Novel fluorescent quater- and quinquifurans: Syntheses and photophysical properties

Article abstract of DOI:10.1002/jhet.5570390519

In the quest for fast fluors for use in waveshifting polystyrene fibers, symmetrical oligofurans were investigated. Furan moieties were coupled by means of the Ullmann Reaction or by palladium-catalyzed unsymmetrical coupling; the latter gave higher yields. While the benzoxazole-terminated quater- and quinquifurans we prepared were both stable and fast, exhibiting a green fluorescence and decay times of about 2.4 nsec, they were inferior to other types of fluors in solubility and emission intensity when incorporated into polystyrene.

Full text of DOI:10.1002/jhet.5570390519

Sep-Oct 2002
Novel Fluorescent Quater- and Quinquifurans: Syntheses and
Photophysical Properties
981
Joel M. Kauffman* and Guillermo Moyna
University of the Sciences in Philadelphia (USP), 600 South 43rd St., Philadelphia, PA 19104-4495
Received February 11, 2002
In the quest for fast fluors for use in waveshifting polystyrene fibers, symmetrical oligofurans were inves-
tigated. Furan moieties were coupled by means of the Ullmann Reaction or by palladium-catalyzed unsym-
metrical coupling; the latter gave higher yields. While the benzoxazole-terminated quater- and quinquifu-
rans we prepared were both stable and fast, exhibiting a green fluorescence and decay times of about 2.4
nsec, they were inferior to other types of fluors in solubility and emission intensity when incorporated into
polystyrene.
J. Heterocyclic Chem., 39, 981(2002).
Introduction.
benzoxazol-2-yl groups have been postulated [10]. Our
earlier efforts to obtain fast fluors by extending the conju-
gation of unsymmetrical ones, that is, those with push-pull
auxofluors, were not successful; for example, see fluor 16c
in ref. [8], and 11a-b in ref. [11]; the molecules would not
form an excited state utilizing the entire fluorophore,
which, in each, encompassed only 4 aromatic rings con-
For more than a decade fluors have been sought for use
as waveshifters in fiber light-guides. The plan was to col-
lect violet-to-blue light pulses from scintillations derived
from the impact of high-energy particles in polystyrene
(PS) tiles by means of imbedded circular fibers, and, by
means of waveshifting fluors dissolved in the fibers, to
change the light emitted from the fibers to green for detec-
tion with green-sensitive photomultiplier tubes (see Figure
3a in [1]). The requirements for such waveshifting fluors
have been detailed [2], and include high fluorescence
quantum yield (F ). Short fluorescence decay times (t )
f
f
have been inversely correlated with large extinction coeffi-
cients (e) for ultraviolet absorption in oligophenylenes [3];
however, the longest practical emission wavelength
obtainable, 390 nm, is from a sexiphenyl with no auxo-
fluors [4]. Carbon-bridging of some of the rings as in an
oligofluorene extends this wavelength to about 437 nm
[4,5]. A recent report on fluors which are agglomerations
of benzoxazoles (or benzothiazoles) and phenylene groups
gave fast fluors, but their solubilities and Stokes' shifts
were too low [6]. While the decay times for 1,6-diphenyl-
1,3,5-hexatriene (12 nsec) and 1,8-diphenyl-1,3,5,7-
octatetraene (6 nsec) are much too slow [7], and their pho-
tochemical stability is poor, their wavelengths were desir-
able; thus we thought that incorporation of the conjugated
polyene system into furan rings might provide molecules
with a more oligophenylene-like decay time. We had
already established that benzimidazolium groups as auxo-
fluors on a fluorene gave a bright fluor, but with too short
an emission wavelength [8]. It has long been known that
benzoxazoles as auxofluors on oligophenylenes or fluo-
renes give fluors that are photostable, bright and fast (t =
f
1-2 nsec), albeit of far too short emission wavelength [9],
so it seemed reasonable to prepare oligofurans terminated
with benzoxazoles for this work. The placement of the
strongly electron-attracting benzoxazol-2-yl group on the
2- and 5-positions of electron-releasing oligofurans was
expected to give reasonably stable compounds. The struc-
tures of the S state of symmetrical fluors terminated with
1

982
J. M. Kauffman and G. Moyna
Vol. 39
nected by single bonds. By contrast, fluors symmetrical
along their long axes can display high epsilons from as
many as 8 aromatic rings connected by single bonds [12].
However, such large and symmetrical molecules are also
high-melting and have limited solubility. For these reasons
we designed and made oligofurans 1 (Scheme 1) and 2
in Dibutyl Carbitol™ (DBC), rather than the commonly
used polyphosphoric acid, in the belief (not tested) that the
furans would not survive hot aqueous acid conditions.
Furan 4 was lithiated with t-butyllithium to give 5 and this
was converted to the 2-chlorozinc, as we reported previ-
ously [11], which was coupled with 3 by means of the air-
.
stable and under-utilized catalyst PdCl dppb [15] to give
2
bifuran 7 as a gum. Iodination gave a poor yield of 8,
1
which, at least, was well-characterized, including H and
13
C nmr spectra. An Ullmann coupling of 8 with copper
bronze gave quaterfuran 1 in good yield, whose structure is
1
supported by a H nmr spectrum and elemental assays.
The coupling method reported to form plain quaterfuran
from bifuran [13] would not have been applicable, since
the n-butyllithium employed would have reacted with the
C=N bond of the benzoxazole.
Quinquifuran 2 (Scheme 2). We had been successful in
using Seha's method [16] for synthesis of a 2-(2-hydrox-
yphenyl)benzoxaxole from 2-amino-4-t-butylphenol 9 in
one step [11], but we obtained from 5-bromo-2-furoic acid
9 mostly amide 10 . This with boric acid in DBC gave 11.
Dilithiation of the terfuran 12, exchange to the 2,5"-
di(chlorozinc), and coupling with 11 in the presence of
.
PdCl dppb gave quinquifuran 2, whose structure was con-
2
1
firmed by 1D and 2D (COSY) H nmr spectroscopy
(Figure 1).
Side-chains were incorporated for greater solubility
(Scheme 3). We had initially prepared a sort of lower
homolog of quaterfuran 1 from 6-amino-m-cresol by the
same method, but in very low yield, and poorly character-
ized. This was 6,6'"-bis(6-methyl-2-benzoxazolyl)-
2,2':5',2":5",2'"-quaterfuran (1a; this was 22b in ref. [17])
of mp 300-304° and t £ 2.5 nsec (see Table 1). Unlike
f
quaterfuran 1, 1a was too insoluble in styrene monomer
for practical use, as was quinquifuran 2. The latter, for
example, was too insoluble in toluene at 20° for us to
obtain UV and fluorescence spectra, which is why we used
-7
chloroform, in which the solubility was 2x10 M or 1.5
mg/L. So we prepared the intermediate 19 (Scheme 3)
with a dodecyl group to enhance the solubility of the corre-
sponding oligofurans in styrene, compared with the oligo-
furans derived from 3 and 11.
(Scheme 2) with alkyl groups to provide easier purfication
and higher solubility in PS.
There is a recent preparation of 4-dodecylphenol 16 in
40% yield by diazotization of 4-dodecylaniline [18], but
the latter was far too expensive for this project. The clas-
sic method of Clemmensen reduction, when applied to
4-dodecanoylphenol, was poorly described with no yields
given [19,20] and failed altogether in our hands. (In fact
the prior Fries rearrangement of the phenyl ester does not
We were unaware of any prior description of the synthe-
ses of any quater- and quinquifurans in the peer-reviewed
literature. After this work was completed, a single report
of the synthesis of unsubstituted quaterfuran appeared,
without detail or characterization [13].
Discussion and Results.
Syntheses.
give a clean 4-isomer [21]). So taking advantage of the
.
discovery that PdCl dppb catalyzed the coupling of
2
Quaterfuran 1 (Scheme 1). Condensation of 5-bromo-2-
furoic acid with 2-amino-4-ethylphenol to give 3 was car-
ried out by an old method [14] using boric acid as catalyst
methylmagnesium chloride with a 2-bromofluorene [22],
we prepared from 1-bromododecane 13 the Grignard
reagent 14 and coupled it with 4-iodoanisole (after finding

Sep-Oct 2002
Novel Fluorescent Quater- and Quinquifurans: Syntheses and Photophysical Properties
983
Figure 1. Optical Spectra of Quaterfuran 1 in Tolulene.
Table 1
Absorption and Fluorescence Properties of Oligofurans and Related Tetracne
Compound
Absorption (nm)
Found
Fluorescence em. (nm)
Found
(Calc) [a]
(—)
(Calc) [a]
(—)
F f
t (nsee)
f
(E,E,E,E)-1,8-Diphenyl-
1,3,5,7-octatetracne [b]
356, 373, 387
420,444
453,475,518*,556
472*,498,530
477*, 507
0.08 [e]
6.2 [g]
2.4 [h]
2.4 [h]
5,5"'-Bis(5-ethyl-2-benzoxazolyl)-
2,2':5',2":5",2"'-quaterfuran (1)[c]
5,5''''-Bis(5-t-butyl-2-benzox-
azolyl)-2,2':5',2":5",2"':5"',2''''-
quinquifuran (2) [d]
(439)
(448)
(478)
(488)
0.77 [f]
0.89 [f]
431,455
[a] As described in ref. [17]; [b] From ref. [3]; [c] The calculations were for 6,6"'-Bis(6-methyl-2-benzoxazolyl)-2,2':5',2":5",2"'-quaterfuran, 22b in ref.
[17]; [d] Samc as 23b in ref. [17]; [e] From [7] adjusted as in [4]; [f] Detd. in toluene as described [4] vs. DPA in toluene with F = 0.82; [g] From ref.
f
[3]; [h] The reciprocal of the slope of intensity vs. time of the fast decay component obtained by laser excitation of the sample; *Indicates the major peak.
that 4-bromoanisole was too unreactive) to give 4-dode-
cylanisole 15 in good yield. The classic ether cleavage with
hydrobromic acid was sluggish, but use of the co-solvents
propanoic acid and its anhydride, and a long reaction time
gave 16 in good yield. The recently reported nitration of
16 [18] failed entirely in our hands, but the use of diluted
nitric acid with the same co-solvent, propanoic acid, gave
nitro 17 in good yield. Instead of the reported use of

984
J. M. Kauffman and G. Moyna
Vol. 39
Raney nickel for the reduction [19] we used palladium on
carbon for an excellent yield of amine 18, which, with
5-bromo-2-furoic acid, boric acid and DBC, gave a low
yield of benzoxazole 19, which was well-characterized
by its H nmr spectrum. This sequence was not carried
further because the shortcomings of 1 and 2 had been
ization in styrene, but 1 partially precipitated on cooling
because of insolubility. Both were deficient in light output
compared with other fluors tested during the project
because of too little Stokes' shift, not because of their
quantum yields, which were high (Table 1). The extinction
coefficient of 1 (77,000) was significantly lower than that
of 2 (108,000), showing that the additional furan ring was
contributing. This De is very large for just one additional
aromatic ring. Perhaps this is due to the existence of an
additional plane of symmetry in 2, as shown by our draw-
ings of these structures in the putative lowest-energy con-
formation with all oxygens anti; or perhaps it is due to the
"quinquiphenyl effect" that has been noted [4]. The
Stokes' shift of 1 was much less than that of the almost iso-
electronic 1,8-diphenyl-1,3,5,7-octatetraene (Table 1); this
1
recognized by this time.
Photophysical Properties.
Both quaterfuran 1 and quinquifuran 2 had fast decay
times of 2.4 nsec, the fastest of any green-emitting fluors
of which we are aware. No decomposition in solution at
20° was observed in contrast to plain quaterfuran, which is
so unstable that spectra had to be determined on freshly
prepared solutions [13]. Both 1 and 2 survived polymer-
1
Figure 2. 1D and 2D (COSY) H nmr spectra of compound 2. Correlations between coupled protons are marked with boxes. See text and Scheme 2 for
assignments.

Sep-Oct 2002
Novel Fluorescent Quater- and Quinquifurans: Syntheses and Photophysical Properties
985
tane (120 mL, 0.18 mol, Acros 18128) was added during 1 hour
to form salt 5; this was exothermic to the last drop. After a 2-
hour period at 0° zinc chloride (24.48 g, 0.18 mol, Cerac Z-1007)
was heated with a gas flame (but not fused) at 0.5 torr, allowed to
cool and added to the solution of 5 in one portion. An exotherm
to 35° occurred with formation of an orange color; all was solu-
ble in 25 minutes, but was kept at 21° overnite to give 6.
means that there is much less conformational change of 1
on excitation.
Predictions of the major absorption and fluorescence
wavelength peaks of 1 and 2 by computerized calculations
[17] were reasonably accurate (Table 1), but the degree of
self-absorption, limiting the light output of these fluors in
PS, was not predicted.
.
The catalyst PdCl2 dppb (1.0 g, 0.0018 mol) was added, then
1/3 of a solution of 2-(5-bromo-2-furyl)-5-ethylbenzoxazole (6,
52.56 g, 0.18 mol) was introduced. Warm water in the bath pro-
voked an exotherm from 35-40° with brisk reflux. The rest of the
solution of E was added, which continued the exotherm, finally
to 53°, after which the bath was heated with a hot-plate for 3
hours more of reflux. The mixture turned black in 1 hour, indi-
cating completion of the reaction. The mixture was cooled in ice
to 10° and quenched in a mixture of 500 g ice, 1 L of water and
100 mL of ethanol with propellor stirring to give a suspension
that was filtered on a 12.5 cm Büchner funnel.
The filtrate in benzene was chromatographed on a 4.5 cm
diam. 30 cm high column of alumina using 1.2 L of benzene to
obtain, after evaporation, 41.7 g (83 % crude) of viscous oil (7),
which did not crystallize even after 1 year. Tlc on Whatman
MK6F Silica Gel with toluene showed the same Rf = 0.28 as 3.
On Analtech Alumina GF, toluene, 3 had Rf = 0.73, a non-fluo-
rescent spot that turned brown; while 7 had Rf = 0.63 with a pow-
erful violet fluorescence under longwave uv. The structure of 7
was confirmed by its conversion to 8 below.
EXPERIMENTAL
General.
Methods previously described [8] were followed with the
exception that the 300 MHz 1H nmr spectrum was determined on
a Varian 300 XL, and the 400 MHz spectra on a Bruker Avance
400, and 13C were recorded at 100 MHz on the latter. Lower-
case letters refer to both carbon and any hydrogen attached to that
carbon for nmr assignments. Dibutyl Carbitol™ (DBC) was
diethylene glycol dibutyl ether, Aldrich 20,562-1. Solvents used
in moisture-sensitive reactions were dried over 3A Molecular
Sieve. Preparation of the catalyst palladium(II) chloride 1,4-
.
bis(diphenylphosphino)butane (PdCl2 dppb) has been described
[15]. Florisil™ was Aldrich 22,074-4. Next to mp, pc means
phase-change, and s means softens.
2-(5-Bromo-2-furyl)-5-ethylbenzoxazole (3).
With magnetic stirring, 5-bromo-2-furoic acid (80.0 g, .419
mol, Aldrich B6,740-6), 2-amino-4-ethylphenol (57.5 g, 0.419
mol, [10]), and 1.0 g of powdered boric acid in 575 mL of DBC
were heated under a spray-trap/air-condenser combination, and
the mixture was heated strongly to distill out water. The reaction
temperature unexpectedly jumped to 275° with concurrent distil-
lation of 500 mL of DBC and 10.5 mL of water. At 65° 500 mL
of hexane was added, the liquid was decanted and cooled at -20°
overnight. The sticky solid was filtered, washed with 200 mL
hexane and dried at 50°/20torr/4hrs to give 87 g gummy solid,
which was mixed with 100 mL of acetone, 400 mL of water and
10 g of sodium carbonate monohydrate to give a solid from
which liquid was decanted. The solid was slurried with 100 mL
of 1:1 methanol:water and air-dried overnight. Now no longer
gummy, it was extracted from a 2-cm high column of Florisil in a
large Ace-Kau with 600 mL of 1,1,2-trichlorotrifluroethane for 4
days. Solvent was distilled and recovered from the extract. The
residual oil was taken up in 150 mL of 95 % ethanol, diluted with
» 30 mL of water to the cloud point, seeded and kept at 0°
overnight. The product was filtered, washed with 100 mL of 80
% ethanol, and dried as above to give 53.3 g (44%), mp 78-79°.
1H nmr: (400 MHz, DMSO-d6): d 1.23 (t, 3H, J = 7.6 Hz, Ha),
2.74 (q, 2H, J = 7.6 Hz, Hb), 6.95 (d, 1H, J = 3.7, Hc), 7.28 (dd,
1H, J = 1.6 Hz, J = 8.3 Hz, He), 7.47 (d, 1H, J = 3.7 Hz, Hd), 7.59
(dd, 1H, J = 0.5 Hz, J = 1.6 Hz, Hf), 7.65 (dd, 1H, J = 0.5 Hz, J =
8.3 Hz, Hg); 13C nmr (100 MHz, DMSO-d6): d 17.0, 29.0, 111.3,
115.8, 118.0, 119.4, 142.0, 142.1, 148.7, 154.5.
5-Ethyl-2-[5'-(5"-iodo-2-furyl)-2-furyl]benzoxazole (8).
The 5-ethyl-2-[5'-(2-furyl)-2-furyl]benzoxazole (7, 41.0 g,
0.147 mol) was dissolved in 300 mL of acetic acid with magnetic
stirring. Then 4.5 mL of sulfuric acid diluted first with 30 mL of
water was added, and the whole warmed to 45° on a mantle. A
mixture of iodine (15.0 g, 0.0589 mol, Fisher I-37) and periodic
acid (7.18 g, 0.0315 mol, Fisher A-223) was ground in a mortar,
then » 10 % was added to the solution of 7 at 45°, then another 10
% every 10 minutes, at 55°, then at 65°, then at 70° thereafter.
The reaction seemed to stop when » 70 % had been added, and
under long UV a weak blue-green fluorescence replaced a power-
ful violet fluorescence, all during an additional 1-hour period at
70°. The product mixture was decanted from » 5 g of gum into
600 mL each of ice and water to give a granular precipitate,
which was filtered, washed with 300 mL of 95 % ethanol and air-
dried overnight, becoming gummy in the process.
The gum was fractionally crystallized from methanol-acetone
mixtures at -20° in several portions to give about 13 g of which
the largest fraction had mp 128-133°. This orange solid was
extracted from a 4-cm high column of Florisil in a medium Ace-
Kau with 250 mL of cyclohexane. The extract at 21° was
decanted from the pure product, which was dried at 80°/20
torr/1hr to give 7.68 g (13 %) of rust-colored needles of 8, mp
1
139-141°; H nmr (400 MHz, DMSO-d6): d 1.24 (t, 3H, J = 7.6
Hz, Ha), 2.74 (q, 2H, J = 7.6 Hz, Hb), 6.92 (d, 1H, J = 3.6 Hz,
Hc), 6.94 (d, 1H, J =3.6 Hz, Hd), 7.03 (d, 1H, J = 3.8 Hz, He),
7.28 (dd, 1H, J = 1.6 Hz, J = 8.3 Hz, Hg), 7.53 (d, 1H, J = 3.8 Hz,
Hf), 7.60 (dd, 1H, J = 0.5 Hz, J = 1.6 HHz, Hh), 7.67 (dd, 1H, J =
0.5 Hz, J = 8.3 Hz, Hi); 13C nmr (100 MHz, DMSO-d6): d 17.0,
29.0, 95.1, 109.4, 111.3, 111.8, 117.6, 119.3, 123.3, 126.5, 141.8,
142.1, 142.3, 148.1, 148.9, 150.0, 155.2.
Anal. Calc. for C13H10BrNO2: C, 53.45; H, 3.45; Br, 27.35; N,
4.80 %. Found: C, 53.60; H, 3.27; Br, 27.18; N, 4.69.
5-Ethyl-2-[5'-(2-furyl)-2-furyl]benzoxazole (7).
Under argon, furan (4, 13.1 mL, 12.25 g, 0.18 mol, Acros
11977) in 200 mL of tetrahydrofuran was cooled below 0° by
means of an ice-acetone bath, then 1.5 M t-butyllithium in pen-
Anal. Calc. for C17H12INO3: C, 50.39; H, 2.99; I, 31.32; N,
3.46 %. Found: C, 50.32; H, 2.99; I, 31.19; N, 3.37.

986
J. M. Kauffman and G. Moyna
Vol. 39
5,5'"-Bis(5-ethyl-2-benzoxazolyl)-2,2':5',2":5",2'"-quaterfuran
Anal. Calc. for C15H14BrNO2: C, 56.27; H, 4.41; N, 4.38 %.
(1).
Found: C, 56.66; H, 4.73; N, 4.76.
Because of a peculiarity in the H nmr spectrum, the above
1
Copper bronze (3.20 g, 0.0500 mol, Aldrich 29,258.3) and 5-
ethyl-2-[5'-(5"-iodo-2-furyl)-2-furyl]benzoxazole (8, 7.30 g,
0.0180 mol) in 75 mL of dry N,N-dimethylformamide was heated
with rapid magnetic stirring at 80°/18 hours, then at reflux (151°)
for 1 hour. At 21° the solids were filtered, washed with 50 mL of
acetone, and dried at 100°/20torr/1hour to give 8.6 g, which was
extracted from a 2-cm high column of Florisil in a medium Ace-
Kau with 125 mL of tetrahydrofuran for 2 days to give a total of
2.85 g (57 %) of orange-yellow microneedles, all of which was
extracted again in the same manner for 18 hours to give 2.48 g
(50%), mp 285-286.5°; uv (toluene, 2.3x10-5 M): 420 nm
(77,000), 444 (55,000); fluorescence emission of same solution:
472 (79 rel int), 498 (100), 530 sh (57), see Figure 1; t f = 2.4
material was further purified by extraction from Merck 10181
Silica Gel in an Ace-Kau with hexanes. The extract was kept at -
20° to give two fractions. The latter, mp 103-105°, showed H
nmr (major conformer, 400 MHz, DMSO-d6): d 1.35 (s, 9H,
(CH3)3-C-), 6.86 (d, 1H, J = 3.7 Hz, He), 7.49 (dd, 1H, J = 1.8
Hz, J = 8.6 Hz, Hc), 7.51 (d, 1H, J = 3.7 Hz, Hd), 7.65 (dd, 1H,
J = 0.5 Hz, J = 8.6 Hz, Hb), 7.74 (dd, 1H, J = 0.5 Hz, 1.8 Hz, Ha).
1
2,2':5',2"-Terfuran (12).
Some 1,4-di(2-furyl)-1,4-butanedione was prepared as
described from furfural and divinylsulfone by the method of El-
Hajj [23] in 18% yield, mp 130-133°. It was not clear which cat-
alyst had been used; we chose 3-ethyl-5-(2-hydroxyethyl)-4-
methylthiazolium bromide, Aldrich 33,124-4. The terfuran 12
was obtained from the dione by the method of Asano [24] in 52%
yield, mp 59-63°; this was raised to 64-65° after extraction of the
crude from Merck 10181 Silica Gel in an Ace-Kau with 1,1,2-
trichlorotrifluoroethane (lit. 62-63° [24]).
1
nsec; H nmr (400 MHz, satd. in DMSO-d6 at 70°): d 1.28 (t,
3H, J = 7.5 Hz, Ha), 2.78 (q, 2H, J = 7.5 Hz, Hb), 7.09 (d, 1H, J =
3.8 Hz, Hc), 7.15 (d, 1H, J = 3.7 Hz, He), 7.17 (d, 1H, J = 3.8 Hz,
Hd), 7.31 (dd, 1H, J = 1.6 Hz, J = 8.3 Hz, Hg), 7.57 (d, 1H, J = 3.7
Hz, Hf), 7.62 (d, 1H, J = 1.6 Hz, Hi), 7.68 (d, 1H, J = 8.3 Hz, Hh).
Anal. Calc. for C34H24N2O6: C, 73.37; H, 4.35; N, 5.03 %.
Found: C, 73.12; H, 4.36;N, 4.98.
5,5""-Bis(5-t-butyl-2-benzoxazolyl)-2,2':5',2":5",2'":5'",2""-
quinquifuran (2).
N-(2-hydroxy-5-t-butylphenyl)-5-bromofuran-2-carboxamide
(10).
First, terfuran (12, 3.00 g, 0.015 mol) in 200 mL of tetrahydro-
furan was cooled to -5°, then 1.5 M t-butyllithium in pentane (20
mL, 0.030 mol, Acros 18128) was added below 0°. The dilithium
derivative separated and the suspension was kept in ice for 2.15
hours before the addition of solid, fused, anhydrous zinc chloride
(4.10 g, 0.030 mol, Cerac). The exotherm to 6° was followed by
heating to 30° during 30 minutes. Then 2-(5-bromo-2-furyl)-5-t-
butylbenzoxazole (11, 9.6 g, 0.030 mol) was added, then
Under argon a mixture of 5-bromo-2-furoic acid (19.1 g, 0.100
mol, Acros 10702), 100 mL of dioxane, thionyl chloride (10.9
mL, 17.85 g, 0.150 mol, Aldrich 23,046-4), and 1 mL of N-
methylpyrrolidinone were heated to reflux, which began at 98°,
and increased over 1.75 hours to 104°, after which 20 mL of
dioxane was distilled out to remove the excess thionyl chloride
from the acid chloride.
Separately, under argon, a solution of 2-amino-4-t-butylphenol
(9, 16.5 g, 0.100 mol, Acros 18583), pyridine (8.5 mL, 8.31 g,
0.105 mol) and 75 mL of dioxane was treated with the acid chlo-
ride solution above below 28°. The mixture was kept for 3 days,
diluted with 20 mL of water, then quenched in 500 mL of water.
The solid product was collected, washed with 50% methanol, then
slurried in 200 mL of the same for 10 minutes, and filtered, and
dried at 100°/30 torr/2 hours to give 32.32 g (96%), mp 203-206°;
ir (2% in CHCl3): 3600-2400 (br, intra-bonded O-H), 3400 (N-H
str), 3000 w, 2958 (-CH3), 2900 w, 2865 w, 1710, 1642 (C=O),
1600, 1587, 1528, 1503, 1467, 1313, 1243, 1200, 1112, 1010 cm-1.
The structure of 10 was confirmed by its conversion to 11
below.
.
PdCl2 dppb (0.30 g, 0.005 mol). The reactor was moved to a man-
tle and heated, to cause an exotherm to reflux at 57°; heat was
stopped, 200 mL of methanol was added, and the mixture was
stirred overnight. The precipitated yellow solid was collected,
washed with methanol and dried to give 7.43 g of orange solid,
mp pc 310-320°, 335-337° dec. This solid was extracted from a
soxhlet with 200 mL of tetrahydrofuran. The extract was kept at
20° overnight, collected, washed with acetone, and dried to yield
1.58 g, which was extracted from 1 cm of 35-70 mesh Silica Gel
in a small Ace-Kau with 3-chlorotoluene to give a total of 1.5 g
(15%) of 2, mp 341-344°; uv (2.2x10-7 M in CHCl3): 431 nm
(108,000), 455 sh (80,000); fluorescence emission of same solu-
1
tion: 477 nm (100), 507 (80); t f = 2.4 nsec; H nmr (400 MHz,
satd. in DMSO-d6 at 70°): d 1.38 (s, 18H, (CH3)3C-), 7.07 (d, 1H,
J = 3.8 Hz, Hh), 7.08 (s, 1H, Hg), 7.15 (d, 1H, J = 3.7 Hz, He),
7.17 (dd, 1H, J = 3.8 Hz, Hf), 7.50 (dd, 1H, J = 1.8 Hz, 8.7 Hz,
Hc), 7.57 (d, 1H, J = 3.7 Hz, Hd), 7.68 (dd, 1H, J = 0.6 Hz, J= 8.7
Hz, Hb), 7.77 (dd, 1H, J = 0.6 Hz, J = 1.8 Hz, Ha); see Fig. 2.
Anal. Calc. for C42H34N2O7: C, 74.32; H, 5.05; N, 4.13 %.
Found: C, 74.11; H, 5.16; N, 4.19.
2-(5-Bromo-2-furyl)-5-t-butylbenzoxazole (11).
With magnetic stirring, N-(2-hydroxy-5-t-butylphenyl)-5-bro-
mofuran-2-carboxamide (10, 9.07 g, 0.0268 mol) and 0.20 g of
powdered boric acid in 75 mL of DBC were heated strongly
under a spray-trap/air-condenser combination to distill out water.
After allowing this to cool, most of the DBC was removed by dis-
tillation at 96°/0.7 torr. The residue was covered with 50 mL of
methanol and kept at -20° overnight to give 5.66 g of 11, mp 98-
100°, which was extracted from 4 cm of Silica Gel in a medium
Ace-Kau with 200 mL of 1,1,2-trichlorotrifluroethane. The
extract was evaporated and the residue recrystallized from 50 mL
of methanol at -20° to give tan spars, mp 100-103°, 3.73 g (43%).
This was extracted from 2 cm each basic and neutral alumina in a
medium Ace-Kau with 75 mL of 1,1,2-trichlorotrifluroethane to
give 3.2 g (37%), mp 102-103°.
4-Dodecylanisole (15).
From 1-bromododecane (13, 123 mL, 127 g, 0.513 mol, Acros
10691) and magnesium (12.48 g, 0.513 mol, Reade RMC-3) in
1.5 L of tetrahydrofuran, the dodecylmagnesium bromide 14 was
prepared; this involved a 1-hour addition of 13 and 2.5 hours of
.
reflux. When the mixture cooled down to 62°, PdCl2 dppb (3.1 g,
0.00513 mol) was added, followed by 4-iodoanisole (90.0 g,
0.385 mol) in multi-gram portions. A mild exotherm from

Sep-Oct 2002
Novel Fluorescent Quater- and Quinquifurans: Syntheses and Photophysical Properties
987
66-69° to reflux was noted, and reflux was maintained overnight.
Separately a solution of ammonium chloride (8.0 g, 0.15 mol) in
25 mL of water was added at 44°; this caused an exotherm to 52°.
After 3 days at 20° the liquid was decanted, evaporated, evapo-
rated again with 300 mL of toluene, the residual oil taken up in
300 mL of hexane, some solid removed, the hexane evaporated,
and the product distilled at 175-195°/0.45 torr to give 61.4 g
chromatographed on Merck 10181 Silica Gel with cyclohexane
and benzene to give an oil which was crystallized from ethanol at
-20°, then twice from 1,1,2-trichlorotrifluoroethane at -20° to
give 3.76 g (12%), mp 64.5-66°; 1H nmr (10% in CCl4, 300
MHz): d = 0.90 (t, 3H, J = 5 Hz, CH3(CH2)9CH2CH2-); 1.31 (m,
18H, CH3(CH2)9CH2CH2-); 1.68 (m, 2H, CH3(CH2)9CH2CH2-);
2.73 (t, 2H, J = 5 Hz, CH3(CH2)9CH2CH2-); 6.549 (d, 1H, J =
3.6 Hz, He); 7.168 (1H, dd, J = 0.6, 8.2 Hz, Hd); 7.185 (1H, d, J
= 3.6 Hz, Hc); 7.434 (1H, d, J = 8.2 Hz, Hb); 7.534 (1H, d, J = 0.6
Hz, Ha).
1
(58%) of 15, fp 23°; H nmr (2.5% in DMSO-d6, 400 MHz): d
0.86 (t, 3H, J = 7.0 Hz, CH3(CH2)9CH2CH2-), 1.24 (m, 18H,
CH3(CH2)9CH2CH2-) 1.51 (m, 2H, 2H, CH3(CH2)9CH2CH2-
), 2.49 (t, 2H, J = 7.6 Hz, CH3(CH2)9CH2CH2-), 3.71 (s, 3H,
Ha), 6.82 (AA'XX', 2H, Hb), 7.08 (AA'XX', 2H, Hc); 13C NMR
(100 MHz, 2.5% in DMSO-d6): d 14.8, 22.9, 29.4, 29.6, 29.7,
29.8, 29.9, 32.1, 32.2, 35.1, 114.5, 129.9, 135.0, 158.1. The struc-
ture of 15 was further confirmed by its conversion to 16, 17 and
18 below.
Anal. Calc. for C23H30BrNO2: C, 63.89; H, 6.99; N, 3.24 %.
Found: C, 64.43; H, 6.61; N, 3.43.
Acknowledgement.
Financial support was provided under an SBIR Phase I con-
tract from the U. S. Department of Energy, #DE-FG03-
97ER82419/A000, "Development of Scintillators and
Waveshifters for Detection of Ionizing Radiation" and Phase II,
#DE-FG03-97ER82419, "Development of Scintillators and
Waveshifters for Detection of Ionizing Radiation and of
Manufacturing Methods to Improve Aging Properties", Charles
R. Hurlbut, P. I., Ludlum Measurements, Inc., Sweetwater, TX
79556. Online literature searches and editing of manuscript were
the effort of Leslie Ann Bowman; UV spectra were determined
by Farzad Kobarfard, some fluorescence spectra by Dwight
Crawford; the F f values were determined by Robert Swoboda
under the direction of Michael Bruist, and the first bifuran of type
7 was prepared by Peter T. Litak; all at USP. The 1H nmr spec-
trum of 19 was run and interpreted by Walter J. Boyko,
Department of Chemistry, Villanova University. James Bishop
determined the decay times under the direction of Randall C.
Ruchti, Department of Physics, University of Notre Dame.
Discussions with Charles R. Hurlbut (who also prepared
Figure 1), Randall C. Ruchti, Walter M. F. Fabian, and Charles J.
Kelley (the latter pair as consultants in the program) were very
valuable.
4-Dodecylphenol (16).
A mixture of 4-dodecylanisole (15, 61.3 g, 0.222 mol) with
150 mL of 48% hydrobromic acid and 135 mL of propanoic acid
was boiled under reflux for 6 hours, then cooled to -20°; the solid
was isolated by decantation, triturated with 150 mL of cold 50%
methanol, filtered and air-dried. This was subjected to the acid
treatment again; but, because the initial reflux temperature was
only 117°, 50 mL was distilled out, then 22 mL of propanoic
anhydride was added, and the mixture was boiled under reflux
(119°) for 50 hours. At 40° the upper organic layer was diluted
with an equal volume of methanol and held at -20° to crystallize
the product, 68 g, mp s. 40°, 55-62°. This was recrystallized
from 300 mL of methanol at -20° to give white plates, 41.2 g
(71%), mp s. 55°, 66-68°, (lit. mp 64-65° [20], 40%, mp 65°
[18]).
4-Dodecyl-2-nitrophenol (17).
Nitration of 4-dodecylphenol (16, 30.0 g, 0.109 mol) was
accomplished by suspending it in 200 mL of propanoic acid and
adding a mixture of 20 mL each of 70% nitric acid and water
below 5°, and stirring for an hour at 5°. The resulting homoge-
neous mixture was diluted with 600 mL of water and extracted
with 300 mL of ether. The extract in 1200 mL of water was
treated with excess sodium bicarbonate, separated, washed with
200 mL of water, dried over magnesium sulfate, and evaporated
to give 51 g of orange solid, which was recrystallized from 200
mL of pentane at -20° to give 26.2 g (75%), mp 53-55°; (lit. mp
55-56° [19], 82%, mp 54° [18]).
REFERENCES AND NOTES
[1] D. Green, Radiat. Phys. Chem., 41 (1/2), 19 (1993).
[2] J. M. Kauffman, Radiat. Phys. Chem., 41 (1/2), 365
(1993).
[3] I. B. Berlman, H. O. Wirth and O. J. Steingraber, J.
Phys. Chem., 75 (3), 318 (1971).
[4] J. M. Kauffman, P. T. Litak , J. A. Novinski, C. J.
Kelley, A. Ghiorghis and Y. Qin, J. Fluorescence, 5, 295 (1995).
[5] C. J. Seliskar, D. A. Landis, J. M. Kauffman, M. A.
Aziz, R. N. Steppel, C. J. Kelley, Yuanxi Qin, A. Ghiorghis,
Laser Chem. 13, 19, (1993).
2-Amino-4-dodecylphenol (18).
Reduction in a Parr hydrogenator beginning with 50 psig
hydrogen pressure was carried out on 4-dodecyl-2-nitrophenol
(17, 31.9 g, 0.100 mol) in 800 mL of tetrahydrofuran with 13.2 g
of recovered 10% Pd/C catalyst overnight. After removal of the
catalyst and solvent, the tan plates were recrystallized from 150
mL of methanol at -20° to give 27.2 g (94%), mp s. 105°, 119-
126°; (lit. 125-126° [19]).
[6] Y.-H. So, J. M. Zaleski, C. Murlick and A. Ellaboudy,
Macromolecules, 29, 2783 (1996).
[7] I. B. Berlman, J. Phys. Chem., 77 (4), 562 (1973).
[8] J. M. Kauffman, A. Khalaj, P. T. Litak , J. A. Novinski
and G. Bajwa, J. Heterocyclic Chem., 31 957 (1994).
[9] A. Reiser, L. J. Leyshon, D. Saunders, M. V. Mijovic,
A. Bright and J. Bogie, J. Am. Chem. Soc., 94, 2414 (1972).
[10] J. M. Kauffman and G. S. Bajwa, J. Heterocyclic
Chem., 30, 1613 (1993).
[11] (The Late) J. H. Hall, J. Y. Chien, J. M. Kauffman, P. T.
Litak. J. K. Adams, R. A. Henry and R. A. Hollins, J.
Heterocyclic Chem., 29,1245 (1992).
2-(5-Bromo-2-furyl)-6-dodecylbenzoxazole (19).
The method used for 11 was applied to 5-bromo-2-furoic acid
and 18 to give a solid (crude 19) from methanol. This in hexane
was passed through alumina, followed by more hexane to give an
eluate, which was evaporated, and the residual oil was recrystal-
lized from ethanol at -20° to give 21%, mp 57-61°. This was

988
J. M. Kauffman and G. Moyna
Vol. 39
[12] P. Leggate and D. Owen, Molecular Crystals, 4, 357
(1968).
[13] J. S. de Melo, F. Elisei, C. Gartner, G. G. Aloisi and R.
S. Becker, J. Phys. Chem. A, 104, 6907 (2000).
[14] I. Rubel, Optical Brighteners, Noyes Data Corp., Park
Ridge, NJ 07656, 1972, p23.
[15] C. J. Kelley, (the late) K. Ansu, (the late) W.
Budisusetyo, A. Ghiorghis, Y. Qin and J. M. Kauffman, J.
Heterocyclic Chem. 38, 11 (2001).
[19] D. Woodcock, J. Chem. Soc., 4391 (1955).
[20] F. Ueda, Chem. Pharm. Bull. Japan, 7, 823 (1959).
[21] A. W. Ralston, M. R. McCorkle and S. T. Bauer, J. Org.
Chem., 5, 645 (1940).
[22] C. J. Kelley, A. Ghiorghis and J. M. Kauffman, J.
Chem. Res. (S), 446-7 (1997); J. Chem. Res. (M), 2701-2733
(1997).
[23] T. El-Hajj, J.-C. Martin and G. Descotes, J.
Heterocyclic Chem., 20, 233 (1983).
[16] Z. Seha and C. D. Weis, Helv. Chim. Acta, 63, 413
(1980).
[24] T. Asano, S. Ito, N. Saito and K. Hatakeda,
Heterocycles, 6, 317 (1977).
[17] W. M. F. Fabian and J. M. Kauffman, J. Luminescence,
85, 137-148 (1999).
[25] I. B. Berlman, Handbook of Fluorescence Spectra of
Organic Molecules, 2nd Ed., Academic Press, NY, NY, 1971, p
323.
[18] C. Soulié and J. Simon, New J. Chem., 17, 267 (1993).