Practical and efficient synthesis of tris(4-formylphenyl)amine, a key building block in materials chemistry

Article abstract of DOI:10.1055/s-2005-865336

A short, practical and efficient preparation of tris(4-formylphenyl)amine, a key building block in materials chemistry, is described. It involves a two-flask synthesis from triphenylamine, which requires shorter overall time than the direct one-flask three-fold Vilsmeier-Haack formylation, while giving higher yields. The reaction levels off at the disubstitution stage, due to the deactivation of the bis-iminium intermediate, but hydrolysis of the latter into a less deactivated dialdehyde allows the third formylation to occur. The simple experimental protocol makes this method more convenient than the previously reported procedures. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2005-865336

PAPER
1771
Practical and Efficient Synthesis of Tris(4-formylphenyl)amine, a Key
Building Block in Materials Chemistry
P
T
ractical
a
nd Ef
h
ficient
S
ynth
o
esis of Tris(
m
an e s Mallegol, Said Gmouh, Mohamed Aït Amer Meziane, Mireille Blanchard-Desce, Olivier Mongin*
Synthèse et ElectroSynthèse Organiques (CNRS, UMR 6510), Université de Rennes 1, Campus de Beaulieu,
Bât. 10A, 35042 Rennes Cedex, France
Fax +33(1)223236955; E-mail: olivier.mongin@univ-rennes1.fr
Received 26 January 2005
was given.11 Finally, the two last pathways, involving a
Abstract: A short, practical and efficient preparation of tris(4-
formylphenyl)amine, a key building block in materials chemistry, is
described. It involves a two-flask synthesis from triphenylamine,
which requires shorter overall time than the direct one-flask three-
8
three-step sequence (based on triple bromination, cyana-
tion and controlled reduction) or a four-step sequence16
(
based on monoformylation, double bromination, protec-
fold Vilsmeier–Haack formylation, while giving higher yields. The tion of the aldehyde, and double bromine-lithium ex-
reaction levels off at the disubstitution stage, due to the deactivation change with t-BuLi followed by treatment with DMF), are
of the bis-iminium intermediate, but hydrolysis of the latter into a
difficult to scale up (Scheme 1).
less deactivated dialdehyde allows the third formylation to occur.
The simple experimental protocol makes this method more conve-
nient than the previously reported procedures.
A shorter, simple and reliable synthesis of 1 was therefore
required. Despite that the Vilsmeier–Haack formylation
of triphenylamine was low-yielding in 1, 4-formyltriphe-
nylamine (3) could be obtained using a single equivalent
Key words: materials chemistry, building block, aldehydes, elec-
trophilic aromatic substitutions, Vilsmeier–Haack reaction
of POCl in 94% yield, and 4,4¢-diformyltriphenylamine
3
(
4) was reported to be obtained in 71% yield using 10
24
equivalents of POCl3. It was obvious that the introduc-
tion of electron-withdrawing groups on the triphenyl-
amine deactivates the system too much for allowing the
third substitution to proceed significantly. The reactions
levels off at the disubstitution stage and competes with
other degradation pathways.
The popularity of tris(4-formylphenyl)amine (1; Figure 1)
as a building block is rapidly growing in materials chem-
istry. It is used for the synthesis of many materials, with
various applications, including electrophotographic pho-
1
,2
3–5
toreceptors, hole transport materials, TiO electrode
2
6
sensitizers for solar cells, polymers for non-doping emit-
ting materials, dendrimers, octupoles for second har- Our basic idea was that under the Vilsmeier–Haack con-
monic generation (SHG),
branched and dendritic chromophores for two-photon ab- 4a) leading to dialdehyde 4 were even more deactivated
sorption (TPA).
7
8,9
1
0,11
as well as octupolar, multi- ditions, the intermediates (as for example the bisiminium
1
2–22
than 4, and that, consequently, a hydrolysis at this stage
should be performed, before attempting the third substitu-
tion (Scheme 2).
OHC
CHO
We describe in this paper the optimization of the synthesis
of dialdehyde 4 and the investigation of the reaction con-
ditions for the conversion of 4 into 1.
N
Repeating the same conditions as in ref.24 (Table 1, entry
CHO
1
3
), we were not able to reproduce the reported results (i.e.
% of 3 and 71% of 4). Indeed, in our hands, the main
1
Figure 1
product was the monoaldehyde 3, along with a smaller
amount of the expected dialdehyde 4. Using larger excess
However, there was until now, no practical and efficient of phosphorus oxychloride and longer reaction times led
synthesis of 1 described in the literature. The shortest to the full conversion of triphenylamine (2) into 4 and 1
pathway could be the direct threefold Vilsmeier–Haack (entries 2 and 3), but also to degradation. It should be not-
formylation of triphenylamine (2), but regardless of some ed that increasing the temperature from 95 °C (entry 2) to
3
,23
erroneous patents, the reported yields remained finally 105 °C (entry 3) does not lead to an increase in the yield
4
,7
low, ranging between 6–7% and 18% (using 40 equiv of of 1. In addition, the yield of 4 significantly decreases, the
2
4
POCl3 at 95–100 °C for 48 h). Tris(4-formylphe- degradation processes being favored. Finally, an impor-
/DMF ratio, and
nyl)amine (1) could also be prepared from tris(4- tant parameter was found to be the POCl
3
lithiophenyl)amine and DMF, but no procedure or yield comparison between entries 2 and 4 is illustrative for that
purpose.
SYNTHESIS 2005, No. 11, pp 1771–1774
x
x.
x
x
.
2
0
0
5
Advanced online publication: 25.04.2005
DOI: 10.1055/s-2005-865336; Art ID: P01405SS
Georg Thieme Verlag Stuttgart · New York
©

1
772
T. Mallegol et al.
PAPER
POCl3, DMF
1
ref. 4: 7%
ref. 7: 6.1%
ref. 24: 18%
Br
Br
NC
CN
N
Br2
N
N
KCN
ref. 8: 40%
DIBAL-H
ref. 8: 45%
1
ref. 8: 94%
2
Br
CN
Br
Br
Br
HOCH2CH2OH
PTSA
Br
1
) t-BuLi
POCl3, DMF
N
Br2
N
N
2) DMF
1
ref. 24: 94%
Ref. 16: 95%
ref. 16: 50%
CHO
CHO
O
O
3
Scheme 1
Indeed, a slight variation of this ratio from 21:25 to 25:23
reduced the degradation very significantly, and the ex-
pected dialdehyde 4 could thus be obtained in 81% yield
along with 10% of trialdehyde 1 (Table 1, entry 4).
N
H
aqueous
workup
N
POCl3, DMF
The conversion of dialdehyde 4 into the trialdehyde 1 un-
der Vilsmeier–Haack conditions was then examined. It
was first found that the reaction temperature should be
kept below 100 °C and that the best reaction time was 1.5
hours. For longer reaction times, the slight increase in
conversion was counterbalanced by a parallel increase of
the degradation. The optimization of the amounts of phos-
phorus oxychloride and DMF is summarized in Table 2.
We started with a fixed amount of 100 equivalents of
2
4a
H
N
OHC
N
POCl3, DMF
1
POCl (Table 2, entries 1–4), varying the quantity of
3
DMF; reduction of the excess of DMF led to very signifi-
CHO
cantly improved conversion rates and yields of 1, together
with lowering of the degradation. For a fixed POCl /DMF
4
3
ratio, decreasing the amount of phosphorus oxychloride,
from 100 equivalents to 50 equivalents (entry 5 vs. entry
Scheme 2
3
, and entry 6 vs. entry 4) did not decrease the conversion
Table 1 Direct Vilsmeier–Haack Formylation of Triphenylamine
OHC
OHC
CHO
N
N
N
N
POCl3, DMF
+
+
CHO
CHO
CHO
2
3
4
1
Entry
POCl (equiv)
DMF (equiv)
Reaction temp (°C) Time (h)
Isolated Yields (%)
3
3
4
1
1
2
3
4
10.5
21
25
25
25
23
95
95
6
15
15
4
57
0
19
49
28
81
0
16
17
10
21
105
95
0
25
1
Synthesis 2005, No. 11, 1771–1774 © Thieme Stuttgart · New York

PAPER
Practical and Efficient Synthesis of Tris(4-formylphenyl)amine
1773
rate. By finely tuning the POCl /DMF ratio around 1 (en- yields. The simple experimental protocol makes this
3
tries 6 and 7), we found that the best results were obtained method more convenient than the previously reported pro-
for a ratio slightly higher than 1, i.e. by using less DMF cedures.
than phosphorus oxychloride. Finally, by using such a ra-
tio (entries 7–9), the quantity of Vilsmeier–Haack reagent
could be reduced to about 20–25 equivalents (such as for
entry 8), allowing minimal degradation while keeping the roy CM4000 HPLC instrument using a RP Nova-Park C18 column;
Phosphorus oxychloride and DMF were dried and distilled prior to
use. Reversed phase HPLC analyses were obtained with a Milton-
yield of 1 at 94%. When the amount of reagent was further eluent: CH CN–aqueous buffer solution (60:40, 20 mM AcONH ,
3
4
1
% AcOH); 1 mL/min; 30 °C; UV detection at 254 nm; retention
reduced to below 20 equivalents, the mixture became
times: t (1) = 2.79 min, t (4) : 4.37 min, t (3) = 8.60 min, t (2) =
highly viscous, preventing stirring (keeping in mind that
R
R
R
R
the reactions were performed without any solvent),25 and
19.66 min. Column chromatography was performed with Merck sil-
ica gel Si 60 (40–63 mm, 230–400 mesh). Melting points were de-
termined on an Electrothermal IA9300 digital melting point
a decrease of the conversion rate was observed (entry 9).26
instrument. NMR spectra were taken on a Bruker ARX 200 spec-
Table 2 Investigation of the Conversion of Dialdehyde 4 into Tri-
aldehyde 1
1
trometer in CDCl solutions; H chemical shifts (d) are given in ppm
3
relative to TMS as internal standard, and J values are given in Hz.
OHC
OHC
CHO
Investigation of the Conversion of 4 into 1; General Procedure
Phosphorus oxychloride and DMF were added into 20 mL vials
containing micro stir bar. Dialdehyde 4 (50 mg, 0.166 mmol) was
added, and the mixture was stirred at 95 °C. The reactions were
monitored by HPLC after 30 min, 45 min, 1 h, 1.5 h, 3 h and 5 h.
The best results were obtained for 1.5 h reaction time.
POCl3, DMF
N
N
95 °C, 1.5 h
CHO
CHO
4
1
Tris(4-formylphenyl)amine (1); Two-Flask Procedure from
Triphenylamine (2)
Phosphorus oxychloride (9.5 mL, 101.9 mmol) was added dropwise
at 0 °C under Ar to DMF (7.26 mL, 93.8 mmol) and the reaction
mixture was stirred for 1 h. Triphenylamine (1.00 g, 4.08 mmol)
was added, and the resulting mixture was stirred at 95 °C for 4 h.
After cooling to r.t., the mixture was poured into ice-water (200
Entry
POCl3
DMF
(equiv)
Unreacted 4 Yield of 1
a
(%)a
28
40
56
82
63
82
95
94
83
(
equiv)
(%)
34
26
20
12
25
16
3
1
2
3
4
5
6
7
8
100
100
100
100
50
800
400
200
110
100
55
mL), and basified with 1 M NaOH. After extraction with CH Cl
2
2
(
200 mL), the organic layer was washed with water (3 × 50 mL),
dried (Na SO ) and filtered over a short pad of silica gel. The sol-
2
4
vent was then removed under reduced pressure and the residue was
added on an ice-cooled mixture of POCl (7.6 mL, 81.5 mmol) and
3
DMF (5.78 mL, 74.6 mmol). The resulting mixture was stirred at 95
50
°
C for 1.5 h, and after cooling to r.t., poured into ice-water (200
mL), and basified with 1 M NaOH. After extraction with CH Cl
2
50
49
2
(
200 mL), the organic layer was washed with water (3 × 50 mL) and
25
23
6
dried (Na SO ). After evaporation of the solvent, the crude product
2
4
was purified by column chromatography (CH Cl ) to yield 1 (0.690
2
2
9
5
4.9
15
g, 52%, HPLC purity >99%) as a yellow solid; mp 233–235 °C
2
4
(
Lit. 230–232 °C).
a
Yields determined by HPLC analysis.
1
H NMR (200.13 MHz, CDCl ): d = 9.95 (s, 3 H), 7.85, 7.26
3
(
AA¢XX¢, JAX = 8.6 Hz, 12 H).
In order to reduce the reaction times, the microwave acti-
vation was also tested. It appears that although the reac-
tion proceeds indeed faster, in our hands, this process is
Acknowledgment
not as efficient as the thermal one, in terms of yield/deg- We acknowledge the financial support from CNRS and Ministère
radation rate trade-off, probably due to the absence of a délégué à la Recherche. We gratefully thank Prof. Jack Hamelin for
access to microwave facilities.
fine local temperature control.
Finally, the two-flask synthesis of trialdehyde 1 from
triphenylamine (2) was carried out with 52% isolated References
yield on the gram scale as follows: i) POCl (25 equiv),
DMF (23 equiv), 95 °C, 4 h; ii) hydrolysis and isolation,
3
(1) Suzuki, H.; Matsumoto, M. Jpn. Kokai Tokkyo Koho,
3163361, 1988.
6
but without purification; iii) POCl (25 equiv), DMF (23
equiv), 95 °C, 1.5 h. This method could easily be scaled
up without any loss of yield.
(2) Goto, S.; Abe, N.; Sasaki, O. Jpn. Kokai Tokkyo Koho,
3
01044452, 1989.
(
3) Kushibiki, N.; Takeuchi, K.; Kobayashi, H.; Masatomi, T.
Eur. Pat. Appl., 771809, 1997.
In summary, we have described a short, practical and effi-
cient preparation of the key building block 1. This two-
flask synthesis requires shorter overall time than the direct
one-flask threefold formylations, while giving higher
(
4) Nam, H.; Kang, D. H.; Kim, J. K.; Park, S. Y. Chem. Lett.
2000, 1298.
(5) Jiang, K. J.; Sun, Y. L.; Shao, K. F.; Yang, L. M. Chem. Lett.
004, 33, 50.
2
Synthesis 2005, No. 11, 1771–1774 © Thieme Stuttgart · New York

1
774
T. Mallegol et al.
PAPER
(17) Spangler, C. W.; Suo, Z.; Drobizhev, M.; Karotki, A.;
Rebane, A. NATO Sci. Ser. II 2003, 100, 139.
(18) Mongin, O.; Porrès, L.; Katan, C.; Pons, T.; Mertz, J.;
Blanchard-Desce, M. Tetrahedron Lett. 2003, 44, 8121.
(19) Lee, H. J.; Sohn, J.; Hwang, J.; Park, S. Y.; Choi, H.; Cha,
M. Chem. Mater. 2004, 16, 456.
(20) Yang, W. J.; Kim, D. Y.; Kim, C. H.; Jeong, M.-Y.; Lee, S.
K.; Jeon, S.-J.; Cho, B. R. Org. Lett. 2004, 6, 1389.
(21) Mongin, O.; Charlot, M.; Katan, C.; Porrès, L.; Parent, M.;
Pons, T.; Mertz, J.; Blanchard-Desce, M. Proc. SPIE-Int.
Soc. Opt. Eng. 2004, 5516, 9.
(22) Hua, J. L.; Li, B.; Meng, F. S.; Ding, F.; Qian, S. X.; Tian,
H. Polymer 2004, 45, 7143.
(
(
(
(
6) Meng, F. S.; Yao, Q. H.; Shen, J. G.; Li, F. L.; Huang, C. H.;
Chen, K. C.; Tian, H. Synth. Met. 2003, 137, 1543.
7) Meng, F.; Liu, C.; Hua, J.; Cao, Y.; Chen, K.; Tian, H. Eur.
Polym. J. 2003, 39, 1325.
8) Elandaloussi, E. H.; Spangler, C. W. Polym. Prepr. (Am.
Chem. Soc., Div. Polym. Chem.) 1998, 39, 1055.
9) Holmberg, S. E.; Spangler, C. W. Polym. Mater. Sci. Eng.
2
001, 84, 717.
(
(
(
10) Stadler, S.; Feiner, F.; Bräuchle, C.; Brandl, S.; Gompper, R.
Chem. Phys. Lett. 1995, 245, 292.
11) Lambert, C.; Gaschler, W.; Schmalzlin, E.; Meerholz, K.;
Brauchle, C. J. Chem. Soc., Perkin Trans. 2 1999, 577.
12) Elandaloussi, E. H.; Spangler, C. W.; Dirk, C.; Casstevens,
M.; Kumar, D.; Burzynski, R. Mater. Res. Soc. Symp. Proc.
(23) Wilson, C. D. US Patent, 2558285, 1951.
(24) Lai, G.; Bu, X. R.; Santos, J.; Mintz, E. A. Synlett 1997,
1275.
1
999, 561, 63.
(
(
(
(
13) Drobizhev, M.; Karotki, A.; Rebane, A.; Spangler, C. W.
Opt. Lett. 2001, 26, 1081.
(25) This could explain why DMF is often used in larger amounts
2
4
14) Yoo, J.; Yang, S. K.; Jeong, M.-Y.; Ahn, H. C.; Jeon, S.-J.;
Cho, B. R. Org. Lett. 2003, 5, 645.
15) Drobizhev, M.; Karotki, A.; Dzenis, Y.; Rebane, A.; Suo, Z.;
Spangler, C. W. J. Phys. Chem. B 2003, 107, 7540.
16) Spangler, C. W.; Suo, Z.; Drobizhev, M.; Karotki, A.;
Rebane, A. Proc. SPIE-Int. Soc. Opt. Eng. 2003, 4797, 146.
than POCl3. To overcome this problem, other authors tried
to use high boiling chlorinated solvents, such as 1,2-
1
9,26
dichloroethane, but the yields remained low.
(26) Jin, S.-H.; Jung, J.-E.; Yeom, I.-S.; Moon, S.-B.; Koh, K.;
Kim, S.-H.; Gal, Y.-S. Eur. Polym. J. 2002, 38, 895.
Synthesis 2005, No. 11, 1771–1774 © Thieme Stuttgart · New York