P.-C. Ioannou et al.
Inorganica Chimica Acta 522 (2021) 120300
well as the repulsion between the Br atoms [76,77]; the Br… Br distance
(3.536 Å for 2a and 3.376 Å for 2b) is slightly shorter than the sum of
bromine Van der Walls radii (ca. 3.7 Å) [78]. On the other hand, the
steric constraints imposed by the biphenyl group on the backbone of the
BIPHEP ligand does not allow for the free rotation of the P atoms that is
necessary to accommodate P-Ni-P bite angles close to 109◦, the ideal bite
angle in a tetrahedral geometry.
probable that the reactions have been quenched due to deactivation of
catalyst. Since this inhibition occurred exclusively in the coupling of
iodoarene, it most probably originates in the interactions of iodide ions
released from iodoarene with catalytic species. As can be seen, not only
the novel catalysts 1 and 2, but also the reference catalyst is subject to
this inhibition, which therefore requires a more detailed mechanistic
study, exceeding the scope of the present work.
It should also be noted that compared to more concentrated reaction
systems, the product selectivity in dilute systems slightly decreased in
favor of the Grignard reagent homo-coupling. This undesired homo-
coupling took a significant part in all catalytic tests (Table 3). It was
most pronounced in the initial stage of the reaction and contributed little
in its final stage. However, it is not the case that the fastest reaction
would lead to the lowest yield of the homo-coupling by-product.
Compare, e.g., entries 3c, 6c and 9b in Table 3: the lowest SDM (7%)
corresponded to the reaction lasting 4 d, while the reaction completed
within 5 min exhibited SDM of 18%.
2.3. Catalytic behavior of complexes 1 and 2
2.3.1. Kumada coupling reactions
Complexes 1 and 2 were tested for their possible use to catalyze the
–
Kumada and the Suzuki-Miyaura C C coupling reactions. The catalytic
efficiency in Kumada coupling was tested on reactions of 4-tolylmagne-
sium bromide with 1-halogeno-4-tert-butyl-benzenes (Scheme 3) and
was compared with the efficiency of [Ni(dppp)Cl2] that is commonly
–
used to catalyze various C C coupling reactions [35–38]. As can be
To examine the possible contributions of uncatalyzed reactions, the
reaction of iodoarene with Grignard reagent was performed in the
absence of catalyst, under reaction conditions A. No product was
observed after 1 h of reaction, and no more than 0.5% yield of the
desired MT product after 20 h. Hence, it is clear that the uncatalyzed
reaction does not interfere with the catalytic one, and that the undesired
homo-coupling of Grignard reagent is a catalytic reaction. These find-
ings support the hypothesis that the iodoarene cross-coupling is actually
quenched owing to catalyst deactivation, as no coupling occurs in these
systems after 5 min of reaction.
seen, the formation of the desired cross-coupling product, 4-tert-butyl-4′-
methylbiphenyl (MT), was always accompanied by the formation of two
homo-coupling by-products: 4,4′-di-tert-butylbiphenyl (DT) formed
from halogenoarene and 4,4′-dimethylbiphenyl (DM) formed from
Grignard reagent (Scheme 3 and Table 3).
The first section of Table 4 shows the results of Kumada coupling
tests with chloroarene. As can be seen, the reaction was slow with all the
catalysts, but proceeded for at least four days. The final conversion was
72% (SMT = 82%) for 1, 62% (SMT 74%) for 2 and 82% (SMT 91%) for the
reference catalyst, i.e. [Ni(dppp)Cl2]. The highest difference between 1
and 2 on the one hand and [Ni(dppp)Cl2] on the other is evident in the
first phase of the reaction, when in systems catalyzed by 1 or 2 homo-
coupling of Grignard reagents (SDM 51–52%) prevailed over hetero-
coupling (SMT 44–46%), while hetero-coupling was clearly dominated
by [Ni(dppp)Cl2] (SMT 78% vs SDM 21%). A significant decrease in SDM in
the next stage of the reaction clearly indicates a substantial reduction in
homo-coupling of the Grignard reagent at this stage. It is worth noting
that the substrate homo-coupling was also significant in the systems
with catalysts 1 and 2 (final SDT 7%) but almost negligible (SDT 1–2%) in
the reference system.
An attempt has been made to correlate the catalytic behavior of [Pd
(P,P)X2] complexes with the magnitude of the P-Pd-P bite angle [37].
We can preliminarily follow this approach for Ni(II) complexes,
although there are not yet enough results in order to draw reliable
conclusions. Such data are available for some [Ni(PNP)X2] complexes
[20,21] with almost planar NiP2X2 coordination sphere and P-Ni-P bite
angles of ca. 73.5◦ [20,21,80,81], much smaller than the angles in
complexes 1 and 2. Complexes 1 and 2 are superior to complexes [Ni
(PNP)X2] in terms of efficiency in catalyzing Kumada reactions.
Compared to that, the reference catalyst [Ni(dppp)Cl2] exhibits a P-Ni-P
angle of ca. 91-92◦ (Table 2), slightly lower than complexes 1 and 2 (ca.
94◦), but afforded slightly higher conversions. Nevertheless, the differ-
ence is rather small to safely probe the bite-angle effect. In addition, Ni
(II) complexes alone can only be catalyst particle precursors, as recently
demonstrated by Jarvo et al. [48]. Our data can also be compared with
the data on the Kumada coupling of but-2-ylmagnesium chloride with
bromobenzene catalyzed by [Pd(BIPHEP)Cl2], affording but-2-
ylbenzene (yield 93%, selectivity greater than 99%) [37], but only
qualitatively, due to the substantial difference between the reactants
employed.
Tests with bromoarene substrate (second section of Table 3) gave
results differing from those obtained for chloroarene. The course of the
reaction was almost the same, exhibiting practically the same final
conversion (60–62%) and product selectivity (SMT 73–74%) with all
three catalysts. The time courses of selectivities to both homo-coupling
by-products were also the same. Thus, it can only be stated that the
catalytic performance of complexes 1 and 2 in bromoarene coupling is
identical to the catalytic performance of the reference catalyst [Ni(dppp)
Cl2].
Reactions with iodoarene, the most reactive of the tested substrates
[79] (third part of Table 3), gave results significantly different from
those obtained for the two previous systems. The reaction reached an
overall conversion of 90–93% within the first five minutes, regardless of
the catalyst used. The product as well as the co-product selectivity was
also practically the same regardless of the catalyst used (SMT 77–80%;
2.3.2. Suzuki-Miyaura coupling reactions
The results of the catalytic tests of complexes in Suzuki-Miyaura
coupling are summarized in Table 4. Unlike the Kumada reactions, the
Suzuki-Miyaura reactions gave only the desired cross product (Scheme
4). All three studied complexes, namely 1, 2 and Ni(dppp)Cl2], showed
negligible activity in this reaction, as did the previously studied [Ni
(PNP)X2] complexes bearing bidentate PNP ligands [20]. We also
examined the previously reported co-catalytic effect of PPh3 [39,40], by
employing the widely applied PEPPSITM-IPr catalyst with two mono-
dentate ligands as a benchmark. A significant increase in the yield (from
1 to 31–34%) was observed for 1 and 2, as well as for [Ni(dppp)Cl2]
(26%). Nevertheless, all these catalysts still showed a lower performance
compared to the Pd(II)-based PEPPSITM-IPr catalyst.
S
DM 17–21%; SDT 2–3%). Remarkably, however, the conversion did not
increase over the next hour. Therefore, another series of tests was per-
formed using five-fold diluted reaction mixtures (reaction conditions B)
to check whether the reactions quenched due to a decrease in reactant
concentration or due to catalyst deactivation. All three catalysts gave an
overall conversion of 21 ± 1% within 5 min of the reaction and no
further growth was observed during the next 24 h. Thus, it is highly
3. Conclusions
–
In this work, the synthesis and catalytic properties in C C coupling
Scheme 3. The Kumada coupling reaction investigated in this work.
reactions of the [Ni(rac-BIPHEP)X2] complexes (X = Cl (1), Br (2) is
4