatible with these conditions is

rather limited. Halogen atoms are among the most promising

candidates R, because they allow the introduction of a broad

variety of organic substituents through Pd-catalyzed C−C-

coupling reactions. The most readily available halogenated 1,2-

bis(trimethylsilyl)benzenes are the compounds 4-ﬂuoro- and 4-

chloro-1,2-bis(trimethylsilyl)benzene,³¹as well as 4-bromo-

(3)³²and 3,4-dichloro-1,2-bis(trimethylsilyl)benzene (4),³¹

which are shown in Scheme 2. For our purposes, 3 appeared

to be best suited, because C−Br bonds are usually easier to

activate than C−F/Cl bonds. Because the 4,5-dibrominated 1,2-

bis(trimethylsilyl)benzene is unknown, we included the corres-

ponding 4,5-dichloro-1,2-bis(trimethylsilyl)benzene (4; Scheme 2)

in our investigation.

Our ﬁrst syntheses of halogenated DBA(Mes)₂species

employing 3 and 4 relied on the protocol developed for the

preparation of pristine 2 (Scheme 1). After workup, we noted

two unexpected results: (i) The ratio between the two isomeric

products 2,6-Br₂-DBA(Mes)₂(5) and 2,7-Br₂-DBA(Mes)₂(6)

was not 1:1 but 3:2. (ii) With both starting materials 3 and 4,

the target compounds 5/6 and 10 were the minor constituents

in the product mixture; as major constituents, we identiﬁed the

unsymmetrically substituted DBAs 7/8 and 12, which had

obviously formed through solvent (toluene) activation (Scheme 2).

Given this background, we performed a systematic screening

of reaction temperature proﬁles and solvents and come to the

conclusion that the optimum conditions are dependent on the

desired product speciﬁcations:

(1) If the focus lies on a maximum overall yield of 5/6 or 10,

the DBA scaﬀold should be assembled under the following

conditions: hexane, 120 °C, sealed glass ampule, 1.5 d (Scheme 2).

As a drawback, this procedure does not discriminate at all

between 5 and 6, but leads to a 1:1 mixture.

B

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

a

Scheme 2. Synthesis of the Halogenated DBA(Mes)₂Derivatives 5−13

a

The table shows the overall yield over the two-step reaction and the proportion of the solvent-activated products 7−9 and 11−13 (n = 0−2),

respectively. (i) (1) +BBr₃(3 equiv), 120 °C, 36 h, solvent as indicated in the table; (2) removal of all volatiles, +MesMgBr (2 equiv in THF),

toluene, 0 °C, 12 h. (a) A 1:1 mixture hexane:o-xylene was used.

(2) If the discrimination between 5 and 6 is the main issue,

the reaction should be carried out in C₆H₆instead of hexane

(Scheme 2).

Scheme 3. Mechanistic Investigation into the Reaction of 3

with BBr₃and Proposed Reaction Mechanism for the

Assembly of the DBA Scaﬀold via the Main Intermediate 14

a

(3) If the amount of unsymmetrically substituted DBAs is to

be maximized, a 1:1 mixture of hexane with the electron-rich

o-xylene is a particularly reactive solvent, which, contrary to

toluene, generates exclusively one monobromo derivative 9

(Scheme 2).

In all cases, best yields were obtained when the air- and

moisture-sensitive halogenated DBA(Br)₂intermediates were

directly converted to the air- and water-stable mesitylated target

products without prior puriﬁcation. We also note that the yield

of the parent compound DBA(Mes)₂(2; R = H) improves

considerably if hexane is used instead of toluene¹⁹as solvent in

the ﬁrst step of the synthesis sequence.

With the aim to learn more about the factors governing the

formation of the isomer 2,6-Br₂-DBA(Mes)₂(5) versus 2,7-Br₂-

DBA(Mes)₂(6), we monitored the reaction between 3 and

BBr₃by NMR spectroscopy. Three key factors were varied: (i)

the amount of BBr₃(substoichiometric 0.35 equiv and the

regular 3 equiv), (ii) the solvent (perdeuterated cyclohexane

(C₆D₁₂) and C₆D₆), and (iii) the reaction temperature (room

temperature and 120 °C).

The room-temperature data are discussed ﬁrst. In C₆D₁₂, 3

and BBr₃showed no reaction within 1 d, irrespective of the

amount of BBr₃employed.³³A signiﬁcantly diﬀerent result was

obtained for the C₆D₆samples: Addition of 0.35 equiv of BBr₃

to 3 essentially gave a mixture of the starting materials and

4-bromo-1-dibromoboryl-2-trimethylsilylbenzene 14 after 1 d

(Scheme 3; cf., the Supporting Information for more details).³³

The qualitative picture did not change when the amount of

BBr₃was raised to 3 equiv, but the ratio 3:14 inverted from 6:1

to 1:6 (1 d); the conversion to 14 was quantitative after 2 d.

Most importantly, we found no indication for the formation of

the other possible isomer 4-bromo-2-dibromoboryl-1-trime-

thylsilylbenzene. Thus, at room temperature, the Si/B exchange

proceeds selectively para to the bromine atom, as to be

expected from the strongly ortho/para directing eﬀect of a

bromo substituent in electrophilic aromatic substitution

a

Bottom: Br/Me rearrangement reaction of 14 leading to the side

products 18 and 19. (i) +BBr₃(0.35 or 3 equiv); C₆D₆, room

temperature (quantitative after 2 d) or C₆D₁₂, 120 °C.

reactions.³⁴Moreover, the introduction of a ﬁrst BBr₂group

apparently deactivates the aromatic system to such an extent

that a second Si/B exchange does not take place under ambient

conditions.³⁵This explains why neither 4-bromo-1,2-bis-

(dibromoboryl)benzene 16 nor dibrominated DBA(Br)₂(15/17)

could be observed in the room-temperature experiments.

The results described in the following for experiments carried

out at 120 °C equally apply for the C₆D₁₂and the C₆D₆

samples. The reaction of 3 with 0.35 equiv of BBr₃completely

consumed the BBr₃during 1 d and yielded 14 together with

C

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

dibrominated DBA(Br)₂(15/17). Further heating of the

sample mainly led to a rearrangement of 14 into 18.³⁶When

using 3 equiv of BBr₃, 15/17 started to precipitate after 2 h, and

a solution containing 14, 18, and 4-bromo-1-dibromoboryl-2-

bromodimethylsilylbenzene (19) was obtained.

(relative to 18) by accelerating the desired one of the two

competing reaction pathways.

With respect to the solvent-activation reaction, the following

trends are worth mentioning: (i) The nonhalogenated bis-

(trimethylsilyl)benzene 1 neither reacts with C₆H₆, nor with

toluene, nor with o-xylene; (ii) the monobrominated derivative

3 is inert toward C₆H₆but activates toluene and, to an even

greater extent, o-xylene; and (iii) the dichlorinated derivative 4

reacts even with C₆H₆but again gives best results with o-xylene

(cf., 10:13 = 1:7; Scheme 2). Highest yields of an unsym-

metrically substituted DBA(Mes)₂are thus obtained when an

electron-poor 1,2-bis(trimethylsilyl)benzene is combined with

an electron-rich aromatic solvent. It is worth mentioning that

the solvent activation mechanism necessarily involves the libera-

tion of HBr, which, in turn, can induce unwanted proto-

desilylation processes (cf., ref 33) and thereby lower the

product yields.

After 1 d, only 18 and 19 had remained in the solution; more

prolonged reaction times led to a gradual decrease of the rela-

tive amount of 18 and to a concomitant increase in the

concentration of 19. 19 is thus likely generated through the

reaction of 18 with BBr₃(signals of the byproduct MeBBr₂also

appeared in the NMR spectra).

1

After mesitylation of the obtained 15/17, H NMR spectro-

scopy indicated isomer distributions of 5:6 = 1:1 (C₆D₁₂) and

3:2 (C₆D₆). The only experimentally established B/Si-sub-

stituted bromobenzene was 14; resonances of its isomer

4-bromo-1-dibromoboryl-2-bromodimethylsilylbenzene never

occurred in the NMR spectra. If 15/17 were solely formed

through the homodimerization of 14, one would consequently

expect perfect selectivity for the isomer 2,6-Br₂-DBA(Br)₂(15)

in the present case. The total lack of selectivity in C₆D₁₂and the

moderate selectivity in C₆D₆solution therefore point toward a

B/Si-exchange on 14 at elevated temperatures furnishing the

1,2-diborylated benzene 16. The more symmetrical compound

16 will react with itself or 14 in a less discriminative fashion

regarding the formation of either 15 or 17. Indeed, during our

high-temperature NMR-monitoring studies with excess BBr₃,

we came across one set of signals not mentioned so far, which

we tentatively assign to 16 (Scheme 3; cf., the Supporting

Information for more details).

Our current working hypothesis is that the solvent activation

step proceeds through electrophilic aromatic borylation, which

requires electron-deﬁcient boron reagents. 1,2-Diborylbenzenes

are already strong Lewis acids, and the introduction of halogen

atoms into the phenylene ring should further enhance their

Lewis acidity (cf., 16; Scheme 3). However, the best established

aromatic borylation mechanisms involve boron cations as key

intermediates.³⁷We therefore assume that the solvent

activation process has to be promoted by bromide abstraction

from the reacting boron center, which could be achieved either

38

by attack of external BBr₃or by the formation of an intra-

molecular B−Br···B acid−base adduct (Figure 2, left). A related

To test this working hypothesis, we ﬁrst heated a solution of

pure 14 in C₆D₆to 120 °C for extended periods of time. As a

result, extensive Br/Me rearrangement (14→18) but no

formation of 15/17 took place. The experiment was next

repeated in the presence of a catalytic amount of BBr₃

(approximately 5 mol %). This time, the reaction gave 18 in

addition to 15/17; however, the DBA assembly was slowed,

and the yield of 15/17 was markedly lower than usual.

Importantly, the isomer ratio determined after mesitylation of

15/17 was again 3:2, thereby conﬁrming the adverse eﬀect of

the likely key intermediate 16 on the isomer selectivity 15

versus 17 (Scheme 3).

Figure 2. Left: Intramolecular B−Br···B adduct as the proposed

intermediate of the electrophilic aromatic solvent borylation reaction

(X = Br, n = 1; X = Cl, n = 2). Right: Structurally characterized

ferrocene-based bidentate B/Sn Lewis acid forming an intramolecular

B−Cl···Sn adduct.³⁹

Our mechanistic investigations not only shed light on the

factors governing the isomer distribution 15/17, but also clarify

why the combined yields of 5/6 never exceeded 60%. After the

usual heating time of 1−1.5 d, the product mixture resulting

from 3 and BBr₃contained 15/17, 18, and 19 irrespective of

the solvent employed. Longer heating times did not eﬀect a

decrease in the concentration of 19. Moreover, we detected

only negligible quantities of Me₂SiBr₂as compared to the

amount of Me₃SiBr present in the solution. Contrary to the

SiMe₃group, the Si(Br)Me₂substituent apparently does not

take part in further B/Si exchange processes, which is why 18

and 19 represent dead ends of the 15/17 synthesis sequence. In

agreement with the fact that the Si−Ph bond cleavage by BBr₃

is less facile in C₆D₁₂than in C₆D₆, we also monitored a much

lower relative content of 19 in heated C₆D₁₂than in heated

C₆D₆, which explains the generally better yields of 9,10-

dihydro-9,10-diboraanthracene syntheses in hexane as com-

pared to aromatic solvents. Moreover, the reaction 14→15/17

requires BBr₃as a catalyst (to generate 16 as key intermediate),

while the rearrangement 14→18 is essentially independent of

the BBr₃concentration. The use of excess BBr₃(i.e., 3 equiv per

molecule 3) consequently improves the yields of 15/17

intramolecular B−Cl···Sn adduct has been characterized by

X-ray crystallography for ferrocene-based bidentate B/Sn Lewis

acids (Figure 2, right).³⁹

All compounds 5−13 compiled in Scheme 2 have been

synthesized on a preparative scale and employed in further

reactions (see below). A comparison of the key optoelectronic

parameters of the toluene-derived DBA(Mes)₂species with

those of the o-xylene-derived ones revealed no signiﬁcant

diﬀerences. From now on, toluene-derived compounds will

therefore be excluded from detailed discussions.

The dibrominated DBA(Mes)₂was obtained as an analyti-

cally pure pair of isomers 5/6 in yields of 43% (hexane; 5:6 =

1:1) or 31% (C₆H₆; 5:6 = 3:2) based on 3. The two isomers

could not be separated from each other, even by HPLC.

However, after substitution of bromine atoms for, e.g., 2-thienyl

groups, chromatographic separation was possible (cf., com-

pounds 23/24 below). The tetrachlorinated DBA(Mes)₂10 is

readily accessible in 60% overall yield. Even though all of the

solvent-activated compounds are inevitably contaminated with

the corresponding symmetrically substituted DBA(Mes)₂

D

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

species, an almost quantitative enrichment can be achieved by

(repeated) HPLC. It is yet more convenient to use the mixtures

for subsequent C−C-coupling reactions and to perform the

separation step afterward by conventional column chromato-

graphy.

Similar to pristine 2, the halogeno derivatives are air- and

moisture-stable both in solution and in the solid state for

extended periods of time. 5−13 are soluble in aromatic

solvents, but they dissolve better in chlorinated hydrocarbons;

NMR data were therefore recorded in CDCl₃(cf., the

Supporting Information for plots of all spectra).

The ¹¹B{¹H} NMR spectra of all DBA(Mes)₂derivatives

5−13 are characterized by very broad resonances in the range

69−74 ppm (cf., 2: δ(¹¹B) = 66¹⁸). 2,6-Br₂-DBA(Mes)₂(5) can

easily be distinguished from 2,7-Br₂-DBA(Mes)₂(6) in the

¹H NMR spectrum, because it gives rise to only one set of

mesityl resonances, whereas two sets appear in the spectrum of

6. Both in 5 and in 6, the o-phenylene protons give rise to one

small doublet (⁴J_HHcoupling), one large doublet (³J_HH

coupling), and one doublet of doublets; corresponding signals

possess only marginally diﬀerent chemical shift values in both

isomers. Because of its high symmetry, Cl₄-DBA(Mes)₂(10)

exhibits only one singlet proton resonance for the DBA

backbone. In the cases of the unsymmetrically substituted

DBA(Mes)₂derivatives, the general NMR features of the

halogenated o-phenylene rings remain essentially the same as in

5/6 or 10. The o-phenylene fragment in 11 gives rise to two

Figure 3. Molecular structure of 5 (left) and 10 (right) in the solid

state. Displacement ellipsoids are drawn at the 50% probability level;

H atoms have been omitted for clarity. The bromine atoms in 5 are

disordered over two positions with an occupancy ratio of

0.858(3):0.142(3); only the major orientation is shown. Selected

bond lengths (Å), bond angles (deg), and dihedral angle (deg): 5,

B(1)−C(1) 1.550(9), B(1)−C(2A) 1.572(10), B(1)−C(11) 1.576(9),

C(1)−C(2) 1.435(8), C(4)−Br(1) 1.878(7); C(1)−B(1)−C(2A)

118.7(5), C(1)−B(1)−C(11) 120.9(6), C(2A)−B(1)−C(11)

120.4(6); Ar(B(1))//Ar(C(11)) 82.2(2). 10, B(1)−C(1) 1.571(5),

B(1)−C(2A) 1.558(5), B(1)−C(11) 1.571(5), C(1)−C(2) 1.417(5),

C(4)−Cl(1) 1.716(3), C(5)−Cl(2) 1.712(3); C(1)−B(1)−C(2A)

118.6(3), C(1)−B(1)−C(11) 119.4(3), C(2A)−B(1)−C(11)

121.9(3); Ar(B(1))//Ar(C(11)) 87.1(1). Symmetry transformation

used to generate equivalent atoms: 5, A, −x + 3/2, −y + 1/2, −z +

3/2; 10, A, −x + 1, −y + 1, −z + 2.

1

multiplets in the H NMR spectrum, reminiscent of the parent

compound 2. The o-xylene-derived species 13 shows one CH

and one CCH₃resonance, while symmetry reduction in 9 leads

to two aromatic CH signals (the two CCH₃resonances are

overlapping).

into the Br−C σ* orbital competes with the reduction of the

low-lying DBA-centered π* orbital at the ﬁrst stages of the

reactions. In line with this view, parent 2 is easily reducible with

elemental Li to the red-colored dilithium salt Li₂[2].^17,25

We identiﬁed the best-suited Pd-catalyzed C−C-coupling

strategy for the further modiﬁcation of 5−13 by screening three

diﬀerent methods to introduce phenyl rings into the molecular

scaﬀolds. Suzuki-type protocols led to extensive decomposition,

because under the basic conditions required to cleave the

(HO)₂B−Ph bond, B−C bonds within the DBA(Mes)₂

frameworks are obviously activated as well.⁴⁰Negishi-type

protocols failed to give any transformation, while the Stille-

coupling reaction was immediately successful.

5/6 and 10 were converted in toluene with PhSnBu₃into

20/21 and 22 using 2.5 mol % [Pd(PtBu₃)₂] per halogen atom

as the catalyst (Scheme 4). The reactions can be conducted

with comparable results either in a closed glass vessel with

Young valve (120 °C, 1 d) or in a microwave synthesizer (170 °C,

45 min). The fully substituted target compounds were easily

separated from partially substituted and protodehalogenated

side products by column chromatography (silica gel, hexane/

CHCl₃) and isolated in 40% yield each.

The molecular structures of 5 and 10 were determined by

X-ray crystallography (Figure 3; cf., the Supporting Information

for the X-ray crystal structure analysis of Cl₄-DBA(Br)₂).

In both centrosymmetric compounds, the planes of the mesityl

substituents are almost perpendicular to the DBA plane with

dihedral angles of 82.2(2)° (5) and 87.1(1)° (10). As a result,

the o-methyl groups are eﬃciently shielding the vacant boron

p-orbitals from above and below, which explains the chemical

inertness of the B−C bonds. All key metrical parameters of the

DBA(Mes)₂fragments in 5 and 10 are essentially the same as in

parent 2¹⁸and therefore do not merit further discussion. In 2,6-

Br₂-DBA(Mes)₂(5), the charge density at Br(1) is 14% less

than expected for one bromo substituent; the missing charge

density is located close to H(5). Thus, 5 either cocrystallizes

together with a small fraction of 2,7-Br₂-DBA(Mes)₂(6) and/

or 5 is disordered over two positions about a 2-fold axis

running through the boron atoms. The second explanation is

supported by the fact that a similar phenomenon has been

observed in the crystal structure analysis of the isomerically

pure compound 2,6-dibromo-9,10-dimesitylanthracene (cf., the

Supporting Information for more details).

Derivatization of Halogenated 9,10-Dimesityl-9,10-dihy-

dro-9,10-diboraanthracenes through Stille Coupling.

Attempts at a metal/bromine exchange on 5/6 with elemental

Mg or n-BuLi in THF were not met with success. An initial

color change from yellow to deep red did not persist, and the

ﬁnal color of the reaction mixtures was a dark brown. When

Me₃SiCl was added as an electrophilic quencher at various

points in time, complex mixtures containing the respective

starting material, free mesitylene, and numerous unidentiﬁed

other species were obtained. We assume that electron injection

The quantitative Br/Ph exchange in 20/21 and 22 is

1

unequivocally evident from the H and ¹³C{¹H} NMR spectra

of the compounds. Not only does the number of resonances

correspond to the proposed symmetry of the molecules, but an

inspection of the proton integral ratios conﬁrms the presence of

2 and 4 phenyl rings in 20/21 and 22, respectively. The

molecular structure of the centrosymmetric isomer 2,6-Ph₂-

DBA(Mes)₂(20) was further corroborated by X-ray crystallog-

raphy (cf., the Supporting Information for full details). In

contrast to 2,6-Br₂-DBA(Mes)₂(5), 20 is not disordered in the

E

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 4. Stille-Type Pd-Catalyzed C−C-Coupling

a

Reactions of 5/6 or 10 with PhSnBu₃

a

MW = microwave irradiation.

crystal lattice; the dihedral angle between one phenyl ring and

the DBA core amounts to 33.9(1)°.

As alluded to above, the DBA(Mes)₂moiety possesses an

energetically low-lying LUMO, essentially identical to the π*

orbital of the DBA core.²⁵We therefore reasoned that pro-

nounced eﬀects on the optoelectronic properties of the DBA-

(Mes)₂fragment will arise if it is equipped with electron-donating

substituents such as 2-thienyl and p-N,N-diphenylaminophenyl.⁴¹

Starting from corresponding halogenated symmetrical or

unsymmetrical DBA(Mes)₂derivatives shown in Scheme 2

and ThSnBu₃or Ph₂NphSnBu₃, we have prepared compounds

23−31 (Th = 2-thienyl, ph = p-phenylene; Figure 4) through

the Stille-type protocol outlined in Scheme 4. The complete

series of possible substitution patterns, which will provide an

opportunity to study detailed structure−property relationships,

was realized for the 2-thienyl derivatives. Isolated yields were

generally excellent, even with the tetrachlorinated starting

material 10. The solubilities of the compounds 23−28 in common

organic solvents are considerably improved as compared to

those of the starting materials, while the air and moisture

stabilities remain the same. Importantly, 2,6-(2-thienyl)₂-

DBA(Mes)₂(23) could be separated from the 2,7-isomer 24

via HPLC. All compounds 23−28 are yellow-colored, some of

them with a greenish tint. Solutions show turquoise to green

emission already when they are exposed to daylight; strong

solid-state ﬂuorescence is observed upon irradiation with a UV

lamp (366 nm).

The introduction of p-N,N-diphenylaminophenyl substitu-

ents was facile for the dibrominated starting materials (i.e.,

5/6→29/30), whereas the yields reproducibly dropped to 45%

in the case of 31 (cf., 81% obtained for 26). Contrary to the

2-thienyl species 23/24 and 26, 29/30 and 31 decompose in

nondried, nondegassed solvents within several hours (NMR

spectroscopic and UV/vis spectroscopic control). This

instability, together with the fact that 29 and 30 showed

virtually identical retention times on the HPLC column,

precluded the isomer separation. The orange- (29/30) to red-

colored (31) compounds exhibit strong orange to red

ﬂuorescence in solution as well as in the solid state (see below).

The identity of compounds 23−31 was conﬁrmed by NMR

spectroscopy, and 23, 26−28, and 31 were further charac-

terized by X-ray crystallography (Figure 5; cf., the Supporting

Information for 27 and 31). Each of the centrosymmetric

compounds 23 and 26 crystallizes with two crystallographically

Figure 4. The complete series of accessible substitution patterns as

exempliﬁed for 2-thienyl-functionalized DBAs (23−28) as well as the

p-N,N-diphenylaminophenyl derivatives 29/30 and 31 were synthe-

sized according to the protocol outlined in Scheme 4 (ph =

p-phenylene).

independent molecules in the asymmetric unit (23_A, 23_B; 26_A,

26_B). In 23_Aand 23_B, the dihedral angles between the 2-thienyl

rings and the DBA planes amount to 28.1(2)° and 30.5(2)°,

respectively. The two vicinal 2-thienyl substituents in 28 deﬁne

one small (30.4(1)°) and one large (68.5(3)°) dihedral angle

with the DBA core; two markedly diﬀerent interplanar angles

are also found in the cases of 26_A(42.1(2)°, 55.9(2)°) and 26_B

(37.8(2)°, 67.0(2)°). We take these structural peculiarities as

an indication that each o-phenylene ring tends to maintain

maximum π-conjugation with one 2-thienyl donor, even if this

requires sacriﬁcing the donor capacity of a second substituent.

All-Carbon Analogues of 2,6-Substituted DBA(Mes)₂

Derivatives. Detailed studies, in which the optoelectronic

properties of a speciﬁc organo-heteroelement system have been

thoroughly characterized and compared to those of an

isostructural all-carbon congener, are lacking. For example,

neutral, closed-shell structural analogues of boroles, borepins,

or the highly popular dimesitylboryl group simply do not exist.

The situation is diﬀerent in the case of DBA derivatives, which

are isostructural to anthracenes and isoelectronic to the elusive

anthracene dications.

F

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Figure 5. Molecular structure of 23_A(left), 26_A(middle), and 28 (right) in the solid state. Displacement ellipsoids are drawn at the 50% probability

level; H atoms have been omitted for clarity. The thiophene moieties in 23_Aare disordered with respect to a 180° rotation about the C(4)−C(21)

bond causing S(22) and C(25) to exchange positions; the occupancy ratio is 0.727(8):0.273(8); only the major orientation is shown. The 2-thienyl

moieties in 28 are disordered with respect to a 180° rotation about the C(34)−C(51) bond [the C(35)−C(41) bond] causing S(52) and C(55)

[S(42) and C(45)] to exchange positions; the occupancy ratio is 0.538(4):0.462(4) [0.750(4):0.250(4)]; only the major orientation is shown.

Selected bond lengths (Å), bond angles (deg), and dihedral angles (deg): 23_A, B(1)−C(1) 1.559(9), B(1)−C(2A) 1.578(8), B(1)−C(11) 1.566(9),

C(1)−C(2) 1.414(8); C(1)−B(1)−C(2A) 117.6(5), C(1)−B(1)−C(11) 120.8(5), C(2)−B(1)−C(11) 121.6(5); Ar(B(1))//Ar(C(11)) 76.5(2),

Ar(C(4))//Ar(C(21)) 28.1(2). 26_A, B(1)−C(1) 1.576(6), B(1)−C(6A) 1.559(7), B(1)−C(31) 1.569(7), C(1)−C(6) 1.413(6), C(11)−S(11)

1.722(5), C(12)−S(11) 1.709(6), C(21)−S(21) 1.708(7), C(22)−S(21), 1.688(9); C(1)−B(1)−C(6A) 117.8(4), C(1)−B(1)−C(31) 118.5(4),

C(6A)−B(1)−C(31) 123.7(4), C(11)−S(11)−C(12) 91.6(3), C(21)−S(21)−C(22) 92.4(4); Ar(B(1))//Ar(C(31)) 85.7(2), Ar(C(3))//

Ar(C(11)) 42.1(2), Ar(C(4))//Ar(C(21)) 55.9(2). 28, B(1)−C(1) 1.562(3), B(1)−C(31) 1.565(3), B(1)−C(11) 1.571(3), B(2)−C(2) 1.556(3),

B(2)−C(32) 1.560(3), B(2)−C(21) 1.569(3), C(1)−C(2) 1.413(3), C(31)−C(32) 1.417(3); C(1)−B(1)−C(31) 118.68(18), C(1)−B(1)−C(11)

120.58(18), C(31)−B(1)−C(11) 120.73(18), C(2)−B(2)−C(32) 118.69(18), C(2)−B(2)−C(21) 121.67(19), C(32)−B(2)−C(21) 119.64(18),

C(35)−C(34)−C(51) 124.96(19), C(34)−C(35)−C(41) 122.75(18); Ar(B(1))//Ar(C(11)) 76.7(1), Ar(B(2))//Ar(C(21)) 88.7(1), Ar-

(C(34))//Ar(C(51)) 30.4(1), Ar(C(35))//Ar(C(41)) 68.5(1). Symmetry transformation used to generate equivalent atoms: 23_A, A, −x + 2,

−y + 1, −z + 1; 26_A, A, −x + 1, −y + 1, −z + 1.

The all-carbon relatives 32−36 of the DBA(Mes)₂deriva-

tives 2, 5, 20, 23, and 29 were prepared with the aim to com-

pare the π charge-density distributions, the crystal packing

motifs, and the optoelectronic properties of both classes of

compounds.⁴²Pristine 32 is literature-known;²⁵2,6-dibromo-

9,10-dimesitylanthracene (33) was synthesized in a one-pot

procedure by treating 2,6-dibromoanthraquinone⁴³ﬁrst with

MesMgBr and then with SnCl₂/HCl_aq(Scheme 5). Subsequent

Suzuki- or Stille-type coupling reactions aﬀorded the correspond-

ing phenyl (34), 2-thienyl (35), and p-N,N-diphenylamino-

phenyl (36) derivatives (Scheme 5).

Scheme 5. Synthesis Route to the All-Carbon Analogues

33−36

a

¹³C{¹H} NMR spectroscopy is a useful tool to map the π

charge-density distributions in arenes, because the shielding of a

speciﬁc arene carbon atom depends linearly on the π-electron

density at this position, and its chemical shift remains largely

unaﬀected by ring current anisotropies.⁴⁴The DBA derivatives

5, 20, 23, and 29 are isoelectronic to [33]²⁺−[36]²⁺and

therefore show the predictable deshielding of their C-1 to C-4

carbon nuclei (e.g., the C-1 nuclei are deshielded by 13.3−14.0

ppm). In contrast, the ¹³C resonance patterns of analogous

DBA and anthracene derivatives are essentially superimposable.

NMR spectroscopy thus provides no evidence for a stronger π

donation of the electron-rich substituents to the DBA moieties

than to the anthracene cores.

The proton chemical shift diﬀerences between the p- and

o-CH₃groups in mesityl-substituted cyclic conjugated π systems

provide a measure of the magnetic anisotropy and thereby of

the degree of aromaticity in these systems.⁴⁵For the four boron

species, these Δδ values range between 0.27 and 0.31 ppm,

whereas the Δδ values of the all-carbon compounds 33−36 are

a

Th = 2-thienyl, ph = p-phenylene. (i) (1) +MesMgBr (3 equiv in

THF), 60 °C, 2 h; (2) +SnCl₂·2H₂O (2 equiv) in 5 M HCl_aq. (ii) 34:

+PhB(OH)₂(2.4 equiv), [Pd(PtBu₃)₂] (6 mol %), Na₂CO₃(5.3

equiv), toluene/H₂O/EtOH (3:2:1), microwave synthesizer 170 °C,

45 min. 35, 36: +RSnBu₃(2.4 equiv), [Pd(PtBu₃)₂] (6 mol %),

toluene, oil bath 120 °C, 1 d or microwave synthesizer 170 °C, 45 min.

found in the interval 0.63−0.72 ppm (Δδ = δ(p-CH₃)−δ(o-

CH₃)). It is thus obvious that the ring currents in the central

anthracene C₆rings are signiﬁcantly larger than those in the

formally antiaromatic central B₂C₄rings. The Δδ values of our

DBA(Mes)₂derivatives compare well with those of other

mesityl-substituted antiaromatic boron heterocycles, that is,

6,13-dimesityl-6,13-dihydro-6,13-diborapentacene (0.37

ppm),²⁹B-mesityldibenzoborole (0.17 ppm),⁴⁶and B-mesityl-

tetraphenylborole (0.27 ppm);⁴⁷in stark contrast, Δδ equals

G

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Figure 6. Overlays of the DBA derivatives (green) with their all-carbon congeners (black) 5/33 (left) and 23_A/35 (right).

a

0.91 ppm for the formally aromatic B-mesityldibenzoborepin,⁴⁸

similar to the anthracenes 33−36.

Table 1. Electrochemical Data

E_1/2in [V] vs FcH/FcH⁺

Compounds 33 and 35 were also characterized by X-ray

crystallography. The bromo-substituted 33 is isostructural with

its DBA congener 5; a least-squares ﬁt of the tricyclic cores

yields an rms deviation of 0.123 Å (cf., Figure 6 for an overlay

of both molecular structures). The structural similarities are

going so far that the bromine atoms in 33 even exhibit the same

kind of disorder as discussed above in the case of 5. The only

noteworthy diﬀerences in the geometrical parameters of 5 and

33 relate to the B−C bonds, which are longer than the

corresponding C−C bonds in 33. Moreover, we note no

signiﬁcant deviation between the lengths of the B−C_exosingle

bond (1.576(9) Å) and the average B−C_endobond (1.561 Å).

On the contrary, an appreciably shorter average C−C_endobond

(1.407 Å) as compared to the C−C_exobond (1.498(7) Å) is

found in 33 (Δ = 0.091 Å). This feature reﬂects a low degree of

BC_endoπ bonding in the DBA(Mes)₂derivative as opposed

to signiﬁcant CC_endodouble-bond character in its anthracene

analogue and nicely matches the diﬀerences in the ring-current

eﬀects assessed above by NMR spectroscopy.

Even though the crystal structures of the 2-thienyl species 23

and 35 are not comparable because 35 crystallizes together with

1 equiv of toluene, their molecular structures are as much alike

as those of the bromo-substituted pair (cf., the dihedral angle

2-thienyl//DBA = 28° (35), 28° (23_A), 30° (23_B); Figure 6

shows an overlay of 35 and 23_A). With respect to the

BC_endo/CC_endodouble-bond character, we come to the

same conclusion as outlined above (cf., the Supporting

Information for more details).

c

5/6

10

−1.59

−1.38

c

d

23/24

25

−1.71, −2.48

−1.83, −2.63

−1.59, −2.39

−1.79, −2.84

d

e

26

27

d

e

28

−1.62, −2.41, −3.16

f f c

b

29/30

31

+0.38, +0.10, −2.11

f f c

+0.60, +0.30, −1.86

e

33

−2.20, −2.33, −2.61, −3.15, −3.39

e

35

−2.37, −3.00, −3.42

+0.55, −2.46, −2.96, −3.4

d

e

36

a

Dried and degassed THF; supporting electrolyte, [nBu₄N][PF₆] (0.1

b

mol L⁻¹); scan rate, 200 mV s⁻¹. E_1/2calibrated against internal

FcH*/FcH*⁺and recalculated versus FcH/FcH⁺(E_1/2(FcH*/FcH*⁺)

c

= −0.45 V vs FcH/FcH⁺; FcH* = (C₅Me₅)₂Fe). Decomposition

occurs at potential values more cathodic than approximately −2.5 V.

d

e

f

Quasi-reversible redox wave. E_pcvalue. E_pavalue.

9,10-dimesitylanthracene (32).²⁵The potential value of E_1/2

=

−1.84 V required for the DBA-centered one-electron reduction

of 2 is more anodic by 0.75 V than that necessary for the

monoreduction of 32 (E_1/2= −2.59 V). Both processes are

reversible. In contrast, the second one-electron reduction is

fully reversible only for 2 (E_1/2= −2.73 V), but shows features

of chemical irreversibility in the case of 32 (E_pc= −3.40 V).

The DBA species compiled in Table 1 reveal a behavior

similar to that of compound 2, with the following

modiﬁcations: (i) Halogenation of the DBA scaﬀold (i.e.,

5/6, 10) somewhat facilitates the injection of a ﬁrst electron.

At potential values more cathodic than approximately −2.5 V,

however, the compounds undergo extensive decomposition.

We currently assume that reductive cleavage of the C−Cl/Br

bonds competes with the incorporation of a second electron

Optoelectronic Properties. The DBA derivatives 5/6, 10,

and 23−31 as well as the all-carbon analogues 33, 35, and 36

were investigated by cyclic voltammetry. For a discussion of

the electrochemical data (cf., Table 1, Figure 7), it is helpful to

ﬁrst consider the cyclic voltammograms of the parent species

9,10-dimesityl-9,10-dihydro-9,10-diboraanthracene (2) and

H

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

The ﬂuorescence spectra of the two classes of compounds

show striking diﬀerences already for the two parent molecules

32 and 2. 9,10-Dimesitylanthracene (32) ﬂuoresces at λ_em

=

403 nm with a high quantum yield of 80%; the position of the

emission band is essentially independent of solvent polarity.

DBA(Mes)₂(2), in contrast, hardly ﬂuoresces at all (λ_em= 460

nm; ϕ_f= 3%), and the emission wavelength shows pronounced

positive solvatochromism (C₆H₁₂, λ_em= 415 nm; CH₂Cl₂, λ_em

=

489 nm; Δ = 3650 cm⁻¹). Anthracene ﬂuorescence is known to

originate from π→π* transitions,⁵⁰whereas the emission of 2

has been attributed to twisted intramolecular charge transfer

(TICT) transitions between the orthogonal DBA and mesityl

units.^21,25

Dibromination of 32 results in considerably reduced quan-

tum yields due to the ﬂuorescence quenching eﬀect of heavy-

atoms⁵⁰(cf., 33; ϕ_f= 53%). Because the boron-containing

molecule 2 already is a very poor emitter, halogenation in this

case has no signiﬁcant further attenuating eﬀect on the

quantum yields of the compounds obtained (5/6, 10).

The introduction of two strong p-N,N-diphenylaminophenyl

electron donors into the parent systems (32, 2) shifts the

emission bands of the all-carbon (36) as well as the boron-

containing species (29/30) by approximately 4000 cm⁻¹to the

bathochromic side (Table 2). A decisive diﬀerence between the

two classes of compounds is, however, apparent for the trend in

the quantum yields. The quantum yield of 32 is very high (ϕ_f=

80%) and does not change much for 36 (ϕ_f= 72%). Yet, in the

case of the boron-doped species, we have literally switched on a

bright ﬂuorescence upon going from 2 (ϕ_f= 3%) to 29/30

(ϕ_f= 69%) by generating a distinct donor−acceptor−donor

triad.

The availability of 23−28 oﬀers a means to investigate in

detail how the optical properties of the DBA(Mes)₂ﬂuoro-

phore change when the number or the positional arrangement

of its moderately electron-donating 2-thienyl substituents is

varied (Table 2; Figure 8). We ﬁrst note a continuous red-shift

of the emission wavelengths with an increasing number of

2-thienyl groups, which covers an overall range of 2800 cm⁻¹

(25 (λ_em= 469 nm) → 23/27/28 (λ_em= 486/515/524 nm) →

26 (λ_em= 540 nm)). Second, the ﬂuorescence maxima of the

2,6- and the 2,3-isomer 23 and 28 diﬀer by as much as

1320 cm⁻¹. Changing the positional arrangement of the pendant

groups is thus a similarly powerful set screw to modify the

optoelectronic properties of DBA ﬂuorophores as is the

introduction of an additional 2-thienyl moiety (cf., 25 → 27:

Δ = 1900 cm⁻¹).

Figure 7. Normalized cyclic voltammograms of 23, 25, 26, and 28 as

solutions in THF at room temperature; supporting electrolyte,

[nBu₄N][PF₆] (0.1 mol L⁻¹); scan rate, 200 mV s⁻¹; vs FcH/FcH⁺.

In the case of 28, an additional irreversible redox process takes place at

E_pc= −3.16 V.

into the DBA π-electron system (cf., the above discussion of

attempted Grignard or lithium−halogen exchange reactions on

5/6 and 10). (ii) With an increasing number of 2-thienyl

substituents, the corresponding DBA derivatives become more

easily reducible (cf., Figure 7). (iii) An introduction of p-N,N-

diphenylaminophenyl substituents leads to the occurrence of

two more redox events in the positive potential regime, likely

involving the electron-rich peripheral groups. In summary,

similar to pristine 2, most of the substituted 9,10-dimesityl-

9,10-dihydro-9,10-diboraanthracenes are electrochemically

well-behaved and signiﬁcantly better electron acceptors than

the all-carbon congeners.

Unless noted otherwise, the UV/vis absorption and ﬂuo-

rescence spectra discussed in the following paragraphs were

recorded in C₆H₆.⁴⁹We ﬁrst turn our attention to the UV/vis

spectral properties of the DBAs and to a comparison with the

organic model systems 32−36 (Table 2). The longest-wavelength

absorptions of the all-carbon compounds 32−36 show the char-

acteristic vibrational ﬁne structure typical of anthracene derivatives.

As compared to 32, any substitutions at the phenylene rings led to

slight bathochromic shifts. Moreover, a number of intense

hypsochromic bands occur (Table 2), which are yet irrelevant

for the color of the molecules or the calculation of Stokes shifts/

band gaps and therefore do not merit further discussions.

Most importantly, no indications for (intramolecular)

charge-transfer transitions are apparent in the UV/vis spectra

of 33−36.

This situation changes upon going to the DBA species for

which the longest-wavelength absorptions generally exhibit tails

extending far into the low-energy regime (e.g., 10, λ_max= 354

nm, onset at 450 nm; 22, λ_max= 370 nm, onset at 440 nm; 26,

λ_max= 388 nm, onset at 485 nm). The red-orange compounds

29/30 and 31, which bear strongly electron-donating p-N,N-

diphenylaminophenyl groups at the electron-poor DBA center,

are characterized by independent, broad charge-transfer

bands at λ_max= 458 nm (ε = 11 600 mol⁻¹dm³cm⁻¹; onset at

530 nm) and 459 nm (ε = 34 000 mol⁻¹dm³cm⁻¹; onset at

540 nm), respectively.

For an optimization of quantum yields, the proper position

of the 2-thienyl rings turns out to be an even more decisive

factor than a large absolute number of substituents. For

example, the quantum eﬃciency of 23 (ϕ_f= 44%) can be

improved to ϕ_f= 70% (28) by moving the two 2-thienyl groups

onto the same phenylene fragment. On the other hand, ϕ_f

decreases considerably upon proceeding from disubstituted 23

or 28 to the tetrasubstituted derivative 26 (ϕ_f= 25%). As a

general trend, best quantum eﬃciencies seem to be achievable

by gathering all 2-thienyl substituents at the same side of the

molecule (cf., 25, 27, and 28). Consequently, donor−acceptor

dyads are apparently more eﬃcient emitters than donor−

acceptor−donor triads (cf., 23, 26).

Finally, Figure 9 compiles the UV/vis ﬂuorescence spectra of

selected DBA(Mes)₂species and of their all-carbon congeners.

While the anthracene derivatives emit in the near-ultraviolet

to blue region of the visible light spectrum, the DBA derivatives

I

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 2. Photophysical Data

in C₆H₆

for thin ﬁlms

opt

λ_abs[nm]

λ_onset

[nm]

E_G

λ_ex

[nm]

λ_em

[nm]

ϕ

Stokes shift

E_HOMO/E_LUMO

λ_abs

[nm]

λ

ϕ

^fb

a

^fb

c

d

^eme

(ε [mol⁻¹dm³cm⁻¹])

[eV]

[%]

[cm⁻¹

]

[eV]

[nm]

[%]

f

5/6

352 (9900)

440

2.8

350

500

<1

8400

−6.0/−3.2

358

fg

520

<1

395 ^,

fg

424 ^,

f

10

354 (10 000)

368 (15 200)

450

430

440

2.7

2.9

2.8

350

365

305

504

<1

6

8400

4800

6200

−6.1/−3.4

−/−

350

411 ^,

488

<1

fg

20/21

22

446

460

481

370

g

462

6

408

f

304 (110 000)

370 (21 700)

11

−/−

306

370

410 ^,

489

22

fg

h

23

307 (43 700)

340 (41 800)

396 (20 800)

306 (66 300)

340 (47 500)

396 (25 700)

319 (23 400)

366 (7900)

470

450

485

460

2.6

2.7

2.6

2.7

2.3

305

315

355

315

486

44

45

66

25

72

70

4800

3300

7300

7500

7900

−/−

306

338

395

306

336

392

316

367

405

325

485

8

9

h

broad

506

h

522

23/24

25

490

−5.7/−3.1

−5.7/−3.0

−5.8/−3.2

−5.8/−3.1

−5.9/−3.2

−5.0/−2.7

499

broad

h

469

486

22

12

9

h

494

broad

406 (5700)

26

332 (66 500)

388 (26 000)

540

515

524

538

529

522

388

g

450

27

316 (42 300)

372 (14 100)

316

371

g

430

28

316 (41 400)

370 (12 800)

318

19

372

g

423

29/30

336 (40 300)

458 (11 600)

308 (110 000)

340 (119 000)

459 (34 000)

530

540

450

460

563

594

69

60

4100

5000

338

464

308

574

599

10

22

i

(2.2)

2.3

(−4.9/−2.7)

−5.2/−2.9

(−5.1/−2.9)

31

i

(2.2)

337

g

505

h

33

34

35

350 (2000)

420

430

455

3.0

2.9

2.7

365

300

328

414

53

72

37

420

560

760

−5.6/−2.6

−/−

350

367

387

409

444

4

1

h

366 (3900)

437

470

h

386 (5600)

462

499

h

407 (5600)

495

h

301 (48 000)

429

304

h

436

h

375 (2900)

455

377

460

h

g

h

396 (4500)

484

396

419

487

h

419 (4100)

h

315 (74 100)

328 (102 000)

452

−5.1/−2.4

319

483

h

479

331

517

h

f

389 (9000)

393

412

438

h

409 (14 300)

h

437 (13 900)

h

f

36

305 (25 000)

360 (32 000)

465

2.7

360

478

72

2000

−5.1/−2.4

(−5.4/−2.4)

310

360

420

493

1.5

i

h

i

f

(3.0)

498

h

418 (8000)

h

fg

443 ^,

437 (8000)

a

b

Band gap values were calculated from the onset wavelengths (λ_onset) of the absorption spectra. Quantum yields were determined by a calibrated

c

d

integrating sphere. The Stokes shifts given represent the diﬀerence between the longest absorption and shortest emission wavelengths. Energy

levels were calculated from CV data (E_LUMO= −4.8 eV − E_1/2^red; FcH/FcH⁺= −4.8 eV vs vacuum level) and from the optically determined band gap

e

f

g

h

i

(E_HOMO= E_LUMO− E_G^opt). λ_ex= 315 nm. Poor ﬁlm homogeneity. Shoulder. Vibronic ﬁne structure. Value calculated exclusively from CV data.

cover the blue to orange wavelengths. In this respect, the two

classes of compounds are therefore not competitors, but nicely

complement each other. Most band gaps fall in the intervals

between 2.6 and 2.9 eV for the DBA species and 2.7−3.0 eV for

their all-carbon analogues (Table 2); signiﬁcantly smaller

band gaps are only found for the p-aminophenyl-DBAs 29/30

J

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Despite their strong ﬂuorescence in dilute solutions,^51,52

many anthracene derivatives tend to show a poor performance

in the solid state, mainly due to π stacking, excimer formation,

and self-quenching phenomena.⁵³Even though we expected

that the bulky orthogonal mesityl substituents in 32−36 should

help to avoid unwanted π stacking, thin ﬁlms of these com-

pounds never gave ϕ_fvalues larger than 4%.⁵⁴The weak solid-

state ﬂuorescence is thus likely attributable to self-quenching

resulting from the small Stokes shifts of 32−36. In line with

that, the corresponding DBA derivatives 20/21, 23, and 29/30,

which possess larger Stokes shifts (Table 2; for a direct

comparison of 23 and 35, cf., Figure 10), give quantum yields

of 6%, 8%, and 10%, respectively. The best solid-state quantum

eﬃciencies are observed for the tetraphenylated DBA 22, the

monothienyl DBA 25 (Figure 8, bottom left), and the

tetrakis(p-aminophenyl) DBA 31 with ϕ_f= 22% for each of

them.

CONCLUSION

■

4-Bromo-1,2-bis(trimethylsilyl)benzene (3) and 4,5-dichloro-

1,2-bis(trimethylsilyl)benzene (4) are excellent starting materi-

als for the synthesis of the C-halogenated, B-mesitylated 9,10-

dihydro-9,10-diboraanthracene (DBA) derivatives 2,6/2,7-Br₂-

DBA(Mes)₂(5/6) and 2,3,6,7-Cl₄-DBA(Mes)₂(10). We have

thoroughly studied the course of the reaction between 3 and

BBr₃by in situ NMR spectroscopy and identiﬁed 4-bromo-1,2-

bis(dibromoboryl)benzene (16) as the key intermediate of the

DBA assembly; we also identiﬁed an intramolecular Br/Me-exchange

reaction in the intermediate 4-bromo-1-dibromoboryl-2-trime-

thylsilylbenzene (14) as the major yield-limiting process. The

transient diborylbenzene 16 can be trapped by electron-rich

aromatics (e.g., o-xylene) in preparatively useful yields,

which provides a convenient access route to unsymmetrically

Figure 8. Normalized emission spectra of the thienylated DBA

derivatives 23 and 25−28 in C₆H₆and photographs (hand-held UV-

lamp with λ_ex= 366 nm) of the ﬂuorescence of 25 (left) and 26 (right)

both in C₆H₆(top) and in the solid state (bottom).

(2.3 eV) and 31 (2.3 eV).⁵⁰Diﬀerences between the acceptor

strengths of a central DBA versus a 9,10-anthrylene moiety are

apparent from much more pronounced solvatochromic eﬀects

in the ﬂuorescence spectra of the DBA derivatives (cf., the

Supporting Information for more details). Comparing DBA

dyads with DBA triads, the still limited data set does not allow

for a clear-cut statement as to whether the former or the latter

show stronger solvatochromism.

Figure 9. Normalized emission spectra of the anthracene derivatives 32,²⁵35, and 36 (top) and the corresponding DBA derivatives 2,²⁵23, and 29/

30 (bottom) in C₆H₆.

K

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Figure 10. Normalized UV/vis absorption (dark green, 23; dark blue, 35) and emission spectra (light green, 23; light blue, 35) in C₆H₆. The Stokes

shifts diﬀer strongly for the two congeners with values of 94 nm (4800 cm⁻¹) for 23 and 15 nm (760 cm⁻¹) for 35.

7.15/128.06 ppm; C₆D₁₂, 1.38/26.43 ppm), external BF₃·Et₂O

(¹¹B{¹H}: 0.00 ppm), or external SiMe₄(²⁹Si INEPT: 0.00 ppm).

Abbreviations: s = singlet, d = doublet, dd = doublet of doublets, t =

triplet, vt = virtual triplet, m = multiplet, n.o. = signal not observed, Ph

= phenyl, ph = p-phenylene, Th = 2-thienyl. Signals of boron-bonded

carbon atoms were typically broadened beyond detection in the

¹³C{¹H} NMR spectra; their chemical shift values were determined

brominated DBAs (the same applies to the dichlorinated

analogue).

Stille-type coupling protocols allow the further functionaliza-

tion of chlorinated as well as brominated DBA(Mes)₂

derivatives. Using 2-thienyl as a model substituent, we have

prepared a number of benchtop stable mono- (25), di- (23, 24,

27, 28), and tetrasubstituted (26) DBA(Mes)₂species and

found that already a variation of the number and/or the

positional distribution of the donor groups enables a tuning of

the emission wavelength from 469 nm (blue) to 540 nm

(green). The corresponding quantum yields range between

25% and 72%. There is no clear-cut correlation between the

number of appended 2-thienyl groups and the quantum

eﬃciency observed; however, donor−acceptor dyads tend to

be better emitters than donor−acceptor−donor triads. Stronger

donors than the 2-thienyl ring exert an even stronger eﬀect on

the photoluminescence properties; that is, the 2,3,6,7-tetrakis-

(p-aminophenyl)-DBA(Mes)₂31 ﬂuoresces at approximately

600 nm (red).

The molecular design achieved for our C-substituted

DBA(Mes)₂species results in air and water stabilities that are

similar to those of classical organic compounds commonly used

in organic electronics. Thus, we can now gain the full beneﬁts

of boron doping, which become evident upon direct

comparison with the isostructural anthracene analogues: (i)

The ﬁrst electron aﬃnities of our DBAs are on average higher

by as much as 0.6 eV. (ii) The full tunable interval of emission

colors falls in the visible range of the spectrum (blue to

orange), while anthracenes emit in the near-ultraviolet to blue

wavelength regime. (iii) Our DBAs show signiﬁcant solid-state

quantum yields, whereas those of the corresponding anthracenes

are negligible.

using the cross peaks in the HMBC spectra. UV/vis absorption and

emission spectra were recorded on a Varian Cary 50 Scan UV/vis

spectrophotometer and a Jasco FP-8300 spectroﬂuorometer, respec-

tively. The absolute ﬂuorescence quantum yields (ϕ_f) were determined

using a calibrated integrating sphere system (ILF-835 100 mm

diameter Integrating Sphere; Jasco), a quantum yield calculation

program (FWQE-880; Jasco), and highly diluted samples of at least

ﬁve diﬀerent optical densities in each measurement. For the

determination of the quantum yields of thin ﬁlms, at least four ﬁlms

of varying thicknesses were generated for each compound by

evaporation of diﬀerently concentrated samples in CHCl₃at 100 °C

on a quartz-glass surface. Electrochemical measurements were

performed at room temperature by using an EG&G Princeton Applied

Research 263A potentiostat with a platinum disk working electrode

(diameter 2.00 mm). The reference electrode was a silver wire on

which AgCl had been deposited by immersing the wire into HCl/

HNO₃(3:1). The solvent (THF) was dried at room temperature over

sodium without added benzoquinone and degassed by ﬁve freeze−

pump−thaw cycles. [nBu₄N][PF₆] (0.1 mol/L) was employed as the

supporting electrolyte. All potential values are referenced against the

FcH/FcH⁺couple (E_1/2= 0 V). Combustion analyses were performed

by the Microanalytical Laboratory of the University of Frankfurt or by

the Microanalytical Laboratory Pascher (Remagen, Germany).

General Procedure for the Preparation of the Halogenated

DBA(Mes)₂Derivatives 5−13. A thick-walled glass ampule was

charged with the respective 1,2-bis(trimethylsilyl)benzene 3 or 4

(3.6 mmol), BBr₃(1.0 mL, 2.6 g, 11 mmol), and the appropriate

solvent or solvent mixture (5 mL). The ampule was ﬂame-sealed under

vacuum and subsequently heated to 120 °C for 1.5 d. The dark yellow-

colored solution obtained was cooled to room temperature, transferred

to a Schlenk ﬂask, and evaporated to dryness at 100 °C in vacuo. The

resulting solid was dissolved in toluene (10 mL) without further

puriﬁcation, and a THF solution of MesMgBr (0.86 M, 8.4 mL,

7.2 mmol) was added dropwise with stirring over 10 min at 0 °C. The

reaction mixture was allowed to warm to room temperature overnight,

quenched with aqueous saturated NH₄Cl (10 mL), and the organic

phase was separated with a separation funnel. The aqueous phase was

extracted with CHCl₃(3 × 10 mL), and all organic phases were

combined, washed with H₂O (2 × 10 mL), dried over anhydrous

MgSO₄, and ﬁltered. All volatiles were removed from the ﬁltrate in

vacuo, and the residual crude material was puriﬁed by column

chromatography. The products obtained are air- and water-stable

EXPERIMENTAL SECTION

■

General Considerations. Unless otherwise speciﬁed, all reactions

were carried out under dry nitrogen using Schlenk techniques. Hexane,

toluene, o-xylene, THF, and C₆D₆were dried over Na/benzophenone

and distilled prior to use. The compounds 4-bromo-1,2-bis-

(trimethylsilyl)benzene 3,³²4,5-dichloro-1,2-bis(trimethylsilyl)-

benzene 4,³¹and tri-n-butyl(p-N,N-diphenylaminophenyl)stannane⁵⁵

were synthesized according to literature procedures. The microwave-

assisted reactions were performed using a Biotage Initiator⁺microwave

synthesizer with vials obtained from Biotage.

NMR spectra were recorded on Bruker DPX 250, Avance 300,

and Avance 400 spectrometers. Chemical shifts are referenced to

(residual) solvent signals (¹H/¹³C{¹H}: CDCl₃, 7.26/77.16 ppm; C₆D₆,

L

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

yellow powders, which show almost no ﬂuorescence in solution or the

solid state. The synthesis has been scaled up to 10 mmol of 1,2-

bis(trimethylsilyl)benzene without losses in isolated yields.

Anal. Calcd for C₃₀H₂₆B₂Cl₄(549.93): C, 65.52; H, 4.77. Found: C,

65.03; H, 4.60.

Synthesis of 2,3-Cl₂-6,7-Me₂-DBA(Mes)₂(13). The synthesis

was performed according to the general procedure for the preparation

of halogenated DBA(Mes)₂derivatives. A 1:1 mixture of hexane/o-

xylene was used in the ﬁrst reaction step. The puriﬁcation of the crude

product by column chromatography (silica gel, hexane:CHCl₃= 3:1)

aﬀorded a mixture of 10 and 13 in a 1:7 ratio as a yellow powder (the

product distribution was determined by ¹H NMR spectroscopy; it

varies between individual experiments, and the average distribution of

four runs is given). Overall yield: 0.32 g (34%) containing 0.27 g (30%

yield) of 13. NMR data of 13, ¹H NMR (300.0 MHz, CDCl₃): δ 2.08

(s, 12H; o-CH₃), 2.25 (s, 6H; DBA-CH₃), 2.43 (s, 6H; p-CH₃), 6.96

(s, 4H; MesH), 7.42 (s, 2H; H-5,8), 7.59 (s, 2H; H-1,4). ¹¹B NMR

(96.3 MHz, CDCl₃): δ 69 (h_1/2= 1300 Hz). ¹³C NMR (75.4 MHz,

CDCl₃): δ 20.2 (DBA-CH₃), 21.4 (p-CH₃), 22.8 (o-CH₃), 127.3

(MesC-3,5), 137.2 (MesC-4), 137.9 (C-2,3), 138.0 (Mes-C2,6), 140.0

(C-1,4), 140.1 (MesC-1), 141.1 (C-5,8), 142.6 (CB), 143.4 (C-6,7),

145.1 (CB).

General Procedure for the Stille-type Coupling Reactions.

The respective halogenated DBA(Mes)₂derivative or derivative

mixture (100 μmol DBA equiv), [Pd(PtBu₃)₂] (2.5 mol % per

halogen atom), and the appropriate tri-n-butylstannane (1.2 equiv per

halogen atom) in toluene (3 mL) were heated with stirring either in an

oil bath (120 °C, 1 d) or in a microwave synthesizer (170 °C, 45 min).

The reaction mixture was allowed to cool to room temperature, all

volatiles were removed in vacuo, and the residual crude material was

puriﬁed by column chromatography. Because all target products are

highly ﬂuorescent even when adsorbed on the column, the product

bands can easily be identiﬁed on the stationary phase with the help of a

hand-held UV-lamp (λ_ex= 366 nm). The synthesis has been scaled up

to 0.5 mmol of halogenated DBA equiv without losses in isolated

yields.

Synthesis of 2,6/2,7-Br₂-DBA(Mes)₂(5/6). The synthesis was

performed according to the general procedure for the preparation of

halogenated DBA(Mes)₂derivatives. For a maximum yield, hexane was

used as the solvent in the ﬁrst reaction step. The puriﬁcation of the

crude product by column chromatography (silica gel, hexane:CHCl₃=

3:1) aﬀorded a mixture of the two isomers 5 and 6 in a 1:1 ratio as a

1

yellow powder (the isomer distribution was determined by H NMR

spectroscopy). Overall yield: 0.44 g (0.77 mmol, 43%). To achieve

maximum selectivity for the 2,6-isomer (5:6 = 3:2), the reaction was

conducted in C₆H₆; workup was the same as described above. Overall

yield: 0.32 g (0.56 mmol, 31%). 5 crystallized upon slow evaporation

of a CHCl₃solution of 5/6 in the form of light brown plates that were

suitable for X-ray crystal structure analysis. Attempts to separate the

two isomers by HPLC using either reversed-phase Reprosil-Pur C18-

AQ or normal phase Nucleosil columns (10 μm, 250 × 20 mm) with

various eluents were not successful. Analytical data for mixtures of 5/6,

1

NMR data of 5, H NMR (300.0 MHz, CDCl₃): δ 2.04 (s, 12H; o-

3

CH₃), 2.40 (s, 6H; p-CH₃), 6.94 (s, 4H; MesH), 7.46 (d, J_HH= 7.8

3

4

Hz, 2H; H-4,8), 7.61 (dd, J_HH= 7.8 Hz, J_HH= 2.1 Hz, 2H; H-3,7),

7.70 (d, J_HH= 2.1 Hz, 2H; H-1,5). ¹¹B NMR (96.3 MHz, CDCl₃): δ

4

71 (h_1/2= 900 Hz). ¹³C NMR (75.4 MHz, CDCl₃): δ 21.5 (p-CH₃),

22.8 (o-CH₃), 127.4 (MesC-3,5), 130.4 (C-2,6), 136.8 (C-3,7), 137.6

(MesC-4), 138.0 (MesC-2,6), 139.2 (MesC-1), 140.9 (C-4,8), 141.5

(C-1,5), 142.8, 146.9 (CB). NMR data of 6 (only the resonances

mentioned could be unequivocally identiﬁed; missing resonances are

1

overlapping with signals of the 2,6-isomer), H NMR (300.0 MHz,

CDCl₃): δ 2.03, 2.05 (2 × s, 2 × 6H; o-CH₃), 2.39, 2.41 (2 × s, 2 ×

3

4

3H; p-CH₃), 7.47 (d, J_HH= 7.8 Hz, 2H; H-4,5), 7.69 (d, J_HH

=

2.1 Hz, 2H; H-1,8). ¹³C NMR (75.4 MHz, CDCl₃): δ 21.5, 21.5

(p-CH₃), 22.8, 22.9 (o-CH₃), 127.4, 127.5 (MesC-3,5), 130.2 (C-2,7),

136.9 (C-3,6), 137.5, 137.7 (MesC-4), 137.9, 138.1 (MesC-2,6), 140.7

(C-4,5), 141.6 (C-1,8), 142.9, 146.9 (CB). Anal. Calcd for

C₃₀H₂₈B₂Br₂(569.96): C, 63.22; H, 4.95. Found: C, 62.60; H, 5.15.

Synthesis of 2-Br-6,7-Me₂-DBA(Mes)₂(9). The synthesis was

performed according to the general procedure for the preparation of

halogenated DBA(Mes)₂derivatives. A 1:1 mixture of hexane/o-xylene

was used as the solvent in the ﬁrst reaction step. The puriﬁcation of

the crude product by column chromatography (silica gel, hexa-

ne:CHCl₃= 3:1) aﬀorded a mixture of 5/6 and 9 in a 3:2:10 ratio as a

Synthesis of Ph₄-DBA(Mes)₂(22). The synthesis was performed

according to the general procedure for the Stille-type coupling

reactions using 10 (100 μmol). The puriﬁcation of the crude product

by column chromatography (silica gel, hexane:CHCl₃= 5:1) aﬀorded

22 as an air- and water-stable yellow powder showing a turquoise

ﬂuorescence both in the solid state and in solution. Yield: 29 mg

1

(40 μmol, 40%). H NMR (300.0 MHz, CDCl₃): δ 2.18 (s, 12H; o-

CH₃), 2.33 (s, 6H; p-CH₃), 6.87 (s, 4H; MesH), 7.05−7.08 (m, 8H;

PhH-2,6), 7.16−7.19 (m, 12H; PhH-3,4,5), 7.70 (s, 4H; H-1,4,5,8).

¹¹B NMR (96.3 MHz, CDCl₃): δ 73 (h_1/2= 1500 Hz). ¹³C NMR (75.4

MHz, CDCl₃): δ 21.4 (p-CH₃), 23.1 (o-CH₃), 126.9 (PhC-4), 127.2

(MesC-3,5), 128.0 (PhC-3,5), 130.0 (PhC-2,6), 136.7 (MesC-4),

137.9 (MesC-2,6), 140.5 (MesC-1), 141.2 (C-1,4,5,8), 141.2 (PhC-1),

144.9 (CB), 145.2 (C-2,3,6,7). Anal. Calcd for C₅₄H₄₆B₂(716.57): C,

90.51; H, 6.47. Found: C, 90.20; H, 6.67.

1

yellow powder (the product distribution was determined by H NMR

spectroscopy). Overall yield 5/6 and 9: 0.32 g (33%) containing 0.21 g

1

(22% yield) of 9. NMR data of 9, H NMR (300.0 MHz, CDCl₃): δ

2.06, 2.07 (2 × s, 2 × 6H; o-CH₃), 2.23 (s, 6H; DBA-CH₃), 2.41, 2.42

3

(2 × s, 2 × 3H; p-CH₃), 6.94 (s, 4H; MesH), 7.39 (d, J_HH= 7.8 Hz,

3

1H; H-4), 7.39, 7.40 (2 × s, 2 × 1H; H-5, H-8), 7.55 (dd, J_HH= 7.8

4

Hz, J_HH= 2.1 Hz, 1H; H-3), 7.63 (d, J_HH= 2.1 Hz, 1H; H-1). ¹¹B

NMR (96.3 MHz, CDCl₃): δ 69 (h_1/2= 800 Hz). ¹³C NMR (75.4

MHz, CDCl₃): δ 20.2 (DBA-CH₃), 21.4, 21.5 (p-CH₃), 22.7, 22.8

(o-CH₃), 127.1, 127.2 (MesC-3,5), 129.5 (C-2), 136.1 (C-3), 136.9,

137.0 (MesC-4), 138.0, 138.1 (MesC-2,6), 140.2 (MesC-1), 140.8,

140.8, 140.9, 141.0 (C-1,4,5,8), 143.1, 143.3 (C-6,7), 147.5 (CB).

Synthesis of Cl₄-DBA(Mes)₂(10). The synthesis was performed

according to the general procedure for the preparation of halogenated

DBA(Mes)₂derivatives. Hexane was used as the solvent in the ﬁrst

reaction step. Cl₄DBA(Br)₂crystallized already in the ampule in the

form of thick colorless needles, which were several centimeters long

and suitable for X-ray crystal structure analysis. The puriﬁcation of

crude 10 by column chromatography (silica gel, hexane:CHCl₃= 3:1)

aﬀorded the product as a microcrystalline yellow solid. Overall yield:

0.59 g (1.1 mmol, 60%). 10 crystallized upon slow evaporation of its

CHCl₃/pentane solution in the form of yellow blocks that were

Synthesis of 2,6/2,7-(2-Thienyl)₂-DBA(Mes)₂(23/24). The

synthesis was performed according to the general procedure for the

Stille-type coupling reactions using 5 (60 μmol) and 6 (40 μmol). The

puriﬁcation of the crude product by column chromatography (silica

gel, hexane:CHCl₃= 3:1) aﬀorded 23/24 (3:2 ratio) as an air-and

water-stable yellow-green powder showing a turquoise ﬂuorescence

both in the solid state and in solution. Yield of 23/24: 46 mg (80

μmol, 80%). 23 can be separated from 24 by HPLC using a normal-

phase Nucleosil semipreparative column (10 μm, 250 × 20 mm) and

hexane as the eluent; the fraction containing 24 was, however, still

contaminated with 23. 23 crystallized in the form of yellow X-ray

quality needles upon layering a CHCl₃solution of 23/24 with

pentane. Analytical data of 23, ¹H NMR (300.0 MHz, CDCl₃): δ 2.10

(s, 12H; o-CH₃), 2.42 (s, 6H; p-CH₃), 6.94 (s, 4H; MesH), 7.05 (dd,

³J_HH= 4.9 Hz, 3.8 Hz, 2H; ThH-4), 7.26−7.29 (m, 4H; ThH-3, ThH-

5), 7.65 (m, 4H; H-3,7, H-4,8), 7.85 (m, 2H; H-1,5). ¹¹B NMR (96.3

MHz, CDCl₃): δ 70 (h_1/2= 1000 Hz). ¹³C NMR (75.4 MHz, CDCl₃):

δ 21.5 (p-CH₃), 22.9 (o-CH₃), 124.6, 126.2 (ThC-3, ThC-5), 127.2

(MesC-3,5), 128.3 (ThC-4), 130.1 (C-3,7 or C-4,8), 136.0 (C-1,5),

137.0 (MesC-4), 138.1 (MesC-2,6), 139.1 (C-2,6), 139.9 (C-3,7 or C-

4,8), 140.6 (MesC-1), 144.0 (ThC-2), 146.3 (CB). The following

analytical data were obtained on mixtures of 23/24. NMR data of 24

1

suitable for X-ray crystal structure analysis. H NMR (300.0 MHz,

CDCl₃): δ 2.03 (s, 12H; o-CH₃), 2.41 (s, 6H; p-CH₃), 6.95 (s, 4H;

MesH), 7.64 (s, 4H; H-1,4,5,8). ¹¹B NMR (96.3 MHz, CDCl₃): δ 74

(h_1/2= 1000 Hz). ¹³C NMR (75.4 MHz, CDCl₃): δ 21.4 (p-CH₃),

22.8 (o-CH₃), 127.6 (MesC-3,5), 137.8 (MesC-2,6), 138.0 (MesC-4),

138.9 (C-2,3,6,7), 140.2 (MesC-1), 140.7 (C-1,4,5,8), 144.3 (CB).

M

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

(only the resonances mentioned could be unequivocally identiﬁed;

missing resonances are overlapping with signals of the 2,6-isomer), ¹³C

NMR (75.4 MHz, CDCl₃): δ 21.4, 21.5 (p-CH₃), 22.8, 22.9 (o-CH₃),

124.5, 126.2 (ThC-3, ThC-5), 127.1, 127.3 (MesC-3,5), 128.4 (ThC-

4), 130.3 (C-3,6 or C-4,5), 136.2 (C-1,8), 137.0 (MesC-4), 137.9,

138.2 (MesC-2,6), 138.9 (C-2,7 or ThC-2), 139.6 (C-3,6 or C-4,5).

Anal. Calcd for C₃₈H₃₄B₂S₂(576.39): C, 79.18; H, 5.95; S, 11.13.

Found: C, 78.27*; H, 6.16; S, 10.84. “*” represents incomplete

combustion due to boron carbide formation.

ThC-5), 127.1 (ThC-4), 127.1 (MesC-3,5), 127.7 (ThC-3 or ThC-5),

136.7 (MesC-4), 138.0 (C-2,3), 138.1 (MesC-2,6), 140.7 (MesC-1),

140.8 (C-5,8), 141.0 (C-1,4), 142.5 (ThC-2), 142.9 (C-6,7), 144.7

(CB). Anal. Calcd for C₄₀H₃₈B₂S₂(604.48): C, 79.48; H, 6.34; S,

10.61. Found: C, 79.08; H, 6.35; S, 10.55.

Synthesis of 2,6/2,7-(p-N,N-Diphenylaminophenyl)₂-DBA-

(Mes)₂(29/30). The synthesis was performed according to the

general procedure for the Stille-type coupling reactions using 5

(60 μmol) and 6 (40 μmol). The puriﬁcation of the crude product by

column chromatography (silica gel, hexane:CHCl₃= 3:1) aﬀorded 29/

30 (3:2 ratio) as an air- and water-stable orange powder showing an

orange ﬂuorescence in the solid state. In solution, the compound

ﬂuoresces with a pronounced positive solvatochromism (cyclohexane,

green; C₆H₆and CHCl₃, orange; THF, red). In nondried, non-

degassed solvents, the compound decomposes over the course of 1 d.

Attempts to separate the two isomers by HPLC using either reversed-

phase Reprosil-Pur C18-AQ or normal phase Nucleosil columns

(10 μm, 250 × 20 mm) with various eluents were not successful. Yield:

70 mg (78 μmol, 78%). Analytical data for mixtures of 29/30, NMR

Synthesis of 2-(2-Thienyl)-6,7-Me₂-DBA(Mes)₂(25). The

synthesis was performed according to the general procedure for the

Stille-type coupling reactions using a mixture of 5/6 (33 μmol) and 9

(67 μmol). 23/24 and 25 can be separated from each other by column

chromatography (silica gel, hexane:CHCl₃= 5:1) and thereby

obtained as analytically pure samples. The target product 25 is an

air- and water-stable yellow-green powder showing a turquoise ﬂuore-

scence both in the solid state and in solution. Yield: 32 mg (61 μmol,

91%). ¹H NMR (300.0 MHz, CDCl₃): δ 2.09 (s, 12H; o-CH₃), 2.24 (s,

6H; DBA-CH₃), 2.42 (s, 6H; p-CH₃), 6.94, 6.94 (2 × s, 2 × 2H;

MesH), 7.03 (dd, ³J_HH= 5.0, 3.7 Hz, 1H; ThH-4), 7.24−7.28 (m, 2H;

1

data of 29, H NMR (300.0 MHz, CDCl₃): δ 2.12 (s, 12H; o-CH₃),

ThH-3, ThH-5), 7.41, 7.42 (2 × s, 2 × 1H; H-5, H-8), 7.55 (d, ³J_HH

=

2.40 (s, 6H; p-CH₃), 6.93 (s, 4H; MesH), 7.01−7.08 (m, 8H; PhH-4,

phH-3,5), 7.10−7.13 (m, 8H; PhH-2,6), 7.24−7.29 (m, 8H; PhH-3,5),

7.37−7.39 (m, 4H; phH-2,6), 7.61−7.68 (m, 4H; H-3,7, H-4,8), 7.82−

7.83 (m, 2H; H-1,5). ¹¹B NMR (96.3 MHz, CDCl₃): n.o. ¹³C NMR

(75.4 MHz, CDCl₃): δ 21.5 (p-CH₃), 22.9 (o-CH₃), 123.3, 123.4

(phC-3,5, PhC-4), 124.9 (PhC-2,6), 127.1 (MesC-3,5), 128.1 (phC-

2,6), 129.5 (PhC-3,5), 131.0 (C-3,7), 134.2 (phC-1), 136.8 (MesC-4),

137.0 (C-1,5), 138.1 (MesC-2,6), 139.7 (C-4,8), 141.1 (MesC-1),

143.7 (CB), 145.4 (C-2,6), 147.6 (PhC-1), 148.0 (phC-4). NMR data

of 30 (only the resonances mentioned could be unequivocally

identiﬁed; missing resonances are overlapping with signals of the

3

4

7.6 Hz, 1H; H-4), 7.62 (dd, J_HH= 7.6 Hz, J_HH= 1.8 Hz, 1H; H-3),

7.78 (d, ⁴J_HH= 1.8 Hz, 1H; H-1). ¹¹B NMR (96.3 MHz, CDCl₃): δ 69

(h_1/2= 1000 Hz). ¹³C NMR (75.4 MHz, CDCl₃): δ 20.2, 20.2 (DBA-

CH₃), 21.5, 21.5 (p-CH₃), 22.8, 22.9 (o-CH₃), 124.4, 126.0 (ThC-3,

ThC-5), 127.0, 127.1 (MesC-3,5), 128.3 (ThC-4), 130.0 (C-3), 135.7

(C-1), 136.7 (MesC-4), 138.0, 138.2 (MesC-2,6), 138.7 (C-2), 139.3

(C-4), 140.7, 140.9 (C-5, C-8), 140.9 (MesC-1), 142.7, 142.9 (C-6, C-

7), 144.2 (ThC-2), 146.1 (CB). Anal. Calcd for C₃₈H₃₄B₂S (576.43):

C, 82.78; H, 6.95; S, 6.14. Found: C, 82.58; H, 6.87; S, 6.31.

Synthesis of (2-Thienyl)₄-DBA(Mes)₂(26). The synthesis was

performed according to the general procedure for the Stille-type

coupling reactions using 10 (100 μmol). The puriﬁcation of the crude

product by column chromatography (silica gel, hexane:CHCl₃= 3:1)

aﬀorded 26 as an air- and water-stable yellow powder showing a yellow-

green ﬂuorescence in the solid state. The compound is completely air-

and water-stable also in solution and ﬂuoresces with a pronounced

positive solvatochromism (cyclohexane, green; C₆H₆and CHCl₃,

yellow-green; THF, yellow; acetone, yellow-orange). Yield: 60 mg

(81 mmol, 81%). 26 crystallized in the form of yellow X-ray quality

needles upon layering its CHCl₃solution with pentane. ¹H NMR

(300.0 MHz, CDCl₃): δ 2.15 (s, 12H; o-CH₃), 2.37 (s, 6H; p-CH₃),

6.88 (dd, ³J_HH= 3.6 Hz, ⁴J_HH= 1.1 Hz, 4H; ThH-3 or ThH-5), 6.90 (s,

1

2,6-isomer), H NMR (300.0 MHz, CDCl₃): δ 2.37, 2.42 (2 × s, 2 ×

3H; p-CH₃), 6.90, 6.96 (2 × s, 2 × 2H; MesH). ¹³C NMR (75.4 MHz,

CDCl₃): δ 21.3 (p-CH₃), 22.8, 23.0 (o-CH₃), 127.1 (MesC-3,5),

137.9, 138.3 (Mes-C2,6). Anal. Calcd for C₆₆H₅₆B₂N₂(898.79): C,

88.20; H, 6.28; N, 3.12. Found: C, 87.47; H, 6.41; N, 3.55.

Synthesis of (p-N,N-Diphenylaminophenyl)₄-DBA(Mes)₂(31).

The synthesis was performed according to the general procedure for

the Stille-type coupling reactions using 10 (100 μmol). The

puriﬁcation of the crude product by column chromatography (silica

gel, hexane:CHCl₃= 3:1) aﬀorded 31 as an air- and water-stable red

powder showing an orange-red ﬂuorescence in the solid state. In

solution, the compound ﬂuoresces intensely with a pronounced

positive solvatochromism (cyclohexane, green; C₆H₆, orange; CHCl₃,

light red; THF, red; acetone, dark red). In nondried, nondegassed

solutions, 31 decomposes over the course of 1 d. Yield: 63 mg (45

μmol, 45%). 31 crystallized in the form of thin red X-ray quality

needles upon layering its CHCl₃solution with hexane. ¹H NMR

(300.0 MHz, CDCl₃): δ 2.19 (s, 12H; o-CH₃), 2.37 (s, 6H; p-CH₃),

6.88−6.94 (m, 20H; phH, MesH), 6.99−7.02 (m, 8H; PhH-4), 7.07−

7.09 (m, 16H; PhH-2,6), 7.21−7.25 (m, 16H; PhH-3,5), 7.66 (s, 4H;

H-1,4,5,8). ¹¹B NMR (96.3 MHz, CDCl₃): n.o. ¹³C NMR (75.4 MHz,

CDCl₃): δ 21.4 (p-CH₃), 23.2 (o-CH₃), 122.6 (phC-3,5), 123.1 (PhC-4),

124.7 (PhC-2,6), 127.2 (MesC-3,5), 129.4 (PhC-3,5), 130.8 (phC-

2,6), 135.1 (phC-1), 136.6 (MesC-4), 138.2 (MesC-2,6), 140.8 (C-

1,4,5,8), 141.1 (MesC-1), 144.4 (CB), 144.9 (C-2,3,6,7), 146.7 (phC-4),

147.7 (PhC-1). Anal. Calcd for C₁₀₂H₈₂B₂N₄(1385.39): C, 88.43; H,

5.97; N, 4.04. Found: C, 88.14; H, 6.14; N, 3.59.

Crystal Structure Determinations. Data for Cl₄-DBA(Br)₂, 5,

10, and 20 were collected on a STOE IPDS II two-circle diﬀracto-

meter with graphite-monochromated Mo K_αradiation (λ = 0.71073

Å). Data for 23, 26, 27, 28, 31, 33, and 35 were collected on a STOE

IPDS II two-circle diﬀractometer with a Genix Microfocus tube with

mirror optics using Mo K_αradiation (λ = 0.71073 Å). For Cl₄-

DBA(Br)₂, 5, 10, and 20, an empirical absorption correction with the

program PLATON⁵⁶was performed. The data for 23, 26, 27, 28, 31,

33, and 35 were scaled using the frame scaling procedure in the

X-AREA program system.⁵⁷The structures were solved by direct

methods using the program SHELXS⁵⁸and reﬁned against F²with full-

matrix least-squares techniques using the program SHELXL-97.⁵⁸

3

4H; MesH), 6.94 (dd, J_HH= 5.0, 3.6 Hz, 4H; ThH-4), 7.26 (dd,

4

³J_HH= 5.0 Hz, J_HH= 1.1 Hz, 4H; ThH-3 or ThH-5), 7.76 (s, 4H;

H-1,4,5,8). ¹¹B NMR (96.3 MHz, CDCl₃): δ 70 (h_1/2= 1500 Hz). ¹³

C

NMR (75.4 MHz, CDCl₃): δ 21.4 (p-CH₃), 23.1 (o-CH₃), 126.9

(ThC-3 or ThC-5), 127.1 (ThC-4), 127.3 (MesC-3,5), 127.8 (ThC-3

or ThC-5), 137.0 (MesC-4), 138.0 (MesC-2,6), 138.4 (C-2,3,6,7),

140.0 (MesC-1), 141.5 (C-1,4,5,8), 142.3 (ThC-2), 144.7 (CB). Anal.

Calcd for C₄₆H₃₈B₂S₄(740.62): C, 74.59; H, 5.17; S, 17.32. Found: C,

73.59*; H, 5.25; S, 17.17. “*” represents incomplete combustion due

to boron carbide formation.

Synthesis of 2,3-(2-Thienyl)₂-6,7-Me₂-DBA(Mes)₂(27). The

synthesis was performed according to the general procedure for the

Stille-type coupling reactions using a mixture of 10 (13 μmol) and 13

(87 μmol). 26 and 27 can be separated from each other by column

chromatography (silica gel, hexane:CHCl₃= 4:1) and thereby

obtained as analytically pure samples. The target product 27 is an

air- and water-stable green powder showing a yellow-green ﬂuo-

rescence in the solid state and in solution. Yield: 50 mg (83 μmol,

95%). 27 crystallized upon slow evaporation of its CHCl₃/hexane

solution in the form of yellow needles that were suitable for X-ray

1

crystal structure analysis. H NMR (300.0 MHz, CDCl₃): δ 2.10 (s,

12H; o-CH₃), 2.24 (s, 6H; DBA-CH₃), 2.39 (s, 6H; p-CH₃), 6.84 (dd,

4

³J_HH= 3.6 Hz, J_HH= 1.2 Hz, 2H; ThH-3 or ThH-5), 6.91−6.93 (m,

3

4

6H; MesH, ThH-4), 7.24 (dd, J_HH= 5.1, J_HH= 1.2 Hz, 2H; ThH-3

or ThH-5), 7.40 (s, 2H; H-5,8), 7.67 (s, 2H; H-1,4). ¹¹B NMR (96.3

MHz, CDCl₃): δ 70 (h_1/2= 900 Hz). ¹³C NMR (75.4 MHz, CDCl₃):

δ 20.2 (DBA-CH₃), 21.4 (p-CH₃), 23.0 (o-CH₃), 126.7 (ThC-3 or

N

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

The Br atom in 5 is disordered over two sites with a site occupation

factor of 0.858(3) for the major occupied site.

(7) Ben

́

ard, C. P.; Geng, Z.; Heuft, M. A.; VanCrey, K.; Fallis, A. G. J.

Org. Chem. 2007, 72, 7229−7236.

Both 2-thienyl rings in 23 are disordered over two sites with a site

occupation factor of 0.939(8) and 0.727(8), respectively, for the major

occupied site. The coordinates and displacement parameters of the

overlying S and C atoms were constrained to be equal.

In one molecule of 26, both 2-thienyl rings are disordered over two

sites with a site occupation factor of 0.709(9) and 0.653(7), res-

pectively, for the major occupied site. The coordinates and displace-

ment parameters of the overlying S and C atoms were constrained to

be equal.

Both 2-thienyl rings in 28 are disordered over two sites with a site

occupation factor of 0.750(4) and 0.538(4), respectively, for the major

occupied site. The coordinates and displacement parameters of the

overlying S and C atoms were constrained to be equal. The absolute

structure was determined (Flack-x-parameter 0.01(9)).

The displacement ellipsoids of all atoms in 31 were restrained to an

isotropic behavior. Bond lengths and angles in the chloroform

molecules were restrained to be equal. One chloroform molecule was

isotropically reﬁned.

(8) Anthony, J. E. Chem. Rev. 2006, 106, 5028−5048.

(9) Selected reviews: (a) Entwistle, C. D.; Marder, T. B. Angew.

Chem., Int. Ed. 2002, 41, 2927−2931. (b) Entwistle, C. D.; Marder, T.

B. Chem. Mater. 2004, 16, 4574−4585. (c) Jakle, F. Coord. Chem. Rev.

̈

2006, 250, 1107−1121. (d) Yamaguchi, S.; Wakamiya, A. Pure Appl.

Chem. 2006, 78, 1413−1424. (e) Matsumi, N.; Chujo, Y. Polym. J.

2008, 40, 77−89. (f) Elbing, M.; Bazan, G. C. Angew. Chem., Int. Ed.

2008, 47, 834−838. (g) Jakle, F. Chem. Rev. 2010, 110, 3985−4022.

̈

(10) Yu, G.; Yin, S.; Liu, Y.; Chen, J.; Xu, X.; Sun, X.; Ma, D.; Zhan,

X.; Peng, Q.; Shuai, Z.; Tang, B.; Zhu, D.; Fang, W.; Luo, Y. J. Am.

Chem. Soc. 2005, 127, 6335−6346.

(11) Shirota, Y.; Kageyama, H. Chem. Rev. 2007, 107, 953−1010.

́

(12) Baumgartner, T.; Reau, R. Chem. Rev. 2006, 106, 4681−4727.

(13) Fichou, D. Handbook of Oligo- and Polythiophenes; Wiley-VCH:

Weinheim, 1999.

(14) Lorbach, A.; Bolte, M.; Li, H.; Lerner, H.-W.; Holthausen, M.

C.; Jakle, F.; Wagner, M. Angew. Chem., Int. Ed. 2009, 48, 4584−4588.

̈

(15) Chai, J.; Wang, C.; Jia, L.; Pang, Y.; Graham, M.; Cheng, S. Z. D.

Synth. Met. 2009, 159, 1443−1449.

CCDC reference numbers: 945733 (Cl₄-DBA(Br)₂), 945734 (5),

945735 (10), 945736 (20), 945737 (23), 945738 (26), 945739 (27),

945740 (28), 945741 (31), 945742 (33), 945743 (35).

(16) Lorbach, A.; Bolte, M.; Lerner, H.-W.; Wagner, M. Chem.

Commun. 2010, 46, 3592−3594.

(17) Lorbach, A.; Bolte, M.; Lerner, H.-W.; Wagner, M. Organo-

metallics 2010, 29, 5762−5765.

ASSOCIATED CONTENT

* Supporting Information

■

S

(18) Agou, T.; Sekine, M.; Kawashima, T. Tetrahedron Lett. 2010, 51,

5013−5015.

Experimental details including assigned NMR data of 7/8, 11,

12, 20/21, 28, 33, 34, 35, and 36. Experimental details and

NMR data of 14, 15/17, 16, 18, and 19 in conjunction with the

mechanistic investigation of the DBA formation. X-ray crystal

structure analyses of Cl₄-DBA(Br)₂, 20, 27, 31, 33, and 35 and

key crystallographic data of Cl₄-DBA(Br)₂, 5, 10, 20, 23, 26, 27,

28, 31, 33, and 35. Cyclic voltammograms of compounds 5/6,

10, 27, 29/30, 31, 33, 35, and 36. Absorption and emission

spectra of the compounds 5/6, 10, 20−31, and 33−36 in C₆H₆.

Solvatochromism of the compounds 23−31, 35, and 36. Mass

(19) Januszewski, E.; Lorbach, A.; Grewal, R.; Bolte, M.; Bats, J. W.;

Lerner, H.-W.; Wagner, M. Chem.-Eur. J. 2011, 17, 12696−12705.

(20) Lorbach, A.; Hubner, A.; Wagner, M. Dalton Trans. 2012, 41,

̈

6048−6063.

(21) Hoffend, C.; Schodel, F.; Bolte, M.; Lerner, H.-W.; Wagner, M.

Chem.-Eur. J. 2012, 18, 15394−15405.

(22) Seven, O.; Qu, Z.-W.; Zhu, H.; Bolte, M.; Lerner, H.-W.;

̈

Holthausen, M. C.; Wagner, M. Chem.-Eur. J. 2012, 18, 11284−11295.

(23) Januszewski, E.; Bolte, M.; Lerner, H.-W.; Wagner, M.

Organometallics 2012, 31, 8420−8425.

1

spectrometric characterization of 36. H and ¹³C NMR spectra

(24) Zhou, Z.; Wakamiya, A.; Kushida, T.; Yamaguchi, S. J. Am.

Chem. Soc. 2012, 134, 4529−4532.

1

for the intermediates 14−19. H and ¹³C NMR spectra in

(25) Hoffend, C.; Diefenbach, M.; Januszewski, E.; Bolte, M.; Lerner,

H.-W.; Holthausen, M. C.; Wagner, M. Dalton Trans. 2013,

DOI: 10.1039/C3DT51035B.

CDCl₃of the new compounds 5−13, 20−31, and 33−36. This

material is available free of charge via the Internet at http://

pubs.acs.org.

(26) (a) Yamaguchi, S.; Shirasaka, T.; Akiyama, S.; Tamao, K. J. Am.

Chem. Soc. 2002, 124, 8816−8817. (b) Wakamiya, A.; Mishima, K.;

Ekawa, K.; Yamaguchi, S. Chem. Commun. 2008, 579−581. (c) For a

2,3,4,5-tetrakis(2-thienyl)borole, see: Araki, T.; Fukazawa, A.;

Yamaguchi, S. Angew. Chem., Int. Ed. 2012, 51, 5484−5487.

(27) (a) Caruso, A.; Siegler, M. A.; Tovar, J. D. Angew. Chem., Int. Ed.

2010, 49, 4213−4217. (b) Caruso, A.; Tovar, J. D. J. Org. Chem. 2011,

76, 2227−2239. (c) Levine, D. R.; Caruso, A.; Siegler, M. A.; Tovar, J.

D. Chem. Commun. 2012, 48, 6256−6258.

AUTHOR INFORMATION

Corresponding Author

matthias.wagner@chemie.uni-frankfurt.de

■

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

We are grateful to the Sonderforschungsbereich/Transregio 49

project for ﬁnancial support of this work.

■

(28) Williams, V. C.; Dai, C.; Li, Z.; Collins, S.; Piers, W. E.; Clegg,

W.; Elsegood, M. R. J.; Marder, T. B. Angew. Chem., Int. Ed. 1999, 38,

3695−3698.

(29) Chen, J.; Kampf, J. W.; Ashe, A. J., III. Organometallics 2008, 27,

3639−3641.

REFERENCES

■

(30) Dou, C.; Saito, S.; Matsuo, K.; Hisaki, I.; Yamaguchi, S. Angew.

Chem., Int. Ed. 2012, 51, 12206−12210.

(1) Muller, T. J. J., Bunz, U. H. F., Eds. Functional Organic Materials:

̈

Syntheses, Strategies and Applications; Wiley-VCH: Weinheim, 2007.

(2) Klauk, H. Organic Electronics: Materials, Manufacturing and

Applications; Wiley-VCH: Weinheim, 2006.

(31) Lorbach, A.; Reus, C.; Bolte, M.; Lerner, H.-W.; Wagner, M.

Adv. Synth. Catal. 2010, 352, 3443−3449.

(32) Reus, C.; Liu, N.-W.; Bolte, M.; Lerner, H.-W.; Wagner, M. J.

Org. Chem. 2012, 77, 3518−3523.

(3) So, F. Organic Electronics: Materials, Processing, Devices and

Applications; CRC Press: Boca Raton, FL, 2010.

(33) Both in C₆D₁₂and in C₆D₆we detected traces of: (i) 1-bromo-

3-trimethylsilylbenzene: Stehouwer, J. S.; Jarkas, N.; Zeng, F.; Voll, R.

J.; Williams, L.; Owens, M. J.; Votaw, J. R.; Goodman, M. M. J. Med.

Chem. 2006, 49, 6760−6767. (ii) 1-Bromo-3,5-bis(trimethylsilyl)-

(4) Mullen, K.; Scherf, U. Organic Light Emitting Devices: Synthesis,

̈

Properties and Applications; Wiley-VCH: Weinheim, 2006.

(5) Brabec, C.; Scherf, U.; Dyakonov, V. Organic Photovoltaics -

Materials, Device Physics, and Manufacturing Technologies; Wiley-VCH:

Weinheim, 2008.

benzene: Beinhoff, M.; Karakaya, B.; Schluter, A. D. Synthesis 2003,

̈

(6) De Angelis, F.; Gaspari, M.; Procopio, A.; Cuda, G.; Di Fabrizio,

E. Chem. Phys. Lett. 2009, 468, 193−196.

79−90. These appeared already 1 h after the sample had been

prepared but did not increase significantly in amount with time.

O

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

We therefore assume that these compounds are primarily formed by

the reaction of 3 with small quantities of HBr present in BBr₃:

Seyferth, D.; White, D. L. J. Am. Chem. Soc. 1972, 94, 3132−3138. At

high temperatures and in the presence of excess BBr₃, 1-bromo-3-

trimethylsilylbenzene and 1-bromo-3,5-bis(trimethylsilyl)benzene are

finally transformed into 1-bromo-3-dibromoborylbenzene and 1-

bromo-3,5-bis(dibromoboryl)benzene.

(50) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.;

Springer: New York, 2006. The band gap values were calculated from

the onset wavelength (λ_onset) of the absorption spectra.

(51) Dawson, W. R.; Windsor, M. W. J. Phys. Chem. 1968, 72, 3251−

3260.

(52) Morris, J. V.; Mahaney, M. A.; Huber, J. R. J. Phys. Chem. 1976,

80, 969−974.

(53) Anthony, J. E. Chem. Rev. 2006, 106, 5028−5048.

(54) For the determination of the quantum yields of thin ﬁlms, at

least four ﬁlms of varying thicknesses were generated for each

compound by evaporation of diﬀerently concentrated samples in

CHCl₃at 100 °C on a quartz-glass surface. Using this experimental

setup and an integrating sphere system, the measured solid-state

quantum yield of a thin ﬁlm of the standard green OLED material

tris(8-hydroxyquinoline)aluminum (Alq₃) is in line with the published

(34) (a) Bruice, P. Y. Organic Chemistry, 5th ed.; Pearson Education,

Inc.: NJ, 2007. (b) Roberts, J. D.; Sanford, J. K.; Sixma, F. L. J.;

Cerfontain, H.; Zagt, R. J. Am. Chem. Soc. 1954, 76, 4525−4534.

(c) Coombes, R. G.; Crout, D. H. G.; Hoggett, J. G.; Moodie, R. B.;

Schofield, K. J. Chem. Soc. B 1970, 347−357.

(35) Kaufmann, D. Chem. Ber. 1987, 120, 901−905.

(36) A related thermal rearrangement reaction has been reported for

1-trimethylstannyl-2-dichloroborylferrocene: (a) Jakle, F.; Lough, A. J.;

̈

value (estimated error margin

Milius, W.; Hillebrecht, H.; Brutting, W. Adv. Funct. Mater. 2003, 13,

10%): (a) Colle, M.; Gmeiner, J.;

̈

Manners, I. Chem. Commun. 1999, 453−454. (b) Gamboa, J. A.;

̈

Sundararaman, A.; Kakalis, L.; Lough, A. J.; Jakle, F. Organometallics

̈

108−112. (b) Kumar, N. S. S.; Varghese, S.; Rath, N. P.; Das, S. J.

2002, 21, 4169−4181.

Phys. Chem. C 2008, 112, 8429−8437.

(37) (a) Muetterties, E. L. J. Am. Chem. Soc. 1959, 81, 2597.

(b) Muetterties, E. L. J. Am. Chem. Soc. 1960, 82, 4163−4166.

(55) Ref 26b.

(56) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7−13.

(57) Stoe & Cie. X-AREA. Diﬀractometer control program system; Stoe

& Cie: Darmstadt, Germany, 2002.

(c) Kolle, P.; Noth, H. Chem. Rev. 1985, 85, 399−418. (d) Piers, W.

̈

E.; Bourke, S. C.; Conroy, K. D. Angew. Chem., Int. Ed. 2005, 44,

5016−5036. (e) De Vries, T. S.; Prokofjevs, A.; Vedejs, E. Chem. Rev.

2012, 112, 4246−4282. (f) Bagutski, V.; Del Grosso, A.; Carrillo, J. A.;

Cade, I. A.; Helm, M. D.; Lawson, J. R.; Singleton, P. J.; Solomon, S.

A.; Marcelli, T.; Ingleson, M. J. J. Am. Chem. Soc. 2013, 135, 474−487.

(38) Ishida, N.; Moriya, T.; Goya, T.; Murakami, M. J. Org. Chem.

2010, 75, 8709−8712.

(58) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112−122.

(39) See ref 36b and also: Boshra, R.; Venkatasubbaiah, K.; Doshi, A.;

Lalancette, R. A.; Kakalis, L.; Jakle, F. Inorg. Chem. 2007, 46, 10174−

̈

10186.

(40) Wang, N.; Hudson, Z. M.; Wang, S. Organometallics 2010, 29,

4007−4011.

(41) These substituents have already been employed for the

development of donor−acceptor−donor triads featuring 9-boraﬂuor-

ene or dibenzo[b,f ]borepin acceptors. See refs 26 and 27b.

(42) Compounds related to 33−36 have appeared in the recent

patent literature as materials for organic electroluminescence devices:

(a) Meng, A. PCT Int. Appl. WO 2006/050496, 2006. (b) Stoessel, P.;

Heil, H.; Parham, A.; Vestweber, H. DE 10 2006/013802, 2006.

(43) Yang, C.; Jacob, J.; Mullen, K. Macromol. Chem. Phys. 2006, 207,

̈

1107−1115.

(44) (a) Karplus, M.; Pople, J. A. J. Chem. Phys. 1963, 38, 2803−

2807. (b) Gunther, H.; Schmickler, H.; Konigshofen, H.; Recker, K.;

̈

Vogel, E. Angew. Chem., Int. Ed. 1973, 12, 243−245. (c) Levin, R. H.;

Roberts, J. D. Tetrahedron Lett. 1973, 14, 135−138. (d) Farnum, D. G.

Adv. Phys. Org. Chem. 1975, 11, 123−175.

(45) Bock, B.; Kuhr, M.; Musso, H. Chem. Ber. 1976, 109, 1184−

1194.

(46) Iida, A.; Sekioka, A.; Yamaguchi, S. Chem. Sci. 2012, 3, 1461−

1466.

(47) Braunschweig, H.; Dyakonov, V.; Jimenez-Halla, J. O. C.; Kraft,

K.; Krummenacher, I.; Radacki, K.; Sperlich, A.; Wahler, J. Angew.

Chem., Int. Ed. 2012, 51, 2977−2980.

(48) Mercier, L. G.; Piers, W. E.; Parvez, M. Angew. Chem., Int. Ed.

2009, 48, 6108−6111.

(49) The inseparable compound mixtures 5/6, 20/21, and 29/30

have been included into the investigation of optoelectronic properties.

We are aware of the fact that an uncritical use of data obtained on

compound mixtures might give misleading results. However, a

comparison of the optoelectronic properties of pure 23 with those

of the isomeric mixture 23/24 revealed essentially no diﬀerences. Any

discussion of data obtained on 5/6, 20/21, and 29/30 is based on the

working hypothesis that the same would be true also in these cases.

Related investigations of photophysical properties of isomeric mixtures

of 2,9- and 2,10-substituted pentacenes can be found in the literature:

Yamada, H.; Yamashita, Y.; Kikuchi, M.; Watanabe, H.; Okujima, T.;

Uno, H.; Ogawa, T.; Ohara, K.; Ono, N. Chem.-Eur. J. 2005, 11, 6212−

6220.

P

dx.doi.org/10.1021/ja406766e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article Doi

C-functionalized, air- and water-stable 9,10-dihydro-9,10- diboraanthracenes: Efficient blue to red emitting luminophores

DOI: 10.1021/ja406766e

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja406766e

Full text of DOI:10.1021/ja406766e

Products guided by the article

R&D Labs maybe for 1449112-81-5

Relevant to this article

Hot Product