Homo- and hetero-bimetallic flexibly linked dinuclear salphen complexes

Article abstract of DOI:10.5560/ZNB.2012-0088

A modification of the previously developed synthetic approach towards dinuclear flexibly linked salphen complexes is successfully utilized for the preparation of heterodinuclear salphen dimers. A dinuclear salphen species with Pd(II) and Cr(III) centers bears a stronger structural resemblance to the related bis-Cr(III) compound than the corresponding mononuclear Cr(III) salphen complex. Therefore, it was considered as a more useful model for the comparison with the homobinuclear Cr(III) complex regarding the catalytic activity in the ring-opening polymerization of b-butyrolactone, for which a bimetallic catalytic mechanism seems to operate. The polymerization results again have shown a higher activity of the homobinuclear Cr(III) complex.

Full text of DOI:10.5560/ZNB.2012-0088

Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen
Complexes
Sergei I. Vagin and Bernhard Rieger
Wacker Lehrstuhl fu¨r Makromolekulare Chemie, Technische Universita¨t Mu¨nchen,
Lichtenbergstraße 4, D-85748 Garching, Germany
Reprint requests to Prof. Dr. B. Rieger. E-mail: rieger@tum.de
Z. Naturforsch. 2012, 67b, 614 – 620 / DOI: 10.5560/ZNB.2012-0088
Received March 29, 2012
Dedicated to Professor Wolfgang Beck on the occasion of his 80th birthday
A modification of the previously developed synthetic approach towards dinuclear flexibly linked
salphen complexes is successfully utilized for the preparation of heterodinuclear salphen dimers.
A dinuclear salphen species with Pd(II) and Cr(III) centers bears a stronger structural resem-
blance to the related bis-Cr(III) compound than the corresponding mononuclear Cr(III) salphen
complex. Therefore, it was considered as a more useful model for the comparison with the ho-
mobinuclear Cr(III) complex regarding the catalytic activity in the ring-opening polymerization of
β-butyrolactone, for which a bimetallic catalytic mechanism seems to operate. The polymerization
results again have shown a higher activity of the homobinuclear Cr(III) complex.
Key words: Ring-opening Polymerization, β-Butyrolactone, Catalysis, Heterobimetallic Complexes
Introduction
Ring-opening polymerization (ROP) of β-bu-
resulting either in isotactically enriched or completely
atactic polymer.
All these observations together with DFT calcula-
tyrolactone (β-BL) to produce poly(3-hydroxy- tions strongly support a bimetallic catalytic mechanism
butyrate) (PHB) has long been considered as a promis- of the ROP of butyrolactone, according to which the
ing alternative to the biological synthesis of this polymer chain growth takes place by way of nucle-
biodegradable material. Successful results in this ophilic attack of a metal-coordinated carboxylato chain
respect have been achieved using various catalytic end onto the activated monomer at another metal cen-
systems including those based on phenolate com- ter, as shown in Scheme 1 [3]. Upon the appropriate
plexes of lanthanides and group III metals [1], zinc spatial orientation of the two complexes the conditions
complexes of β-diiminates [2], chromium complexes for stereocontrolled β-BL ring opening can be real-
of salphens [3], etc. In some cases, stereoregularly ized.
enriched PHBs could be obtained from racemic
In order to verify the occurrence of bimetallic catal-
β-BL upon utilization of certain achiral catalysts [4], ysis in ROP of β-BL in the presence of chromium
whereas their structurally most closely analogous salphen complexes, we have recently synthesized a se-
complexes do not necessarily exhibit stereocontrol ries of mono- and dinuclear chromium salphens, where
during the polymerization. This is especially inherent the chromium centers were considered to have the
to the chromium(III) salphen complexes, which can same electronic environment [6]. With the chromium
even switch the stereoregularity of produced PHB to β-BL loading of 0.1 mol-%, the dinuclear com-
from isotactically enriched (bulk polymerization) to plexes have shown approximately 5 times higher ac-
syndiotactically enriched (in the presence of solvent, tivity than the mononuclear one. This was attributed to
such as chlorobenzene) [5]. The substituents at the the effect of a higher local concentration of chromium
phenylene ring or at the phenolate moieties of the centers in dinuclear complexes, thus testifying to the
salphen ligand have also been shown to influence involvement of bimetallic processes in the studied ring-
the tacticity of PHB in bulk polymerization of β-BL, opening polymerization.
c
ꢀ 2012 Verlag der Zeitschrift fu¨r Naturforschung, Tu¨bingen · http://znaturforsch.com
Unauthenticated
Download Date | 6/9/17 11:36 AM

S. I. Vagin – B. Rieger · Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen Complexes
615
Scheme 1. Possible path for the ROP of β-
BL in the presence of Cr-salphens.
Despite the similarity of the compared complexes in rity in this product was a dehydrobrominated derivative
terms of electronic properties of the central metal, still of 4.
there can be some difference in solution behavior of
Compound 4 can be subsequently coupled with
monomeric and dimeric complexes due to the distinct a metal complex of 1 possessing enough stability un-
bulk structure (difference in molecular weight, dipole der the conditions of the coupling reaction. For ex-
moment, number of functional centers per molecule, ample, the Zn complex proved to be unsuitable due
etc.). Thus, the solubility and aggregation tendency of to scrambling. That is, the signals of metal-free di-
these catalytic systems due to specific as well as van salphen molecules together with mono- and dimetal-
der Waals interactions between catalyst molecules can lated species were detected by ESI mass spectral analy-
be distinct, which may be reflected in the catalytic ac- sis of such a reaction mixture. This cannot be attributed
tivity, as was noticed for example in the experiments on to an artifact of the ESI measurement, since according
copolymerization of propylene oxide and carbon diox- to our experience a demetallation of Zn-salphen com-
ide [7]. In this respect, a comparison of the polymer- plexes is usually not observed under ESI conditions.
ization activity of two dinuclear systems, one bearing
In contrast, palladium(II) forms a very stable com-
two active metal centers and the other bearing active plex with salphen 1, which therefore can be applied in
and inactive metal centers, would be more reasonable. the coupling reaction. Furthermore, this complex is co-
Here we report on the synthesis and β-BL poly- ordinatively saturated, which makes it inactive in catal-
merization activity of a heterometallic Cr/Pd catalyst, ysis. In this respect, the heteronuclear Cr/Pd species
which resembles structurally the previously reported would be ideal for the comparison with the previously
bis-chromium salphen system, but possesses only one reported homodinuclear Cr-salphen species [6].
ROP-active center.
The reaction of equimolar amounts of 4 and 5 pro-
vided a di-salphen species 6 with one coordinated Pd
ion. Simple precipitation of this compound from ace-
Results and Discussion
tonitrile yielded the product in ca. 90% purity as es-
1
The previously described synthetic approach to- timated by H NMR spectroscopy. High purity was
1
wards dinuclear salphen systems can be modified to achieved by column chromatography. The H NMR
afford heterodimetallic complexes (Scheme 2).
spectrum of 6 represents a superposition of signals of 4
The resorcine-functionalized product 1 is a ver- and 5 in the aromatic region as well as in the region of
satile precursor for modular design of di-salphens, the t-Bu proton signals, indicating high conformational
where the length of the spacer and the nature of the flexibility of 6 in CD2Cl2 solution (Fig. 1).
metal centers can be easily varied. To afford the het-
Representative is the disappearance of -OH (resor-
erobimetallic systems, the alkylation of 1 with ex- cin) and of -CH2-Br resonances in the spectrum of 6.
cess α,ω-dibromoalkanes has to be carried out first. This complex could be easily metallated further, as
For example, reaction with ca. 20 equivalents of 1,3- shown by its reactions with CrCl2 or with Zn(OAc)2
dibromopropane allowed to isolate 69% of the 3- to give the heterodinuclear complexes 7 and 8, respec-
bromopropyl-functionalized salphen 4 in 92% purity tively. The first reaction was performed on a prepara-
without chromatographic purification. The main impu- tive scale, whereas the second one was only a quali-
Unauthenticated
Download Date | 6/9/17 11:36 AM

616
S. I. Vagin – B. Rieger · Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen Complexes
Scheme 2. Synthesis of difunctional salphen ligands: i 0.5 equiv. 1,3-dibromopropane, Cs2CO3 in acetonitrile; ii CrCl2 in
THF, lutidine, air; iii 20 equiv. 1,3-dibromopropane, Cs2CO3 in acetonitrile; iv Pd(OAc)2, DMF; v Cs2CO3 in acetonitrile; vi
CrCl2 in THF, lutidine, air (7) or Zn(OAc)2 in DMF (8).
tative test. The recorded ESI-MS spectra of the com- heterogeneous multicomponent systems. Nevertheless,
plexes display neither signs of metal scrambling or the main object pursued in the present work was to
substitution, nor of incompletely metallated products investigate the effect of heterodinuclear salphen com-
(Fig. 2).
plexes on the catalytic performance in the synthesis of
We believe that salphen building units like 1 and PHB.
4 can also be utilized for grafting reactions as well
Thus, polymerization of β-BL in the presence of
as for the construction of various homogeneous and complexes 3 and 7 was compared under equal con-
Unauthenticated
Download Date | 6/9/17 11:36 AM

S. I. Vagin – B. Rieger · Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen Complexes
617
1
Fig. 1. Comparison of H NMR spectra (CD2Cl2) of monomeric precursors 4 and 5 and their condensation product 6. The
spectra are cut for better visualization. Signals originating from solvent or impurities are marked with asterisks. In the
spectrum of 4 the low-intensity low-field-shifted satellites of the signals of the inner OH groups are due to the marginal
deuterium exchange.
ditions (100 ◦C). In order to minimize an error in ratio of 1 to 1000 [6]. In the experiment with a ratio
the results, the polymerization was performed simul- of 1 to 631, complex 3 afforded nearly twice the con-
taneously for both systems using the same lactone version than with a 1 : 1000 ratio to β-BL: The yields
batch as well as the same heating device. The ratio of PHB after 5 hours of reaction were 67% and 35%,
of chromium centers to β-BL was also practically the respectively, which is in a good agreement with the
same for both experiments (1 : 631 for 3 and 1 : 623 expectations. However, the difference in activity (per
for 7). The progress of the reaction was monitored chromium center) between complexes 3 and 7 is not
1
by H NMR measurements of aliquots, as shown in that drastic as the difference between productivities of
Fig. 3. The reaction seems to accelerate with time up the di- and mononuclear chromium salphens described
to a certain time, after which a typical dependency before [6].
for a zero order reaction for lactone is observed. The
It is doubtful that the palladium center in a salphen
acceleration effect found for the reaction cannot be can participate in the cooperative polymerization mech-
unambiguously ascribed yet. The plots of conversion anism. In this respect, the higher activity of the hetero-
versus reaction time can be very reliably approximated dinuclear Pd/Cr complex compared to the mononu-
by polynomial functions of third order. This allows an clear Cr complex can rather be attributed to the presence
easy determination of the reaction rate at any time of of a covalently linked Pd-salphen unit which causes
the reaction. The ratio of reaction rates for complexes an enhanced intermolecular interaction between two
3 and 7 at β-BL conversions of 0%, 20% and 40% has Pd/Cr molecules, e. g. by π stacking (dimerization or
been determined to be 1.88, 1.92 and 2.04, respectiv- even a higher degree of aggregation). Indeed, the planar
ely.
Pd-salphen unit is unpolar and may have a tendency
Previously we have reported on β-BL polymeriza- to aggregate in a polar solvent like β-BL. This would
tion with the dinuclear complex 3 at a Cr to β-BL lead to an increased local concentration of Cr centers
Unauthenticated
Download Date | 6/9/17 11:36 AM

618
S. I. Vagin – B. Rieger · Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen Complexes
Fig. 2. ESI-MS spectra of
6, 7 and 8.
and explain a relatively high activity of the Pd/Cr PHB with Mw of 24.5 kg mol−1 and a PDI of 2.2 at
complex. It is, however, interesting to note that the 36% conversion, whereas PHB obtained with 7 had Mw
average molecular weights of polymers obtained with of 13 kg mol−1 and a PDI of 3.5 at 45% conversion.
catalysts 3 and 7 at nearly the same conversion differ 1H NMR analysis proved that PHB formed with 7 has
by approximately a factor of 2. That is, catalyst 3 gave a higher content of crotonic ester end groups compared
to PHB obtained with 3, which is a consequence of
the thermal degradation of PHB due to a longer re-
80
action time at 100 ◦C [8]. Indeed, catalyst 7 requires
nearly the double time to reach the same conversion
70
7
3
of β-BL as compared to catalyst 3, during which the
degradation proceeds with the rate proportional to the
polymer concentration.
60
50
40
30
20
10
0
In conclusion, by synthesizing heterodinuclear bis-
salphens we have demonstrated the versatility of the
earlier developed approach towards dinuclear salphen
complexes. Comparison of two dinuclear species,
namely one bearing Pd and Cr centers and another
possessing two Cr centers, has been performed on the
ground of their activity in the ring-opening polymer-
ization of β-butyrolactone. The bis-chromium species
was found to be more active than its Pd/Cr analog,
thus once again underlining the role of a cooperative
bimetallic mechanism in the studied reaction.
0
10000
20000
Reaction time (s)
30000
40000
Fig. 3. Plots of conversion versus reaction time for catalysts
3 and 7 in ROP of β-BL.
Unauthenticated
Download Date | 6/9/17 11:36 AM

S. I. Vagin – B. Rieger · Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen Complexes
619
Experimental Part
8.62 (s, 1H), 8.68 (s, 1H), 13.46 (s, 1H), 13.60 (s, 1H). – MS
((+)-ESI; isopropanol): m/z (%) = 790 (100), [M+Na]+,
768 (10), [M+H]+. – Analysis for C45H57BrN2O4: calcd.
C 70.21, H 7.46, N 3.64; found C 70.47, H 7.57, N 3.77.
Starting materials and solvents were purchased from com-
mercial sources and were used as received, unless mentioned
otherwise. Bis-1,2-(3,5-bis-tert-butylsalicylaldimino)-4-(3-
hydroxyphenoxy)-benzene 1, the difunctional ligand 2 and
its bis-chromium complex 3 were prepared as described pre-
viously [6].
Mono-palladium complex 6 of the difunctional ligand 2
100 mg (0.133 mmol) of complex 5, 120 mg (ca.
0.143 mmol) of 92% pure compound
4 and 29 mg
Instrumentation
(0.089 mmol) Cs2CO3 were stirred at 80 ◦C in 2 mL ace-
tonitrile for 2 h. The formed precipitate was separated by de-
cantation, and 20 mg of 4 together with 20 mg of Cs2CO3
were added to the residual solution. This was heated with
stirring for additional 3 h, followed by decantation of the
formed precipitate. The precipitates were combined and re-
precipitated from acetonitrile by heating up to 80 ◦C and
cooling down, giving 130 mg of 90% pure (1H NMR) com-
pound 7. All the residual solutions were also combined, evap-
orated to dryness and exposed to column chromatography on
silica gel using a pentane / dichloromethane mixture with
gradient. By this procedure additional 33 mg of relatively
pure compound 6 was isolated. Finally, overall product 6 was
chromatographed to give 126 mg (66%) of 1H NMR-pure 6.
It has to be noticed that chromatography causes a partial ir-
reversible adsorption / decomposition of the product on sil-
ica gel according to the residual coloring of the phase. – 1H
NMR (CD2Cl2): δ = 1.28 – 1.31 (4s, 36H), 1.40 (2s, 18H),
1.50 (2s, 18H), 2.22 (p, 2H), 4.10 – 4.15 (m, 4H), 6.62 – 6.74
(m, 6H), 6.92 – 7.01 (m, 3H), 7.18 – 7.29 (m, 7H), 7.43 (t,
2H), 7.52 – 7.54 (m, 3H), 7.83 (d, 1H), 8.43 (s, 1H), 8.47 (s,
1H), 8.61 (s, 1H), 8.67 (s, 1H), 13.46 (s, 1H), 13.60 (s, 1H). –
FT-IR (ATR): ν = 2952 s, 2905 m, 2868 m, 1611 sh, 1579 s,
1518 s, 1483 s, 1467 s, 1439 m, 1417 s, 1387 m, 1360 s,
1331 w, 1268 s, 1250 m, 1166 vs, 1131 vs, 1100 w, 1058 w,
1026 w, 986 w, 932 w, 915 w, 858 m, 841 m, 805 w, 785 m,
772 m, 729 w, 684 m, 643 w, 635 w, cm−1. – MS ((+)-ESI;
isopropanol): m/z (%) = 1441 (100), [M+H]+, 1463 (25),
[M+Na]+. – Analysis for C87H106N4O8Pd: calcd. C 72.45,
H 7.41, N 3.88; found C 72.06, H 7.43, N 3.78.
FT-NMR: Bruker ARX 300 MHz 1H, 75 MHz 13C. FT-
IR: Bruker Vertex 70 with Bruker Platinum ATR-unit. ESI-
MS: Varian LC-MS 500 (50 – 2000 Da). GPC: Varian GPC-
50 (chloroform with 0.1% Bu4NBF4, polystyrene narrow
standard calibration, Varian Olexis column set 600 mm).
Elemental analysis: EA Euro 3000 (Kehatech), Elementar
Vario EL. EDX analysis: Tabletop SEM Hitachi TM 1000
equipped with Oxford Instruments detector.
Synthesis
Palladium complex of 1 (compound 5)
150 mg (0.23 mmol) of 1 and 65 mg (0.29 mmol) of
Pd(OAc)2 in 4 mL DMF were stirred at room temperature
for 6 h. The product was precipitated by addition of wa-
ter, centrifuged and passed through a short silica column
with dichloromethane. Yield after drying 158 mg (90%). –
1H NMR (CD2Cl2): δ = 1.30 (s, 9H), 1.32 (s, 9H), 1.50
(2s, 18H), 5.11 (s, 1H), 6.55 (t, 1H), 6.61 – 6.65 (dd, 2H),
6.99 – 7.03 (dd, 1H), 7.19 – 7.25 (m, 3H), 7.53 (m, 3H),
7.84 (d, 1H), 8.42 (s, 1H), 8.47 (s, 1H). – FT-IR (ATR):
ν = 2951 s, 2904 m, 2866 m, 1578 s, 1546 vw, 1517 s,
1483 s, 1461 s, 1416 s, 1384 s, 1357 s, 1332 m, 1266 s,
1237 m, 1186 w, 1164 vs, 1128 vs, 1098 m, 1025 w, 986 m,
950 w, 933 m, 914 w, 887 w, 861 w, 840 m, 803 w, 784 m,
749 w, 682 m, 634 m, cm−1. – MS ((+)-ESI; isopropanol):
m/z (%) = 1528 (100), [2M+Na]+, 775 (40), [M+Na]+,
753 (20), [M+H]+. – Analysis for C42H50N2O4Pd: calcd.
C 66.97, H 6.69, N 3.72; found C 66.92, H 6.89, N 3.38.
Bis-1,2-(3,5-bis-tert-butylsalicylaldimino)-4-(3-(3-bromo-1-
propyloxy)phenoxy)-benzene (4)
Hetrodinuclear palladium/chromium complex 7 of the
dimeric ligand 2
Compound 1 (1 g, 1.54 mmol) and Cs2CO3 (0.4 g,
1.2 mmol) in 10 mL acetonitrile were stirred at 60 ◦C for
30 min, followed by addition of 1,3-dibromopropane (6.2 g,
The complex was prepared by a quantitative reaction of 6
30.8 mmol). After 1 h the reaction mixture was cooled, wa- with CrCl2 as described previously [6]. – FT-IR (ATR): ν =
ter was added, and the mixture was vigorously stirred and 2952 s, 2904 m, 2867 m, 1578 br. s, 1520 s, 1484 s, 1462 s,
centrifuged. The water phase was decanted, and the residue 1418 s, 1384 s, 1358 s, 1330 m, 1261 br. s, 1166 vs, 1130 vs,
was washed with small portions of methanol, followed by 1099 w, 1058 w, 998 w, 955 w, 931 w, 915 w, 841 s, 808 w,
re-precipitation from a minimal quantity of hot acetonitrile. 784 s, 749 m, 684 m, 635 m, cm−1. – MS ((+)-ESI; iso-
Yield ca. 0.8 g (69%) of > 92% pure (1H NMR) product. – propanol): m/z (%) = 1490 (100), [M–Cl]+. – EDX anal-
1H NMR (CD2Cl2): δ = 1.28 (s, 9H), 1.30 (s, 9H), 1.40 (2s, ysis for the Cr/Cl/Pd ratio: 1.00 : 0.99 : 1.06. – Analysis for
18H), 2.28 (p, 2H), 3.59 (t, 2H), 4.07 (t, 2H), 6.62 – 6.70 (m, C87H104N4O8ClCrPd: calcd. C 68.40, H 6.86, N 3.67; for
3H), 6.94 – 7.01 (m, 2H), 7.21 – 7.29 (m, 4H), 7.43 (t, 2H), C87H104N4O8ClCrPd·H2O: calcd. C 67.60, H 6.91, N 3.62;
Unauthenticated
Download Date | 6/9/17 11:36 AM

620
S. I. Vagin – B. Rieger · Homo- and Hetero-bimetallic Flexibly Linked Dinuclear Salphen Complexes
for C87H104N4O8ClCrPd·THF: calcd. C 68.32, H 7.06, N tively. Purified β-butyrolactone (vacuum distillation from
3.50; found C 67.94, H 7.42, N 3.43.
CaH2) was added subsequently (2 mL for compound 3 and
1 mL for compound 7), both flasks were tightly closed and
immersed into the preheated oil bath (100 ◦C). During the
Polymerization of β-butyrolactone
Two small baked out flasks equipped with magnetic stirrer course of the reaction, small aliquots were quickly taken for
were charged with 27.8 mg of 3 and 28.5 mg of 7, respec- analysis. All manipulations were performed under air.
[1] A. Amgoune, C. M. Thomas, S. Ilinca, T. Roisnel, J.- [5] R. Reichardt, S. Vagin, R. Reithmeier, A. Ott, B. Rieger,
F. Carpentier, Angew. Chem. 2006, 118, 2848 – 2850;
Angew. Chem. Int. Ed. 2006, 45, 2782 – 2784.
Macromolecules 2010, 43, 9311 – 9317.
[6] S. I. Vagin, R. Reichardt, S. Klaus, B. Rieger, J. Am.
Chem. Soc. 2010, 132, 14367 – 14369.
[7] S. Klaus, S. I. Vagin, M. W. Lehenmeier, P. Deglmann,
A. K. Brym, B. Rieger, Macromolecules 2011, 44,
9508 – 9516.
[8] M. Kawalec, G. Adamus, P. Kurcok, M. Kowalczuk,
I. Foltran, M. L. Focarete, M. Scandola, Biomacro-
molecules 2007, 8, 1053 – 1058.
[2] L. R. Rieth, D. R. Moore, E. B. Lobkovsky, G. W.
Coates, J. Am. Chem. Soc. 2002, 124, 15239 – 15248.
[3] M. Zintl, F. Molnar, T. Urban, V. Bernhart, P. Preishuber-
Pflu¨gl, B. Rieger, Angew. Chem. 2008, 120, 3508 – 3510;
Angew. Chem. Int. Ed. 2008, 47, 3458 – 3460.
[4] J.-F. Carpentier, Macromol. Rapid Commun. 2010, 31,
1696 – 1705.
Unauthenticated
Download Date | 6/9/17 11:36 AM