Synthesis, Structure, and Magnetic Properties of Dy₂Co₂<i>L</i>₁₀(bipy)₂ and <i>Ln</i>₂Ni₂<i>L</i>₁₀(bipy)₂, <i>Ln</i> = La, Gd, Tb, Dy, and Ho: Slow Magnetic Relaxation in Dy₂Co₂<i>L</i>₁₀(bipy)₂ and Dy₂Ni₂<i>L</i>₁₀(bipy)₂

The 3,5-dichlorobenzoate anion, <i>L</i>^–, serves as a bridging ligand and 2,2′-bipyridine, bipy, as a terminal bidentate ligand to yield, through hydrothermal syntheses, the tetranuclear clusters Dy₂Co₂<i>L</i>₁₀(bipy)₂, <b>1</b>, and <i>Ln</i>₂Ni₂<i>L</i>₁₀(bipy)₂, where <i>Ln</i> is the trivalent La, <b>2</b>, Gd, <b>3</b>, Tb, <b>4</b>, Dy, <b>5</b>, or Ho, <b>6</b>, ion. Single-crystal X-ray diffraction reveals that the six complexes are all isomorphous with the monoclinic <i>P</i>2₁/<i>c</i> space group and with lattice parameters that decrease with the lanthanide contraction. The two cobalt­(II) or nickel­(II) and two <i>Ln</i>(III) cations are linked by the 10 <i>L</i>^– anions to generate Dy₂Co₂ or <i>Ln</i>₂Ni₂ 3d–4f cationic heteronuclear clusters with a slightly bent Co···Dy···Dy···Co or Ni···<i>Ln</i>···<i>Ln</i>···Ni arrangement. Direct current magnetic susceptibility studies reveal that the complexes are essentially paramagnetic, with room-temperature χ_M<i>T</i> values close to the expected values for two cobalt­(II) or nickel­(II) and two <i>Ln</i>(III) cations. The temperature dependence of χ_M<i>T</i> for <b>1</b> and <b>5</b> is well reproduced by <i>ab initio</i> calculations with the inclusion of weak magnetic exchange between the cobalt­(II) or nickel­(II) and a dysprosium­(III) and between two dysprosium­(III) ions. The calculated magnetic exchange parameters are <i>J</i>_Dy–Co = 0.2 cm^–1 and <i>J</i>_Dy–Dy = 0.02 cm^–1 for <b>1</b> and <i>J</i>_Dy–Ni = −0.2 cm^–1 and <i>J</i>_Dy–Dy = 0.03 cm^–1 for <b>5</b>. Alternating current magnetic susceptibility studies reveal that <b>1</b> and <b>5</b> exhibit slow magnetic relaxation with effective energy barriers, <i>U</i>_eff, for the reversal of the magnetization for <b>1</b> of 82(2) cm^–1 in a 0 Oe dc bias field and 79.4(5) cm^–1 in a 1000 Oe dc bias field and, for <b>5</b>, 73(1) cm^–1 in a 0 dc bias field; the calculated energies of 66.1(1) and 61.0(1) cm^–1 for the first excited spin–orbit state of dysprosium­(III) in <b>1</b> and <b>5</b> agree rather well with these effective energy barriers. The entire Arrhenius plots of the logarithm of τ, the relaxation rate of the magnetization in <b>1</b> and <b>5</b>, have been fit with contributions from quantum tunneling, direct Raman scattering, and Orbach thermal processes. The observation of a low-temperature magnetization reversal mechanism in <b>5</b> but not in <b>1</b> may be understood through the calculated exchange energy spectrum in their ground state