A family of wheel-shaped charge-neutral heterometallic {FeIII4LnIII2}- and {FeIII18MIII6}-type coordination clusters demonstrates the intricate interplay of solvent effects and structure-directing roles of semiflexible bridging ligands. The {Fe4Ln2}-type compounds [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(EtOH), Ln = Dy (1a), Er (1b), Ho (1c); [Fe4Tb2(O2CCMe3)6(N3)4(Htea)4] (1d); [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(CH2Cl2), Ln = Dy (2a), Er (2b); [Fe4Ln2(O2CCMe3)4(N3)6(Htea)4]·2(EtOH)·2(CH2Cl2), Ln = Dy (3a), Er (3b) and the {Fe18M6}-type compounds [Fe18M6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·n(solvent), M = Dy (4, 4a), Gd (5), Tb (6), Ho (7), Sm (8), Eu (9), and Y (10) form in ca. 20-40% yields in direct reaction of trinuclear FeIII pivalate or isobutyrate clusters, lanthanide/yttrium nitrates, and bridging triethanolamine (H3tea) and azide ligands in different solvents: EtOH for the smaller {Fe4Ln2} wheels and MeOH/MeCN or MeOH/EtOH for the larger {Fe18M6} wheels. Single-crystal X-ray diffraction analyses revealed that 1-3 consist of planar centrosymmetric hexanuclear clusters built from FeIII and LnIII ions linked by an array of bridging carboxylate, azide, and aminopolyalcoholato-based ligands into a cyclic structure with a cavity, and with distinct sets of crystal solvents (2 EtOH per formula unit in 1a-c, 2 CH2Cl2 in 2, and 2 EtOH and 2 CH2Cl2 in 3). In 4-10, the largest 3d/4f wheels currently known, nearly linear Fe3 fragments are joined via mononuclear Ln/Y units by a set of isobutyrates and amino alcohol ligands into virtually planar rings. The magnetic properties of 1-10 reveal slow magnetization relaxation for {Fe4Tb2} (1d) and slow relaxation for {Fe4Ho2} (1c), {Fe18Dy6} (4), and {Fe18Tb6} (6).
A family of wheel-shaped charge-neutral heterometallic {FeIII4LnIII2}- and {FeIII18MIII6}-type coordination clusters demonstrates the intricate interplay of solvent effects and structure-directing roles of semiflexible bridging ligands. The {Fe4Ln2}-type compounds [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(EtOH), Ln = Dy (1a), Er (1b), Ho (1c); [Fe4Tb2(O2CCMe3)6(N3)4(Htea)4] (1d); [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(CH2Cl2), Ln = Dy (2a), Er (2b); [Fe4Ln2(O2CCMe3)4(N3)6(Htea)4]·2(EtOH)·2(CH2Cl2), Ln = Dy (3a), Er (3b) and the {Fe18M6}-type compounds [Fe18M6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·n(solvent), M = Dy (4, 4a), Gd (5), Tb (6), Ho (7), Sm (8), Eu (9), and Y (10) form in ca. 20-40% yields in direct reaction of trinuclear FeIII pivalate or isobutyrate clusters, lanthanide/yttrium nitrates, and bridging triethanolamine (H3tea) and azide ligands in different solvents: EtOH for the smaller {Fe4Ln2} wheels and MeOH/MeCN or MeOH/EtOH for the larger {Fe18M6} wheels. Single-crystal X-ray diffraction analyses revealed that 1-3 consist of planar centrosymmetric hexanuclear clusters built from FeIII and LnIII ions linked by an array of bridging carboxylate, azide, and aminopolyalcoholato-based ligands into a cyclic structure with a cavity, and with distinct sets of crystal solvents (2 EtOH per formula unit in 1a-c, 2 CH2Cl2 in 2, and 2 EtOH and 2 CH2Cl2 in 3). In 4-10, the largest 3d/4f wheels currently known, nearly linear Fe3 fragments are joined via mononuclear Ln/Y units by a set of isobutyrates and amino alcohol ligands into virtually planar rings. The magnetic properties of 1-10 reveal slow magnetization relaxation for {Fe4Tb2} (1d) and slow relaxation for {Fe4Ho2} (1c), {Fe18Dy6} (4), and {Fe18Tb6} (6).
Wheel-shaped high-nuclearity
metal clusters of paramagnetic ions, which can be regarded as chains
with a finite number of spin centers, have received substantial attention,
largely stimulated by their fascinating physical properties, including
single-molecule magnet (SMM) and spin qubit characteristics.[1] In this field pioneering synthesis efforts have
mainly focused on the design and investigation of pure 3d systems:
the “ferric wheels” reported by Lippard, Christou, and
their cowokers[2] or by “manganese
wheels” prepared by Christou et al.,[3] a “nickel wheel”[4] and a
“chromium wheel”[5] reported
by Winpenny and Timco et al., lately extended to fascinating heterometallic
{Cr7M} and {Cr8M} wheels.[6] Increasing interest in the molecular magnetic anisotropy
of such clusters led to the integration of lanthanide (4f) ions.[7] In 2010 the group of Murray[8] reported the first {Dy6} wheel that exhibits
slow magnetization relaxation. This development has been followed
by the construction of heterometallic d/f wheel-shaped clusters, that
combine highly anisotropic LnIII ions with the large spin
values derived from exchange-coupled transition metal ions, but only
very few d/f wheel-based SMMs have been discovered thus far.[9]Of special interest in this context are
wheel-shaped polynuclear FeIII-LnIII SMM clusters,
and to our knowledge, only {Fe4Dy4} and {Fe4Ln2} (LnIII = Eu, Lu) wheels were described
in literature by Powell et al.[10] Moreover,
the development of general synthetic routes toward novel heterometal
coordination clusters with wheel topologies represent an ongoing challenge.Among different strategies for the preparation of metallic wheels,
the use of alkoxides as chelating and bridging ligands has been successful
in many cases.[8,11] We also explored the potential
of using both structure-directing amino alcohols and carboxylate bridges
for the synthesis of heterometallic coordination cluster families.[12,13a] Using recent advances in synthetic methodology and expanding our
development efforts for large polynuclear iron systems,[13] we herein report a straightforward protocol
to heterometallic Fe/Ln (or Fe/Y) wheels using trinuclear μ-oxo
FeIII pivalate or isobutyrate precursors and tetratopic
triethanolamine (H3tea) and azide ligands, both capable
to bridge different types of metal ions (Figure ). This approach resulted in a remarkable
family of hexanuclear {Fe4Ln2}-type wheel-shaped
compounds, namely, [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(EtOH), LnIII = Dy (1a), Er (1b), Ho (1c), and [Fe4Tb2(O2CCMe3)6(N3)4(Htea)4] (1d), [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(CH2Cl2), LnIII = Dy (2a), Er (2b), and [Fe4Ln2(O2CCMe3)4(N3)6(Htea)4]·2(EtOH)·2(CH2Cl2), LnIII = Dy (3a),
Er (3b), as well as in the currently largest (24-nuclear)
3d/4f wheels, [Fe18M6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·n(solvent), MIII =
Dy (4, 4a), Gd (5), Tb (6), Ho (7), Sm (8), Eu (9), and Y (10), where H3tea = triethanolamine.
Figure 1
Schematic
syntheses of {Fe4Ln2}-type (1)
and {Fe18M6}-type (M = Ln: 4–9; M = Y: 10) compounds.
Schematic
syntheses of {Fe4Ln2}-type (1)
and {n class="Chemical">Fe18M6}-type (M = Ln: 4–9; M = Y: 10) compounds.
Experimental Section
Materials and Instrumentations
All reagents were purchased from commercial sources and used without
further purification. All synthetic procedures have been performed
under aerobic conditions using commercial-grade solvents. [Fe3O(O2CCMe3)6(H2O)3]O2CCMe3·2(Me3CCO2H), [Fe3O(O2CCMe3)6(EtOH)3]NO3·(EtOH), [Fe6O2(OH)2(O2CCMe3)12], and [Fe3O(O2CCHMe2)6(H2O)3]NO3·2(MeCN)·2(H2O) were prepared by using method described in the literature.[13a,13b] A Bandelin Sonorex RK-100H ultrasonic bath operating at 35 kHz with
a maximum power output of 160 W was used for ultrasonic irradiation
in the syntheses of 4–10.IR
spectra (4000–400 cm–1) were recorded on
a PerkinElmer Spectrum One spectrometer using KBr pellets. Thermogravimetric
analysis/differential thermal analysis (TGA/DTA) measurements were
performed with a Mettler Toledo TGA/SDTA 851 in dry N2 (60
mL min–1) at a heating rate of 10 K min–1 from 25 to 800 °C for 1–3 (Figures S24–S30) and at a heating rate
of 5 K min–1 from 25 to 1000 °C for 4–10 (Figures S31–S38). Metal analysis was performed using inductively coupled plasma-optical
emission spectroscopy (ICP-OES) at ZEA-3, Research Centre Jülich.
Syntheses
of [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(EtOH)
(LnIII = Dy (1a); Er (1b); Ho
(1c))
Compounds 1a–c can be prepared by two methods starting from trinuclear
(A) or hexanuclear (B) ferric pivalates:Method A. A solution of [Fe3O(O2CCMe3)6(EtOH)3]NO3·(EtOH) (0.208 g, 0.2
mmol), Ln(NO3)3·xH2O (0.3 mmol), NaN3 (0.038 g, 0.6 mmol), and H3tea (0.084 g, 0.6 mmol) in 10 mL of EtOH was heated under
reflux (ca. 78 °C) for 30 min and then filtered. The filtrate
was kept in a closed vial at room temperature. After two weeks, yellow
crystals of 1a–c, suitable for single-crystal
X-ray analysis, were filtered off, washed with EtOH, water, and MeCN,
and dried in air. Yield (based on Fe): 0.084 g, 28% (1a); 0.100 g, 33% (1b); 0.065 g, 22% (1c).
After removal of 1 from the mixture, well-defined dark-brown
crystals of [Fe8O3(O2CCMe3)6(N3)3(tea)(Htea)3]·0.5(EtOH)[13a] formed in ca. 30% yield if the filtrate was
kept in a closed vial at room temperature for several additional weeks.Method B. A solution of [Fe6O2(OH)2(O2CCMe3)12] (0.161
g, 0.1 mmol), Ln(NO3)3·xn class="Chemical">H2O (0.3 mmol), NaN3 (0.038 g, 0.6 mmol), and
H3tea (0.084 g, 0.6 mmol) in 10 mL of EtOH was heated under
reflux for 30 min and then filtered. The filtrate was stored in a
closed vial at room temperature. After two weeks, the yellow crystals
of 1a–c suitable for a single-crystal
X-ray diffraction analysis were filtered off, washed with EtOH, water,
and MeCN, and dried in air. Yield (based on Fe): 0.075 g, 25% (1a); 0.130 g, 43% (1b); 0.060 g, 20% (1c). Dark brown crystals of [Fe8O3(O2CCMe3)6(N3)3(tea)(Htea)3]·0.5(EtOH) also slowly formed in the mother liquor once
the yellow crystals of 1 had been removed.
The
identity of 1a–c prepared by method
B was established from their IR data, elemental and TG analyses, as
well as by a single-crystal X-ray diffraction analysis.
1a {Fe4Dy2}
Elemental analysis calcd.
for C58H118Dy2Fe4N16O26 (2004.1 g mol–1): C 34.76,
H 5.93, N 11.15%; found: C 33.61, H 5.68, N 10.46%; calcd. mass ratio
of Dy/n class="Chemical">Fe: 1.45; found: 0.85. IR (KBr): ν = 3434 (br, m), 2958
(m), 2899 (m), 2865 (m), 2098 (s), 2076 (vs), 2058 (sh), 1582 (sh),
1536 (vs), 1481 (s), 1459 (m), 1420 (s), 1402 (s), 1353 (m), 1284
(m), 1226 (m), 1152 (w), 1098 (m), 1074 (m), 1024 (m), 918 (sh), 902
(m), 836 (w), 787 (w), 741 (w), 650 (m), 624 (m), 603 (m), 569 (m),
553 (sh), 489 (m), 428 (m) cm–1.
1b {Fe4Er2}
Elemental analysis calcd.
for C58H118Er2Fe4N16O26 (2011.6 g mol–1): C 34.60,
H 5.91, N 11.13%; found: C 32.72, H 5.92, N 11.20%; calcd. mass ratio
of Er/n class="Chemical">Fe: 1.50; found: 1.50. IR (KBr): ν = 3422 (br, m), 2958
(m), 2900 (m), 2866 (m), 2099 (s), 2076 (vs), 2058 (sh), 1582 (sh),
1535 (vs), 1481 (s), 1459 (m), 1420 (s), 1403 (s), 1354 (m), 1284
(m), 1226 (m), 1153 (w), 1098 (m), 1075 (m), 1024 (m), 920 (sh), 902
(m), 836 (w), 787 (w), 742 (w), 655 (m), 626 (m), 602 (m), 568 (m),
555 (sh), 492 (m), 426 (m) cm–1.
1c {Fe4Ho2}
Elemental analysis calcd.
for C58H118Fe4Ho2N16O26 (2008.9 g mol–1): C 34.68,
H 5.92, N 11.16%; found: C 32.77, H 5.87, N 11.51%; calcd. mass ratio
of Ho/n class="Chemical">Fe: 1.48; found: 1.52. IR (KBr): ν = 3440 (br, m), 2958
(m), 2899 (m), 2865 (m), 2099 (s), 2076 (vs), 2058 (sh), 1582 (sh),
1535 (vs), 1481 (s), 1459 (m), 1420 (s), 1402 (s), 1353 (m), 1284
(m), 1227 (m), 1152 (w), 1098 (m), 1074 (m), 1024 (m), 919 (sh), 902
(m), 787 (w), 742 (w), 653 (m), 625 (m), 602 (m), 567 (m), 554 (sh),
491 (m), 426 (m) cm–1.
Synthesis of [Fe4Tb2(O2CCMe3)6(N3)4(Htea)4] (1d) {Fe4Tb2}
A solution containing [Fe6O2(OH)2(O2CCMe3)12] (0.161 g, 0.1 mmol), n class="Chemical">Tb(NO3)3 (0.103 g, 0.3
mmol), NaN3 (0.038 g, 0.6 mmol), and H3tea (0.084
g, 0.6 mmol) in 10 mL of EtOH was refluxed for 30 min and then filtered.
The filtrate was stored in a closed vial at room temperature. After
two months, red and yellow crystals formed and were filtered off,
washed with EtOH, water, and MeCN, and dried in air. Yellow crystals
of 1d suitable a single-crystal X-ray analysis were separated
manually in the yield of 0.020 g, 7% (based on Fe).
Syntheses
of [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(CH2Cl2) (LnIII = Dy (2a);
Er (2b))
Recrystallization of 1a or 1b from hot dichloromethane yields well-shaped yellow
crystals of [n class="Chemical">Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(CH2Cl2) suitable for a single-crystal X-ray diffraction
analysis.
2a {Fe4Dy2}
Elemental analysis calcd. for C56H110Cl4Dy2Fe4N16O24 (2081.8 g mol–1): C 32.31, H 5.33, N 10.77%; found:
C 32.99, H 5.21, N 10.92%; calcd. mass ratio of Dy/Fe: 1.47; found:
1.52. IR (KBr): ν = 3417 (br, m), 2957 (m), 2900 (m), 2864 (m),
2098 (s), 2076 (vs), 1582 (sh), 1536 (vs), 1482 (s), 1459 (m), 1421
(s), 1402 (s), 1353 (m), 1284 (m), 1226 (m), 1153 (w), 1098 (m), 1068
(m), 1023 (sh), 918 (sh), 901 (m), 788 (w), 740 (w), 651 (m), 623
(sh), 603 (m), 569 (w), 553 (sh), 489 (m), 428 (m) cm–1.
2b {Fe4Er2}
Elemental
analysis calcd. for C56H110Cl4Er2Fe4N16O24 (2091.3 g mol–1): C 32.16, H 5.30, N 10.72%; found: C 32.18, H 5.15,
N 10.94%; calcd. mass ratio of Er/Fe: 1.50; found: 1.50. IR (KBr):
ν = 3422 (br, m), 2957 (m), 2900 (m), 2866 (m), 2098 (s), 2076
(vs), 1582 (sh), 1535 (vs), 1481 (s), 1459 (m), 1421 (s), 1403 (s),
1361 (m), 1284 (m), 1226 (m), 1154 (w), 1099 (m), 1076 (m), 1024 (m),
919 (sh), 902 (m), 788 (w), 741 (w), 655 (m), 626 (m), 602 (m), 569
(m), 555 (sh), 492 (m), 428 (m) cm–1.
Syntheses
of [Fe4Ln2(O2CCMe3)4(N3)6(Htea)4]·2(CH2Cl2)·2(EtOH) (LnIII = Dy (3a); Er (3b))
A solution of [Fe3O(O2CCMe3)6(EtOH)3]NO3·(n class="Chemical">EtOH) (0.208 g, 0.2 mmol), Ln(NO3)3·xH2O (0.3 mmol), NaN3 (0.058 g, 0.9 mmol), and H3tea (0.084 g, 0.6 mmol)
in 20 mL of a EtOH/CH2Cl2 (1:1) mixture was
refluxed for 4 h and then filtered. The filtrate was kept at room
temperature. After 10 d, yellow crystals of 3a and 3b were filtered off, washed with EtOH and ether, and dried
in air. Yield (based on Fe): 0.105 g, 34% (3a); 0.085
g, 28% (3b).
3a {Fe4Dy2}
Elemental analysis calcd. for C50H104Cl4Dy2Fe4N22O22 (2055.8 g mol–1): C 29.21, H 5.10, N 14.99%; found:
C 29.37, H 5.01, N 15.70%; calcd. mass ratio of Dy/Fe: 1.45; found:
1.51. IR (KBr): ν = 3421 (br, m), 2960 (m), 2900 (m), 2862 (m),
2097 (s), 2077 (vs), 2051 (s), 1535 (vs), 1483 (s), 1459 (m), 1422
(s), 1365 (m), 1352 (sh), 1285 (m), 1228 (m), 1149 (w), 1094 (m),
1072 (m), 1062 (sh), 1024 (m), 920 (sh), 900 (m), 788 (w), 740 (w),
649 (m), 626 (sh), 608 (m), 552 (m), 489 (m), 428 (m) cm–1.
3b {Fe4Er2}
Elemental
analysis calcd. for C50H104Cl4Er2Fe4N22O22 (2065.3 g mol–1): C 29.08, H 5.08, N 14.92; found: C 29.07, H 4.85,
N 15.93%; calcd. mass ratio of Er/Fe: 1.50; found: 1.47. IR (KBr):
ν = 3422 (br, m), 2961 (m), 2900 (m), 2862 (m), 2098 (s), 2077
(vs), 2051 (s), 1533 (vs), 1483 (s), 1460 (m), 1422 (s), 1365 (m),
1352 (sh), 652 (m), 626 (sh), 602 (m), 555 (m), 493 (m), 428 (m) cm–1.
Syntheses of [Fe18M6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·n(solvent),
(MIII = Dy (4, 4a), Gd (5), Tb (6), Ho (7), Sm (8), Eu (9), and Y (10))
4 {Fe18Dy6}
A solution of [Fe3On class="Chemical">(O2CCHMe2)6(H2O)3]NO3·2(MeCN)·2(H2O) (0.082
g, 0.087 mmol), Dy(NO3)3·6(H2O) (0.064 g, 0.18 mmol), sodium azide (0.02 g, 0.3 mmol), and triethanolamine
(0.26 g, 1.74 mmol) in MeOH/MeCN (1:1, 12 mL) was ultrasonicated for
35 min at room temperature and then filtered. The filtrate was allowed
to slowly evaporate at room temperature. This resulted in yellow crystals
suitable for single-crystal X-ray diffraction of the first polymorph
of 4 after one week. The crystals were filtered off,
washed with methanol, and dried under vacuum. Crystals of 4, [Fe18Dy6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·(MeCN)·8(H2O), lost their crystallinity upon
loss of crystal solvent. Yield: 0.03 g (30% based on Fe). Elemental
analysis calcd. for 4·8(H2O), C192H406Dy6Fe18N42O104, 6947.7 g mol–1): C, 33.19; H, 5.89; N, 8.47;
Dy, 14.03; Fe, 14.47%. Found: C, 32.01; H, 5.63; N, 8.08; Dy, 13.7;
Fe, 14.3%. IR (KBr): 3423 (br), 2964 (sh), 2862 (s), 2064 (vs), 2036
(sh), 1568 (m), 1472 (m), 1384 (sh), 1359 (m), 1284 (w), 1096 (vs),
898 (m), 750 (w), 1019 (sh), 592 (vw), 559 (vw), 457 (vw) cm–1.
4a {Fe18Dy6}
A solution
of [Fe3On class="Chemical">(O2CCHMe2)6(H2O)3]NO3·2(MeCN)·2(H2O) (0.082 g, 0.087 mmol) in EtOH (6 mL) was added to a solution of
Dy(NO3)3·6(H2O) (0.031 g, 0.09
mmol), sodium azide (0.02 g, 0.3 mmol), and triethanolamine (0.26
g,1.47 mmol) in MeOH (6 mL). The resulting solution was ultrasonicated
for 75 min at room temperature. Following filtration, the filtrate
was left for slow evaporation, producing yellow crystals of the second
polymorph, 4a. Yield: 0.028 g (26% based on Fe). Elemental
analysis calcd. for 4a [Fe18Dy6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·6(MeOH)·30(H2O), C198H474Dy6Fe18N42O132, 7536.5 g mol–1: C 31.55,
H 6.34, N 7.81; Dy, 12.94; Fe, 13.34%. Found: C 32.47, H 5.90, N 8.28;
Dy, 12.9; Fe, 14.5%. IR (KBr): 3423 (br), 2964 (sh), 2864 (s), 2064
(vs), 2036 (sh), 1630 (vw), 1565 (m), 1471 (w), 1415 (w), 1359 (w),
1284 (vw), 1162 (sh), 1096 (vs), 898 (m), 750 (vw), 560 (vw), 469
(vw) cm–1. Compounds 5–10 were obtained in an analogous manner to that of 4.
5 {Fe18Gd6}
Yield:
0.02 g (20% based on Fe), [n class="Disease">Fe18Gd6(O2CCHMe2)12(Htea)18(tea)6(N3)6](MeCN)·10(H2O). Elemental
analysis calcd. for 5·10(H2O), C194H413Gd6Fe18N43O106, 6952.2 g mol–1: C 33.17, H 5.94,
N 8.46; Gd, 13.57; Fe, 14.46%. Found: C 33.46, H 5.63, N 8.33; Gd,
13.4; Fe, 12.5%. IR (KBr): 3423 (br), 2962 (s), 2920 (sh), 2854 (m),
2064 (m), 2036 (sh), 1570 (m), 1473 (s), 1471 (w), 1415 (w), 1359
(w), 1281 (w), 1164 (w), 1075 (m), 898 (m), 748 (sh), 560 (vw) cm–1.
6 {Fe18Tb6}
Yield: 0.017 g (16% based on Fe), [Fe18Tb6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·2(MeCN)·8(MeOH)·7.5(H2O). Elemental analysis calcd. for 6·8(MeOH)·7.5(H2O), C200H437Fe18N42O111.5Tb6, 7173.6 g mol–1: C, 33.49; H, 6.14; N, 8.20; Tb, 13.29; Fe, 14.01%. Found: C, 33.08;
H, 6.00; N, 8.29; Tb, 13.4; Fe, 14.3%. IR (KBr): 3418 (br), 2963 (sh),
2861 (s), 2063 (vs), 2036 (sh), 1569 (m), 1472 (m), 1415 (m), 1358
(m), 1284 (w), 1163 (sh), 1086 (vs), 897 (m), 829 (sh), 749 (vw),
591 (w), 559 (w), 460 (w) cm–1.
7 {Fe18Ho6}
Yield: 0.017 g (16% based
on Fe). [n class="Chemical">Fe18Ho6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·2(MeCN)·3(MeOH)·11.5(H2O). Elemental
analysis calcd. for 7·3(MeOH)·11.5(H2O), C195H425Fe18Ho6N42O110.5, 7121.5 g mol–1: C, 32.89;
H, 6.01; N, 8.26%. Found: C, 33.43; H, 6.03; N, 7.47%. IR (KBr): 3423
(br), 2964 (sh), 2864 (s), 2063 (vs), 2036 (sh), 1571 (m), 1472 (m),
1416 (m), 1359 (w), 1285 (w), 1164 (sh), 1076 (vs), 898 (m), 750 (vw),
593 (vw), 559 (vw), 459 (vw) cm–1.
8 {Fe18Sm6}
Yield: 0.017 g (16% based
on Fe). [n class="Chemical">Fe18Sm6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·3(MeCN)·14(MeOH)·24(H2O). Elemental
analysis calcd. for 8·14(MeOH)·24(H2O), C206H494Fe18N42O134Sm6, 7611.7 g mol–1: C, 32.50;
H, 6.54; N, 7.73%. Found: C, 33.13; H, 5.87; N, 8.31%. IR (KBr): 3426
(br), 2960 (sh), 2920 (sh), 2852 (m), 2064 (m), 2036 (sh), 1629 (w),
1565 (m), 1469 (w), 1414 (m), 1359 (w), 1283 (vw), 1075 (s), 898 (m),
488 (vw) cm–1.
9 {Fe18Eu6}
Yield: 0.02 g (20% based on Fe).
[n class="Chemical">Fe18Eu6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·2(MeCN)·5(MeOH)·8(H2O). Elemental analysis
calcd. for 9·5(MeOH)·8(H2O), C197H426Eu6Fe18N42O109, 7044.7 g mol–1: C, 33.59; H, 6.03;
N, 8.35%. Found: C, 33.73; H, 6.36; N, 8.67%. IR (KBr): 3340 (br),
2962 (sh), 2920 (sh), 2853 (m), 2065 (m), 2036 (sh), 1631 (w), 1568
(m), 1470 (w), 1415 (w), 1359 (w), 1284 (vw), 1076 (s), 898 (m), 560
(vw), 489 (vw) cm–1.
10 {Fe18Y6}
Yield: 0.017 g (17% based on Fe).
[n class="Chemical">Fe18Y6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·3(MeCN)·9(MeOH)·11(H2O). Elemental analysis
calcd. for 10·11(H2O), C198H412Fe18N42O107Y6, 6560.2 g mol–1: C, 35.15; H, 6.33; N,
8.97%. Found: C, 33.68; H, 6.12; N 8.40%. IR (KBr): 3422 (br), 2963
(sh), 2963 (s), 2064 (vs), 2063 (sh), 1544 (m), 1471 (m), 1416 (m),
1359 (m), 1284 (w), 1164 (sh), 1097 (vs), 898 (m), 830 (sh), 750 (vw),
593 (vw), 559 (vw), 462 (vw) cm–1.
Magnetic
Measurements
Magnetic susceptibility data of 1–10 were recorded using a Quantum Design MPMS-5XL
SQUID magnetometer for direct current (dc) and alternating current
(ac) measurements. The polycrystalline samples were immobilized into
polytetrafluoroethylene capsules. The dc susceptibility data were
acquired as a function of the field (0.1–5.0 T) and temperature
(2.0–290.0 K). The ac susceptibility data were measured in
the absence of a static bias field in the frequency range of 3–1500
Hz (T = 2.0–50.0 K, Bac = 3 G). All data were corrected for diamagnetic contributions
from the sample holder and the compounds (as calculated from Pascal’s
constants; χdia (1 × 10–3 cm3 mol–1): 1a: −1.00, 1b: −1.01, 1c: −1.00, 1d: −0.99, 2a: −1.05, 2b: −1.04, 3a: −1.03, 3b: −1.03, 4: −3.46, 5: −3.45, 6: −3.45, 7: −3.47, 8: −3.43, 9: −3.43, 10: −3.24).
X-ray Crystallographic
Analyses
Single-crystal X-ray diffraction experiments for 1a–c and 2–10 were performed at 100(2) K on a Bruker diffractometer with APEX
II CCD detector using graphite-monochromated Mo Kα radiation.
Diffraction data sets for 1d were collected at 100(2)
K on an Oxford Xcalibur CCD with graphite-monochromatized Mo Kα
radiation. The summary of the data collection and the crystallographic
parameters of compounds 1–10 are
given in Table S1. Full crystallographic
data and refinement details of 1–10 are provided in Supporting Information. The positions of metal atoms were found by the direct methods.
The remaining atoms were located in an alternating series of least-squares
cycles and difference Fourier maps. All non-hydrogen atoms were refined
in full-matrix anisotropic approximation using the SHELX suite of programs.[14] All hydrogen atoms
were placed at idealized positions and were allowed to ride on the
neighboring atoms. Some hydrogen atoms of solvent molecules: H2O, MeCN, and MeOH, could not be located in 4–10. In 1a–c, methyl groups
for one monodentate pivalate ligand as well as the solvent ethanol
molecule were found to be disordered. In 2a and 2b, some of the methyl groups in carboxylates and solvate
dichloromethane molecule also revealed disorders. In 4–10 methyl groups for some isobutyrate ligands,
ethanol groups of triethanoamine ligands, as well as azide were found
to be also disordered. Therefore, SIMU, DELU, SADI, ISOR, and DFIX
restraints were used to deal with the disordered moieties in the structures
and to obtain reasonable geometrical parameters and thermal displacement
coefficients. Selected bond distances for 1–10 are listed in Supporting Information (Table S1), and hydrogen bonding interactions in 1–10 are presented in Table S2. Additional crystallographic
information and full experimental details can be found in Supporting Information.
Results and Discussion
The reaction of tri- or hexanuclear pivalate clusters with lanthanide
salts, azide, and triethanolamine (H3tea) ligands in EtOH
under reflux leads to wheel-shaped {Fe4Ln2}-type
cluster compounds [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(EtOH) (LnIII = Dy (1a), Er (1b), and Ho (1c)), [Fe4Tb2(O2CCMe3)6(N3)4(Htea)4] (1d) in 7–43% yield, and
the homometallic octanuclear FeIII cluster [Fe8O3(O2CCMe3)6(N3)3(tea)(Htea)3]·0.5(EtOH) in 30% yield.
Recrystallization of 1a and 1b from hot
dichloromethane afforded crystals of [Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]·2(CH2Cl2) (LnIII = Dy (2a) and Er (2b)). Boiling
of the same starting materials in mixture of CH2Cl2/EtOH (1:1) resulted in the heterometallic wheels [Fe4Ln2(O2CCMe3)4(N3)6(Htea)4]·2(EtOH)·2(CH2Cl2) (LnIII = Dy (3a) and
Er (3b), where two terminal carboxylic groups were replaced
by two azide ligands.By changing the reaction conditions (using
MeOH/MeCN or MeOH/EtOH (1:1) mixtures at room temperature) and introducing
trinuclear FeIII μ3-oxo isobutyrate as
precursor, we obtained ultralarge {Fe18M6}-type
wheels, isolated as [Fe18M6(O2CCHMe2)12(Htea)18(tea)6(N3)6]·n(solvent) for MIII = Dy (4, 4a), Gd (5), Tb
(6), Ho (7), Sm (8), Eu (9), and Y (10), which accumulated in 16–30%
yields over one week.The IR spectra of 1–10 display strong bands in the 1565–1533 cm–1 and 1417–1406 cm–1 regions that arise from
the asymmetric and symmetric vibrations of the carboxylate groups
from the coordinated pivalates or isobutyrates, respectively. The
C–H asymmetric and symmetric stretching vibrations for methyl
groups of carboxylates are observed between 2959 and 2870 cm–1, while the asymmetric and symmetric bending vibrations for these
groups produce a strong single band at 1483 cm–1 and a doublet (1374–1355 cm–1), respectively.
The uncoordinated hydroxyl groups of Htea2– and
solvent EtOH or H2O molecules result in broad bands (3420–3375
cm–1). A sharp peak at 2062 cm–1 corresponds to the N≡N stretching vibrations of azide ligands
in all compounds.TGA of 1–10 showed that all compounds exhibit similar thermal decomposition
behavior and release the remaining solvent molecules below 190 °C
(Figures S24–S38). When further
heated, the wheels start to decompose in several weakly resolved steps
until ∼800 °C. The decomposition of organic ligands in
the smaller {Fe4Ln2} wheels is accompanied by
exothermic peaks at 209 and 222 °C (1a), 200 °C
(2a), 199 °C (3a), 212 °C (1b), 212 °C (2b), 199 °C (3b), 189 °C (1c), and 187 °C (2c).
Structural Description
{Fe4Ln2} Wheels
Single-crystal X-ray diffraction analysis reveals that 1–3 comprise wheel-shaped centrosymmetric hexanuclear
[Fe4Ln2(O2CCMe3)6(N3)4(Htea)4]
(1 and 2) and [Fe4Ln2(O2CCMe3)4(N3)6(Htea)4] (3) clusters (see Figure ) built from four
FeIII and two LnIII ions that are linked by
an array of bridging carboxylate, azide, and aminopolyalcoholato-based
ligands into a cyclic structure with a cavity, and crystal solvent
molecules: two EtOH per formula unit in 1a–c, two CH2Cl2 in 2, and
two EtOH and two CH2Cl2 in 3. The
asymmetric units of 1 and 2 contain one
LnIII and two FeIII centers, two monodentate
and one bridging carboxylate groups, two Htea2– and
two end-on bridging N3– ligands, and
one solvent molecule, whereas the asymmetric unit of 3 contains one LnIII and two FeIII ions, one
monodentate and one bridging carboxylate groups, two Htea2–, one monodentate and two end-on bridging N3– ligands, and two solvent molecules.
Figure 2
Structures of wheel-shaped {Fe4Ln2} clusters in 1 and 2 (a)
and 3 (b). Color scheme: C(tea), dark gray; C(pivalate),
light gray; O, red; N(tea), blue; N(azide), light blue); Ln, green;
Fe, dark yellow spheres. Monodentate N3– groups in 3 are highlighted as light blue spheres.
Hydrogen atoms are omitted for clarity.
Structures of wheel-shaped {Fe4Ln2} clusters in 1 and 2 (a)
and 3 (b). Color scheme: C(tea), dark gray; C(pivalate),
light gray; O, red; N(tea), blue; N(n class="Chemical">azide), light blue); Ln, green;
Fe, dark yellow spheres. Monodentate N3– groups in 3 are highlighted as light blue spheres.
Hydrogen atoms are omitted for clarity.
Each FeIII atom in 1 and 2 resides in a distorted octahedral N2O4 environment
made from two azide ligands [Fe–Nazide, 2.085(6)–2.152(7)
Å], two O sites from two carboxylates [Fe–Ocarb, 1.963(4)–2.071(6) Å], and two alcoholate sites of two
Htea2– [Fe–Oalc, 1.937(4)–1.982(5)
Å]. In 3, the two independent FeIII centers
adopt different environments: one FeIII in a distorted
octahedral N2O4 environment analogous to 1 and 2 [Fe–Nazide, 2.095(10)–2.137(10)
Å, Fe–Ocarb, 1.969(9)–2.053(8) Å,
Fe–Oalc, 1.925(8)–1.949(12) Å]. The
second FeIII adopts a distorted octahedral N3O3 geometry arising from three N atoms of three azide
ligands [Fe–Nazide, 2.036(10)–2.135(13) Å],
one carboxylate O atom [Fe–Ocarb, 2.018(14)–2.031(9)
Å], and two alcoholates from two Htea2– [Fe–Oalc, 1.933(12)–1.940(8) Å]. The azido-bridged Fe···Fe
distances are 3.221(2)–3.279(2) Å, with Fe–Nazide–Fe bridging angles of 98.4(2)–101.6(4)°
in 1–3.Each LnIII atom in 1–3 is eight-coordinated
by donor atoms of two Htea2– molecules: two N [Ln–N,
2.602(5)–2.612(5) (1a); 2.586(5)–2.591(5)
(1b); 2.579(5)–2.581(5) (1c); 2.598(5)–2.601(5)
(2a); 2.603(7)–2.607(7) (2b); 2.597(13)–2.598(13)
(3a); 2.576(9)–2.584(9) (3b) Å]
and six O sites [Ln–O, 2.283(3)–2.414(3) (1a); 2.272(4)–2.389(4) (1b); 2.258(4)–2.380(4)
(1c); 2.284(4)–2.396(4) (2a); 2.270(6)–2.397(5)
(2b); 2.294(12)–2.417(11) (3a); 2.278(8)–2.367(8)
(3b) Å]. Neighboring FeIII polyhedra
share a common N–N edge, while LnIII–FeIII polyhedra share O–O edges in 1–3. Nearest-neighbor Fe···Fe distances range
from 3.220(2) to 3.279(2) Å, Fe···Ln from 3.361(1)
to 3.398(1) Å.In 1–3,
the uncoordinated pivalate C=O group and the OH group in Htea2– as well as the solvate EtOH and CH2Cl2 molecules participate in an extensive network of hydrogen
bonds with both strong O–H···O and O–H···N
interactions, and C–H···O, C–H···N,
and C–H···Cl contacts (for more details see Supporting Information and Figures S9–S15);
these interactions significantly influence the crystal packing.[15]In the crystal lattices of 1–3, coplanar wheels are stacked along columns.
Interestingly, in 1 and 2 these columns
are related by translations b and c and thus have the same orientation of wheels, while in 3, the columns of stacked wheels alternate in their orientation with
respect to the position of Fe or Ln ions (Figures
S16–S19).
{Fe18M6} Wheels
Single-crystal X-ray analyses revealed that compounds 4, 5, 6, 7, 9,
and 10 crystallize in the triclinic space group P1̅ and have C symmetry, whereas 4a and 8 crystallize
in the trigonal space group R3̅ and have C3 crystallographic symmetry.
The molecular core structures in 4–10 involve 18 FeIII and 6 LnIII/YIII ions interconnected by six isobutyrate and 24 amino alcohol ligands
into a ring with alternating three FeIII and one LnIII/YIII ions. The wheels are slightly puckered
with Fe···Fe···Fe angles of ca. 140°
and Fe···Ln/Y···Fe angles of ca. 114°
and form a central cavity with a diameter of ca. 1 nm, Figure , while the external diameter
of ultralarge wheels is ca. 3.7 nm. The crystal structures of ultralarge
wheels display infinite channels filled by solvent molecules (Figure c) parallel to the
crystallographic c axis for 4a and 8 and along a for the remaining structures.
Upon removal of neutral solvent molecules, the triclinic structures
reveal a large total potential solvent area volume of ca. 27%, and
trigonal one ca. 44% per unit cell volume, as calculated by PLATON.
Figure 3
(a) Molecular
structure of {Fe18M6} wheels in 4–10. Fe/Ln/Y centers are highlighted as their
coordination polyhedra (color code as in Figure ). Hydrogen and disordered atoms are omitted
for clarity. (b) Side view of the {Fe18M6} metal
skeleton; the transparent yellow plane serves to emphasize the small
deviations from a fully planar ring structures. (c) Formation of infinite
channels in the trigonal crystal lattices of 4a and 8 along c (for triclinic structures see Supporting Information).
(a) Molecular
structure of {Fe18M6} wheels in 4–10. Fe/Ln/Y centers are highlighted as their
coordination polyhedra (color code as in Figure ). Hydrogen and disordered atoms are omitted
for clarity. (b) Side view of the {Fe18M6} metal
skeleton; the transparent yellow plane serves to emphasize the small
deviations from a fully planar ring structures. (c) Formation of infinite
channels in the trigonal crystal lattices of 4a and 8 along c (for triclinic structures see Supporting Information).All Fe centers in 4–10 are
in a distorted octahedral ligand environment: 12 Fe atoms have a NO5donor set, and the remaining six Fe atoms have a N2O4donor set. The NO5 environments are formed
by a bridging isobutyrate (one O), three amino alcoholate ligands
(three O), and a doubly deprotonated Htea2– (one
N) or by one tea3– and two doubly deprotonated Htea2– (five O, one N). The N2O4 set
stems from three amino alcohols (four O), from a doubly deprotonated
Htea2– (one N) and one azide N atom. All of the
Fe–O bond distances in 4–10 are in the range of 1.876(12)–2.060(5) Å (Fe–Ocarb: 2.021(11)–2.060(5), Fe–Oalc:
1.876(12)–2.051(11) Å). Fe–Nazide distances
range from 1.992(1) to 2.064(12) Å, whereas Fe–Nalc distances are distinctly longer (2.201(7)–2.327(15) Å).
All Ln and Y sites are in distorted square-antiprismatic NO7 environments (Ln/Y–O: 2.259(6)–2.469(6), Ln/Y–N:
2.602(7)–2.698(16) Å) made from two Htea2– and one tea3– (five O, one N), and O atoms of
the bridging and monodentate carboxylates. The latter is additionally
linked by a strong intramolecular O–H···O bond.
Fe···Fe distances amount to 3.162(3)–3.222(4)
Å, Ln/Y···Fe to 3.354(1)–3.534(1) Å.
Neighboring coordination polyhedra in 4–10 share common O–O edges (Figure a).
Magnetic Properties
Magnetic ac and dc susceptibility
data of compounds 1–3 were collected
using a SQUID magnetometer. The magnetic dc data of 1–3 are presented as χmT versus T plots in Figure . At 290 K, the χmT data of 1–3 reach values
in the interval 39–46 cm3 K mol–1. All values are in the upper half of the ranges that are expected
for four FeIII and the two respective LnIII non-interacting,
high-spin centers: 44.2 (1a), 44.7 (2a),
45.8 (3a), expected:[16] 42.3–46.1
cm3 K mol–1. Found: 39.9 (1b), 39.8 (2b), 39.6 (3b), expected: 38.4–40.6;
44.6 (1c), expected: 42.8–45.6; 41.1 (1d), expected: 39.8–42.0 cm3 K mol–1. χmT steadily increases with decreasing T, exhibiting a steep slope at T ≤
50 K revealing dominant ferromagnetic exchange interactions within
the compounds, potentially between all pairs of next neighbors. Note
that compounds 3a and 3b containing additional
azide ligands exhibit maxima at larger χmT values than their analogues, pointing to an effective
enhancement of ferromagnetic exchange interactions due to these ligands.
Figure 4
Temperature
dependence of χmT at B = 0.1 T of 1–3.
Temperature
dependence of χmT at B = 0.1 T of 1–3.Susceptibility measurements (ac) were performed
to probe slow magnetization relaxation of compounds 1–3. At zero static dc field, only 1c and 1d showed significant out-of-phase signals down
to 2 K and up to 1500 Hz. Out-of-phase signals could be detected up
to 3 K for 1c (see Figure S39). Because of the constraints of the parameter range of our experimental
setup, however, the collected data were insufficient for further phenomenological
modeling, for example, via an Arrhenius-like expression. For 1d, out-of-phase signals at zero bias field are present up
to 5.5 K (see Figures and S40). A generalized Debye expression[17] is simultaneously fit to the frequency-dependent
in-phase χ′m and out-of-phase χ″m components of the magnetic ac susceptibility in the range
of 2.2 K ≤ T ≤ 4.2 K. Analysis of the
resulting attempt time τ versus T–1 data in the 2.6–4.0 K interval in terms of an Arrhenius expression
(τ = τ0·exp(Ueff/(kB·T))) yields
parameters typically observed[7,18] for Orbach relaxation
processes: τ0 = (1.2 ± 0.2) × 10–9 s and Ueff = (26.4 ± 0.4) cm–1.
Figure 5
(top) Cole–Cole plot of in-phase χ′m and out-of-phase χ″m susceptibility
of 1d at various temperatures and zero static bias field:
experimental data (●), least-squares fits (solid lines). (bottom)
Corresponding Arrhenius plot of relaxation time τ vs T–1 (2.2 K ≤ T ≤ 4.2 K), solid line shows fit to an Orbach relaxation expression
(2.6 K ≤ T ≤ 4.0 K).
(top) Cole–Cole plot of in-phase χ′m and out-of-phase χ″m susceptibility
of 1d at various temperatures and zero static bias field:
experimental data (●), least-squares fits (solid lines). (bottom)
Corresponding Arrhenius plot of relaxation time τ vs T–1 (2.2 K ≤ T ≤ 4.2 K), solid line shows fit to an Orbach relaxation expression
(2.6 K ≤ T ≤ 4.0 K).The magnetic properties of 4–10 were
investigated by both dc and ac measurements. Analysis of 10 (comprising diamagnetic yttrium centers instead of lanthanides)
reveals the magnetism of the FeIII constituents. In a first
approximation, these form six linear Fe3 trimers, separated
by the YIII centers. Assuming a symmetric cluster and normalizing
the data to such a linear trimer results in the magnetic properties
shown in Figure as
open circles.
Figure 6
Magnetic properties of 10, normalized to
a FeIII3 unit: Temperature dependence of χmT at B = 0.1 T (○:
experimental data, straight lines: calculated data (Seff = 5/2 spin system)); inset: molar magnetization Mm vs applied field B at T = 2.0 K.
Magnetic properties of 10, normalized to
a FeIII3 unit: Temperature dependence of χmT at B = 0.1 T (○:
experimental data, straight lines: calculated data (Seff = 5/2 spin system)); inset: molar magnetization Mm vs applied field B at T = 2.0 K.The χmT value of 7.99 cm3 mol–1 at 290 K (B = 0.1
T) is well below the range of 12.19–13.51 cm3 mol–1 that is expected for three non-interacting FeIII centers.[16] By lowering the temperature,
χmT continuously decreases to 3.59
cm3 mol–1 at 14 K and, below that point,
more rapidly to 2.36 cm3 mol–1 at 2 K.
The molar magnetization Mm versus B at 2 K is likely to saturate in the range of 5–7 NAμB revealing an upper limit
of the total spin of the ground state, which is below 15/2, the maximum
spin value for three ferromagnetically coupled spin-like FeIII centers. All three observations combined indicate predominant anti-ferromagnetic
exchange interactions within each Fe3 unit.The magnetic
data can thus be modeled as six weakly coupled, identical linear trimers
of three spin-5/2 centers for slightly distorted octahedrally coordinated
FeIII centers using the computational framework CONDON
2.0.[19] Potential interactions between neighboring
Fe3 groups are accounted for by an (isotropic) molecular
field approach. The best fit (SQ = 4.1%) reproduces
the experimental data as shown in Figure , yielding JFe3 = −12.4 cm–1 (Hex = −2JFe3 (S1·S2 + S2·S3), g =
2.0) for both intratrimer exchange interaction energies and λmf = −0.412 mol cm–3, that is, zJinter = −0.21 cm–1. According to the fit, the magnitude and sign of JFe3 thus reveal anti-ferromagnetic exchange interactions
within an Fe3 group characterized by a ground state of Seff,total = 5/2. Additionally, such a trimer
and the molecular field are linked by weak anti-ferromagnetic interactions.
Note that the obvious deviation of experiment and fit in the 5–90
K interval is presumably due to a slightly more complex exchange interaction
scheme than the employed model and, to a much lesser or even negligible
extent, caused by ligand-field effects linked to the distortion of
the octahedral site symmetry of the FeIII centers. Nevertheless,
the main contributions of the exchange interactions are reasonably
described by the presented model.The magnetic data of 10 are compared to the data for 4–9 in Figure . The χmT values at 290 K of the
compounds are ca. 25–30 cm3 K mol–1 below the expected values[19] for the respective
non-interacting centers of a {Fe18Ln6} ring
(4: 140.5 cm3 K mol–1, 5: 101.2 cm3 K mol–1, 6: 117.3 cm3 K mol–1, 7:
127.2 cm3 K mol–1, 8: 57.4
cm3 K mol–1, 9: 58.9 cm3 K mol–1, 10: 47.9 cm3 K mol–1). It is reasonable to infer an almost
constant high-temperature χmT offset
for all compounds, including 10, that primarily arises
from the anti-ferromagnetic interactions within the linear FeIII3 trimers discussed above. Additionally, the
χmT curves of 4, 5, 6, and 7 also show ferromagnetic
interactions within the ring structure revealed by the distinct maxima
at ∼3–4 K. In case of 8 and 9, no maxima are observed, which is due to the almost temperature-independent
paramagnetism (TIP)-like behavior of the SmIII centers
(8, multiplets characterized by energetically very close
states) and EuIII centers (9, m = 0 ground state), respectively.[16] The ferromagnetic exchange interactions are
most likely between the terminal FeIII centers of the linear
trimers and the next-neighbor LnIII centers, since the
corresponding signal is only observed in the presence of the centers
that are not characterized by temperature-independent contributions.
The similar decrease in χmT upon
cooling the compounds from 290 K to ca. 100 K is thus potentially
only caused by anti-ferromagnetic coupling interactions between the
FeIII centers of the trimer units.
Figure 7
Temperature dependence
of χmT at B = 0.1
T of 4–10.
Temperature dependence
of χmT at B = 0.1
T of 4–10.The molar magnetization measured at different applied fields B and at 2.0 K (see Figure S41) is consistent with the conclusions derived from the χmT data: the magnetizations of all compounds
approximately add up to the magnetization of 10 and up
to 6 times the maximum contribution of the respective Ln (gJ NAμB). Note that Mm of 10 hints at a saturation value that is much lower than the
maximum possible value of 90 NAμB for 18 FeIII centers indicating the anti-ferromagnetic
interactions within a ring. For 4, 5, 6, and 7, the deviations from the maximum Ln
contributions may be due to anisotropic Fe–Ln exchange interactions,
due to different magnitudes of the exchange coupling energies or due
to single ion contributions of the LnIII centers, that
is, the ground state of the LnIII centers may be characterized
by m ≠ ±J and thus by a less-than-maximum magnetic moment.In addition to the dc measurements, 4–10 were measured in an ac magnetic field in absence of a static field.
Only 4 and 6 show any out-of-phase signal
and thus slow relaxation above 2.0 K and below 1500 Hz (see Figures S42 and S43, respectively). Because of
the limits of our experimental setup, only small temperature intervals
with χ″m ≠ 0 and no distinct features
like maxima in χ″m are observed; thus, the
ac data are insufficient for further analysis of relaxation characteristics.
We thus can only conclude that slow relaxation occurs for 4 at T ≤ 2.3 K and for 6 at T ≤ 2.6 K.
Conclusion
In
summary, a synthetic strategy exploiting the potential of structure-directed
semiflexible bridging ligands and solvent effects has been successfully
applied to design and prepare a new family of hexanuclear {Fe4Ln2} wheels as well as the currently largest d/f
coordination wheels, the tetraicosanuclear {Fe18Ln6} clusters. {Fe4Tb2} (1d), {Fe4Ho2} (1c), {Fe18Dy6} (4), and {Fe18Tb6} (6) were found to exhibit slow magnetization relaxation.
Authors: Anastasios J Tasiopoulos; Alina Vinslava; Wolfgang Wernsdorfer; Khalil A Abboud; George Christou Journal: Angew Chem Int Ed Engl Date: 2004-04-13 Impact factor: 15.336
Authors: Mengyuan Li; Ayuk M Ako; Yanhua Lan; Wolfgang Wernsdorfer; Gernot Buth; Christopher E Anson; Annie K Powell; Zheming Wang; Song Gao Journal: Dalton Trans Date: 2010-02-26 Impact factor: 4.390
Authors: Philippa King; Theocharis C Stamatatos; Khalil A Abboud; George Christou Journal: Angew Chem Int Ed Engl Date: 2006-11-13 Impact factor: 15.336
Authors: Joris van Slageren; Roberta Sessoli; Dante Gatteschi; Andrew A Smith; Madeleine Helliwell; Richard E P Winpenny; Andrea Cornia; Anne-Laure Barra; Aloysius G M Jansen; Eva Rentschler; Grigore A Timco Journal: Chemistry Date: 2002-01-04 Impact factor: 5.236
Authors: Muralee Murugesu; James Raftery; Wolfgang Wernsdorfer; George Christou; Euan K Brechin Journal: Inorg Chem Date: 2004-07-12 Impact factor: 5.165
Authors: Ruslan A Polunin; Igor S Evstifeev; Olivier Cador; Stéphane Golhen; Konstantin S Gavrilenko; Anton S Lytvynenko; Nikolay N Efimov; Vadim V Minin; Artem S Bogomyakov; Lahcène Ouahab; Sergey V Kolotilov; Mikhail A Kiskin; Igor L Eremenko Journal: Molecules Date: 2021-02-15 Impact factor: 4.411