58
D. Tzimopoulos et al. / Journal of Molecular Structure 965 (2010) 56–64
procedure there were added to the mixture under continuous stir-
ring 10.0 mmol of the appropriate triorganotin chloride in toluene
(20.0 mL). The solution became clear instantly and was left to react
under reflux for 5–6 h during which KCl deposited gradually on the
vessel walls. The solution was then filtered while hot in order to re-
move the remaining KCl, the solvent was removed at the rotary
evaporator and the solid (R = Me, Cy) or oily product (R = Du, Ph,
Bz) isolated was dried under vacuum. The products obtained were
dissolved in the minimum amount of hot hexane-dichloromethane
(10:1) and the precipitate formed upon cooling was filtrated and
recrystallized from hot toluene–hexane (5:1).
needed in order to verify the coordination mode present in every
case. In the IR spectrum the symmetric and asymmetric carboxyl-
ate stretching modes provide the widely accepted criterion of their
spacing,
Dm. Generally Dm
differences less than 240 cmꢀ1 are
accounting for bridging or chelating bonding modes while ones
above 260 cmꢀ1 for monodentate, [17]. In borderline cases and in
absence of crystallographic evidence only speculations can be
made, i.e. assumption o asymmetric bridging [18]. In addition,
individual peak positions have been sporadically used as indication
for the presence of specific coordination modes, i.e. the existence of
a carboxylate band in the proximity of 1630 cmꢀ1 is related to the
asymmetric bridging of the carboxylate [19].
The yellow or brownish-yellow products isolated in the case of
the methyl derivatives were further dissolved in hot toluene and
upon addition of petroleum ether the corresponding crystals used
for X-ray determination were deposited for compound 7a. In the
case of the butyl and benzyl derivatives it proved necessary to
use molecular sieves in order to avoid turpidity of the initial mix-
ture and undesired by-products. For the compounds of the 3,5-
disubstituted benzoic acid, the synthesis of the desired products
was carried out after in the same manner but the refluxing periods
had to be expanded to 3 and 8 h, respectively, for the two steps de-
scribed previously. In the case of compound 7a, upon standing at
room temperature a light yellow solid deposited within minutes
and was isolated and submitted for structural analysis which
proved that it was actually the hydrate of 7a. Upon further stand-
ing for a few days a second batch of crystals was deposited and
these were also collected and analyzed in the same way, proving
to be another polymorph of the same compound. Both structures
are reported and they are denoted as polymorphs B and A of com-
pound 7aꢂH2O, respectively. Selected crystallographic data are in-
cluded in Table 5.
It should be noted at this point that the same products can be
obtained by the interaction of the appropriate triporganostannyl
esters of the aminobenzoates, prepared and isolated as described
elsewhere [12] with a slight excess of pyridine-2-aldehyde in
refluxing methanol for a period of 2 h.
The complexes were characterized spectrophotometrically
(FTIR and UV–vis) and by elemental analysis; (theoretical values
are shown in brackets):
10: C, 68.87% (69.02%); H, 4.73% (4.46%); N, 12.34% (12.38%); 11:
C, 68.42% (69.02%); H, 4.81% (4.46%); N, 12.06% (12.38%); 12: C,
68.56% (69.08%); H, 4.54% (4.27%); N, 16.71% (16.96%); 7a: C,
49.07% (49.40%); H, 4.67% (4.66%); N, 6.99% (7.20%); 8a: C, 48.79%
(49.40%); H, 4.70% (4.66%); N, 6.79% (7.20%); 9a: C, 53.35%
(53.58%); H, 4.50% (4.66%); N, 11.77% (11.36%); 7b: C, 57.52%
(58.27%); H, 7.16% (7.04%); N, 5.19% (5.44%); 8b: C, 57.32%
(58.27%); H, 7.18% (7.04%); N, 5.05% (5.44%); 9b: C, 59.63%
(60.11%); H, 6.97% (6.51%); N, 8.75% (9.05%); 7c: C, 62.38%
(62.75%); H, 7.30% (7.13%); N, 3.99% (4.72%); 8c: C, 62.60%
(62.75%); H, 7.49% (7.13%); N, 4.12% (4.72%); 9c: C, 64.01%
(63.71%); H, 6.74% (6.65%); N, 7.86% (8.03%); 7d: C, 64.86%
(64.73%); H,4.55% (4.21%); N, 4.38% (4.87%); 8d: C, 64.55%
(64.73%); H,4.50% (4.21%); N, 4.50% (4.87%); 9d: C, 65.32%
(65.40%); H,4.23% (4.15%); N, 8.33% (8.25%); 7e: C, 65.78%
(66.15%); H, 5.08% (4.90%); N, 4.43% (4.54%); 8e: C, 65.92%
(66.15%); H, 4.94% (4.90%); N, 4.68% (4.54%); 9e: C, 66.08%
(66.59%); H, 4.79% (4.75%); N, 7.63% (7.77%).
The relevant vibrational information for the ligands (10–12),
their potassium salts (4–6) and the triorganotin derivatives (7–9)
are reported in Table 1. The Dm discussed above are also computed
and tabulated in order to aid the interpretation of the carboxylate
bonding mode.
The imine part of the ligands does not appear to take part in the
coordination to tin because the corresponding bands do not sift
from their position in the free ligands. Notable exceptions are 7b,
7c and 9a. In the first two the strengthening of the C@N bond
may be attributed to some twisting of the two rings of the ligand
therefore precluding participation of the specific bond in any ex-
tended aromatic system. In the latter case the lower wavenumber
of the bond stretching relative to the ligand may indicate a partic-
ipation of the imine N atom in an interaction which may be either
hydrogen-bond formation or coordination to tin. Regarding the
coordination mode of the carboxylate moiety the trimethyltin
derivatives 7a and 8a are proposed to be polymeric with carboxyl-
ate bridging adjacent tin centers while most of the rest compounds
appear to have monodentate carboxylate. There exist a number of
compounds, namely 7e, 8e, 9a and 9c for which no clear conclusion
may be drawn; the compounds may be of polymeric nature like the
trimethyltin derivatives discussed previously.
The above assignments for the asymmetric carboxylate and the
imino C@N bond stretchings are aided by the known spectra of the
corresponding unsubstituted benzoates [10,11,18,20] and by their
correlation with the spectra of the studied compounds. Further
support is provided by computations carried out on single mole-
cules which were performed using the PM6 Hamiltonian [21] as
implemented in the MOPAC2009 program [22]. Simple semi-
empirical computations are not expected to result the accurate
vibrational eigenvalues and there is always need for scaling in
Table 1
Relevant vibrational spectral data. Band maxima are given in wavenumbers.
Compound
m
as(COO)
m
s(COO)
D
m
m
(C@N)
m(Sn–O)
m(Sn–C)
10
4
1703
1557
1594
1578
1585
1587
1597
1701
1544
1599
1600
1604
1600
1598
1704
1563
1589
1587
1586
1587
1596
1293
1395
1374
1332
1328
1333
1335
1272
1403
1367
1338
1315
1331
1366
1294
1390
1338
1331
1326
1331
1332
410
162
220
246
257
254
232
429
141
232
262
289
269
232
410
173
251
256
250
256
267
1628
1629
1627
1645
1643
1632
1630
1629
1623
1625
1630
1637
1633
1627
1635
1633
1621
1636
1638
1636
1633
–
–
–
–
7a
7b
7c
7d
7e
11
5
8a
8b
8c
8d
8e
12
6
417
407
410
407
410
–
553
517
492
446
452
–
–
–
439
407
418
406
406
–
551
510
493
447
452
–
3. Results and discussion
–
–
3.1. Vibrational spectra
9a
9b
9c
9d
9e
407
407
406
407
406
550
526
482
448
452
The structure adopted by organotin carboxylates is subtle since
the coordination of the carboxylate ion may be monodentate,
bridging or chelate. A compilation of spectroscopic evidence is