LETTER
Novel and Unexpected Deamination during the Palladium-Catalysed Cross-Coupling
439
and the other product to be the deaminated analogue (4a).
Compounds 3a and 4a were easily isolated (46% and 41%
yields respectively) by column chromatography, and
spectroscopic and elemental analyses confirmed the struc-
tures.
C4H9S
ZnCl
+
I
NH2
5
2b
Pd(PPh3)4, THF
Further investigations through the careful monitoring of a
repeat of entry 1 (Table 1) by GLC analysis did not reveal
the presence of any bromobenzene throughout the reac-
tion and failed to detect any deamination after the cou-
pling reaction was complete. These results suggest that
the deamination process occurs simultaneously with the
coupling reaction, or perhaps simply requires the presence
of the arylboronic acid.
C4H9S
NH2
6
Scheme 2
ronic acid, as reported by Novak and co-workers. 15 How-
ever, the deaminated products 4e-f (entries 7 and 8) retain
the lateral substituent from the starting amine and could
not possibly have arisen from this route.
The coupling reaction was carried out using 4-iodoaniline
(2b) (entry 2, Table 1) in an attempt to reduce deamina-
tion by increasing the rate of coupling. However, contin-
uous GLC analysis revealed similar results to those
obtained for entry 1 with a similar ratio (1:1) of compound
Control experiments were carried out in which amines
that are incapable of coupling (4-pentylaniline and 4-chlo-
roaniline) and 4-methoxyphenylboronic acid were subject
separately to the usual reaction conditions described for
entry 1, Table 1. In each case, no deamination was detect-
ed by GLC analysis, which appears to support our original
evidence that the deamination occurs simultaneously with
the coupling.
3
a to compound 4a, although isolated yields are slightly
different (Table 1).
Further combinations of arylboronic acids and haloa-
nilines in palladium-catalysed cross-coupling reactions
have all resulted in substantial amounts of a deaminated
biaryl product. Entries 3 and 4 (Table 1) are analogous to
entries 1 and 2 respectively, but employed an alkylsulfa-
nyl-substituted boronic acid (1b). However, the results of
the coupling reactions are identical with a 1:1 ratio (GLC
analysis) of the desired aminobiaryl (3b) and the deami-
nated biaryl (4b) recorded, and yields of around 40% of
each were isolated. Similarly, when the cyano-substituted
boronic acid (1c) was used (entry 5) isolated yields were
It is also interesting that of the organometallic units (zinc
and tin) that we have used in palladium-catalysed cross-
coupling reactions, only those involving the boronic acid
species cause deamination. For example, the coupling of
the phenylethynylzinc chloride (5) to 4-iodoaniline (2b)
gave a good isolated yield (67%) of amine 6 without any
deamination (Scheme 2).
5
3
9% and 42% for the aminobiaryl (3c) and the deaminated
biaryl (4c) respectively. The use of the naphthylboronic Deamination is a relatively difficult process to achieve,
acid (1d) did generate a higher proportion of the desired yet it has been shown to occur under the relatively mild
aminobiaryl (3d) (52%), but a 32% yield of the deaminat- conditions of palladium-catalysed cross-coupling reac-
ed biaryl (4d) was also isolated. These results (entries 1- tions employing a range of arylboronic acids and haloa-
6
) show that the nature of the boronic acid does not seem nilines. So far, deamination has not been achieved without
to affect significantly the outcome of the coupling reac- concomitant coupling, and although such deamination is
tion and deamination.
novel, interesting and provides additional background
knowledge, the scope of the reaction is somewhat limited.
Additionally, at this stage, it is impossible to suggest a
mechanism for this unusual deamination. However, in-
vestigations are continuing in order to find a suitable reac-
tion system that will generate ’deaminated starting
materials’rather than ’deaminated products’, which would
be of considerable synthetic application and complemen-
tary to alternative procedures based on diazotisation
methods. 1
Different anilines (2c and 2d), each with an electron-with-
drawing substituent ortho to the amine group, were also
used (entries 7 and 8). When 4-bromo-2-fluoroaniline (2c)
was used in a palladium-catalysed cross-coupling reaction
with boronic acid 1b, the desired aminobiaryl (3e) was ob-
tained in just 12% yield, and a 72% yield of the deaminat-
ed biaryl (4e) was isolated. It is probable that the electron-
withdrawing effect of the ortho fluoro substituent is re-
sponsible for the enhanced degree of deamination, a view
supported by the result shown in entry 8 for the coupling
of 2-amino-5-bromobenzonitrile (2d) with boronic acid
6-18
Typical Experimental Procedure (Entry 1)
1
b. In the case of entry 8, the degree of deamination is not
A solution of compound 1a (5.30 g, 0.028 mol) in DME
as significant as seen for entry 7, but 64% of the deaminat-
ed biaryl (4f) was isolated in comparison to only 20% of
the desired aminobiaryl (3f).
(
30 ml) was added dropwise to a stirred mixture of com-
pound 2a (3.90 g, 0.023 mol), sodium carbonate (4.00 g)
and tetrakis(triphenylphosphine)palladium(0) (0.80 g, 0.7
mmol) in DME (30 ml) and water (40 ml). The stirred
mixture was heated under reflux for 30 min. (GLC analy-
sis revealed a complete reaction and the presence of two
In four cases, it is possible that the deaminated products
4
a-d (entries 1-6) could have resulted from the coupling
of the phenyl from triphenylphosphine with the arylbo-
Synlett 1999, No. 4, 438–440 ISSN 0936-5214 © Thieme Stuttgart · New York
Hird, Michael
