Whither Second-Sphere Coordination?

Abstract

Open AccessCCS ChemistryMINI REVIEW1 Mar 2022Whither Second-Sphere Coordination? Wenqi Liu, Partha J. Das, Howard M. Colquhoun and J. Fraser Stoddart Wenqi Liu Department of Chemistry, Northwestern University, Evanston, IL 60208 Google Scholar More articles by this author , Partha J. Das Department of Chemistry, Northwestern University, Evanston, IL 60208 Google Scholar More articles by this author , Howard M. Colquhoun *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, University of Reading, Reading RG6 6DX Google Scholar More articles by this author and J. Fraser Stoddart *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, Northwestern University, Evanston, IL 60208 School of Chemistry, University of New South Wales, Sydney, NSW 2052 Department of Chemistry, Stoddart Institute of Molecular Science, Zhejiang University, Hangzhou 310027 ZJU-Hangzhou Global Scientific and Technological Innovation Center, Hangzhou 311215 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101286 SectionsAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail The properties of coordination complexes are dictated by both the metals and the ligands. The use of molecular receptors as second-sphere ligands enables significant modulation of the chemical and physical properties of coordination complexes. In this minireview, we highlight recent advances in functional systems based on molecular receptors as second-sphere coordination ligands, as applied in molecular recognition, synthesis of mechanically interlocked molecules, separation of metals, catalysis, and biomolecular chemistry. These functional systems demonstrate that second-sphere coordination is an emerging and very promising strategy for addressing societal challenges in health, energy, and the environment. Download figure Download PowerPoint Introduction Coordination chemistry,1 the investigation of linking ligands directly to a central atom—most often a metal center—is at the heart of modern inorganic, organic, and materials chemistry. The specific combinations of ligands and metals dictate the structures and properties of the resulting coordination complexes, which are applied in numerous technologies, including metal–organic frameworks (MOFs),2–16 optical and magnetic materials,17,18 catalysis,19,20 and biomedical theranostics.21–24 In the resulting coordination complexes, the ligands linked directly to the metal center are referred to (Figure 1) as the first sphere of coordination for the metal.25–28 Meanwhile, another set of ligands can bind to the first-sphere ligands of the coordination complex through noncovalent bonding interactions, leading to second-sphere adducts. Thus, second-sphere coordination of metal centers affords adducts that are essentially complexes of complexes. Figure 1 | Structural formulas depicting the similarity of the three-point binding models for (a) a primary alkyammonium ion [RNH3]+ and (b) a transition-metal ammine. (c) Graphical representation of the concept of second-sphere coordination. A transition metal (M) is coordinated with first-sphere ligands L1, which interact with second-sphere ligands L2 through noncovalent bonding interactions. Adapted with permission from ref 35. Copyright 1983 Royal Society of Chemistry. Download figure Download PowerPoint The concept of second-sphere coordination (Figure 1c) was advanced in the first instance by Chemistry Nobel Prize Laureate Alfred Werner29,30 in 1913. He showed that this concept could explain a number of experimental observations that could not be understood solely on the basis of first-sphere coordination.29 These observations included (1) the formation of adducts between amines and coordinatively saturated complexes such as [M(acac)3]n+, (2) the presence of solvents in the crystal structures of many coordination complexes, and (3) the solvent and counterion-dependent properties associated with the optical rotations of chiral coordination complexes. Beginning in the early 1980s, we sought to gain insight into the nature of second-sphere coordination by investigating the interactions between transition-metal complexes and synthetic molecular receptors.31 Crown ethers are well-known to form complexes with primary alkylammonium ions, [RNH3]+, through hydrogen bonding and ion-dipole interactions.32–34 It occurred to us that the geometric and electronic (stereoelectronic) features of an [RNH3]+ ion closely resemble (Figures 1a and 1b) those of a transition-metal ammine complex,35 [MLx(NH3)]n+. The recognition of this similarity led to our initial investigations of crown ether receptors as second-sphere ligands, resulting in the isolation of a novel series of adducts between crown ethers and transition-metal complexes,36–39 whose superstructures were quickly established by X-ray crystallography. Figure 2 presents a collage [(a)–(s)] of some of the solid-state superstructures that are reproduced from a review25 that was published in Angewandte Chemie on “Second-Sphere Coordination—A Novel Rôle for Molecular Receptors” in 1986. The review was the first one to be published in the chemical literature with a copious use of color: at that time, it cost us the equivalent of £10,000. One of the most remarkable examples (Figures 2j and 2k) was the 1∶1 adduct formed38 between dibenzo[30]crown-10 (DB30C10) and [Pt(bpy)(NH3)2]2+. The fascinating feature of this superstructure is that, in addition to the three pairs of [N–H⋯O] hydrogen bonds between the oxygen atoms in the crown ether and the hydrogen atoms of the NH3 ligands, there are close, complementary [π⋯π] stacking (charge transfer) interactions between the two π-electron-rich aromatic (catechol) rings in the receptor molecule and the π-electron-poor 2,2′-bipyridine ligand in the platinum-based substrate. This second-sphere coordination adduct was shown by 1H NMR spectroscopy to exist as a stable complex in solution as well as in the solid state.38 The high stability of this adduct is sustained by multiple weak interactions between the receptor and the ligands on the transition-metal complex. In contrast, conventional second-sphere ligands, such as water or alcohols, only form weak adducts with metal complexes through monovalent binding, and the well-defined superstructures observed crystallographically in the solid state are not sustained in solution. For example, in the crystal superstructure38 of a lanthanum complex [La(tptz)(NO3)3(H2O)]•2C2H5OH, the two ethanol molecules form hydrogen bonds with the coordinated H2O molecule. Although the solid-state superstructure is well-defined, there is no suggestion that it exists in solution. Molecular receptors, when employed as second-sphere ligands, lead to the formation of stable, robust, and predictable adducts in contrast to those adducts formed by weakly binding ligands observed only in the solid state. Figure 2 | A sampling of solid-state superstructures of 1∶1 adducts between transition-metal complexes and macrocyclic receptors taken directly from the 1986 review on second-sphere coordination published in Angewandte Chemie, where the superstructures are presented with space-filling or ball-and-stick models for (a) [{trans-Pt(PMe3)Cl2(NH3)}2•l8C6]; (b) [{cis-Pt(NH3)2Cl2•dma}2•l8C6]; (c and d) [Cu(NH3)4(H2O)•18C6]n2n+; (e) [Rh(cod)(NH3)2•DB21C7]+; (f) [Rh(nbd)(NH3)2•DB24C8]+; (g) [diammine-1,1-cyclobutanecarboxylatoplatinum(II)-α-CD]; (h) [Pt(en)2•18C6]n2n+; (i) diamminebis-(1,5-cyclooctadiene)(μ-l,4-l0,l3-tetraoxa-7,l6-diazacyclooctadecane-N7,N16)-dirhodium bis(hexafluorophosphate); (j and k) [Pt(bpy)(NH3)2•DB30C10]2+; (l) [SnCl4(H2O)2•18C6][H2O]2; (m) [{trans-Ir(CO)(CH3CN)(PPh3)2}2•18C6]2+; (n) [Rh(cod)(NH3)2•α-CD]+; (o) [Mn(H2O)6•18C6]nn+ (ClO4− salt); (p and q) [trans-Pt(PMe3)Cl2(NH3)•DB18C6]; (r) [Rh(cod)(NH3)2•DB24C8]+; (s) [Rh(cod)(NH3)2•DB30C10]+. The X-ray superstructure (j and k) of [Pt(bpy)(NH3)2•DB30C10]2+ inside the highlighted rectangle will be discussed in Box 1. Adapted with permission from ref 25. Copyright 1986 John Wiley and Sons. Download figure Download PowerPoint It may be noted that the superstructure of the bipyridine-based second-sphere adduct shown in Figure 2 as a space-filling representation (j) and as a ball-and-stick representation (k) provided (Box 1) inspiration for subsequent research into complex formation between dibenzo-crown ethers and the dicationic bipyridinium herbicides Diquat, derived from 2,2′-bipyridine, and Paraquat, derived from 4,4′-bipyridine. The success of the latter investigation in turn led to the discovery of the tetracationic cyclophane, cyclobis(paraquat-p-phenylene), known as Blue Box, which has since proved a hugely valuable component in developing the field of mechanically interlocked molecules (MIMs) and eventually artificial molecular machines (AMMs). Box 1. A 40-year research trail illustrating how one observation in 1981 led to another line of research and so on—and how this dynamic is as sustained over four decades. Following our initial work on using synthetic molecular receptors to bind coordination complexes, there was a surge in the study of second-sphere coordination. Numerous macrocyclic and polycyclic crown ethers,39–50 cyclodextrins (Figures 3a and 3b),51–65 calixarenes (Figure 3e),66–71 resorcinarenes (Figure 3d),72 pillararenes (Figure 3f),73 cucurbiturils (Figure 3c),74 and cyclophanes75–79 all proved to be efficient second-sphere ligands and even, in some examples, acted simultaneously as both first- and second-sphere coordination ligands (Figure 2i).80–83 As a result, many second-sphere coordination adducts were discovered by exploiting the whole range of noncovalent bonding interactions, including hydrogen bonding, charge transfer, [π⋯π] stacking, hydrophobic effects, and van der Waals interactions. Summaries of these early investigations of second-sphere coordination chemistry may be found in several reviews,25–28,84–90 published by us as well as by other groups. Figure 3 | Representative examples of solid-state superstructures for second-sphere coordination adducts reported by other researchers. (a) Tubular and space-filling representation of a 2∶2∶1 adduct of γ-cyclodextrin•12-crown-4•K+. (b) Tubular and space-filling representation of a 4∶5 adduct of β-cyclodextrin•ferrocene. (c) Tubular and ball-and-stick representation of a 1∶1 adduct of cucurbit[6]uril•[Yb(OH2)8]3+[ReO4]−. (d) Tubular and space-filling representation of a 1∶1 adduct of resorcinarene•ferrocene. (e) Tubular and space-filling representation of a 1∶1 adduct of [Cu(NC5H5)2(H2O)4•calix[4]arene sulfonate]. (f) Tubular and space-filling representation of a 1∶1 adduct of (n-octyl)MgBr(THF)2•pillar[5]arene. Download figure Download PowerPoint Over the past two decades, there has been a resurgence of interest in second-sphere coordination. Advances in supramolecular, organic, inorganic, biomolecular, and materials chemistry have created91,92 new demands for molecular receptors capable of binding a wide range of substrates, constructing MIMs, separating different metals, and catalyzing organic reactions. In such contexts, second-sphere coordination has emerged as a valuable strategy for solving previously intractable problems. This minireview highlights many new adducts based on molecular receptors as second-sphere coordination ligands. Molecular Recognition Second-sphere coordination can be applied to the design and construction of new supramolecular complexes by incorporating noncovalent binding sites into ligands, followed by their coordination to transition metals. The metal ion can participate in molecular recognition by a receptor molecule either indirectly via second-sphere coordination or directly via simultaneous first- and second-sphere coordination, targeting specific substrates by tailored receptor-substrate and metal-ligand interactions. Loeb et al.50,80–83,93,94 have demonstrated the design and synthesis of several molecular receptors that interact with substrates through second-sphere interactions. One example involves80,94 a family of thiacyclophane receptors that recognize DNA nucleobases. The molecular recognition displayed by these receptors involves several different types of interaction, including (1) first-sphere coordination from a N atom of the nucleobase substrate to a Pd(II) metal center, (2) second-sphere hydrogen bonding between a NH2 group of the substrate and polyether O atoms of the receptors, and (3) [π⋯π] stacking between the π-electron-rich aromatic spacing units of the receptors and the π-electron-poor aromatic rings of the substrates. Single-crystal X-ray analyses of two adducts involving a receptor binding either with adenine or guanine—the latter as a BF3 adduct—reveals that all three interactions mentioned above exist in these adducts with near-ideal interaction parameters (Figures 4a and 4c). The molecular recognition of cytosine was realized by a receptor having coordination sites for Pd(II) and polyether functional groups. Single-crystal X-ray analysis (Figure 4b) reveals that adduct formation between this receptor and cytosine is sustained by direct first-sphere coordination from a N atom of cytosine to the Pd(II) metal center, aided and abetted by hydrogen-bonding interactions between the NH2 group of cytosine and polyether O atoms of the macrocyclic receptor. These adducts between the DNA nucleobases and their receptors have been proved to exist both in the solid state and in solution. The same principles have been applied81–83 in designing synthetic receptors for molecular recognition of other substrates, including barbiturates, amino acids, amines, and hydrazinium ions. Figure 4 | Stick representations of the X-ray superstructures of thiacyclophane receptors complexed with (a) adenine, (b) cytosine, and (c) guanine (BF3) through a combination of several interactions, including (1) first-sphere coordination from N atoms to Pd(II) metal center, (2) second-sphere hydrogen bonds between the NH2 group of the substrate and polyether O atoms of the receptors, and (3) [π–π] stacking between the electron-rich aromatic spacing units of the receptors and the electron-poor aromatic rings of the substrates. Download figure Download PowerPoint Huc et al.95–98 have presented (Figure 5a) the chemical community with several examples of molecular recognition through second-sphere coordination using molecular capsules formed by metal-coordination-directed folding of a helical oligomer. A key chain segment, pyridazine–pyridine–pyridazine (pyz–pyr–pyz), was introduced into the oligomer.95 This segment exists in an anti–anti conformation that is favored by repulsion between the endocyclic N atoms, directing the oligomer into an extended conformation instead of a capsular shape. As a result, this oligomer displays no propensity for molecular recognition in the absence of metal ions. Coordination of transition-metal ions Cu+, Cu2+, and Ag+, as well as the alkali metal ions Na+ and K+, induces folding of the oligomer into a capsular shape by favoring a syn–syn conformation for the pyz–pyr–pyz segment.96 In all cases, the metal ions were positioned on one side of the cavity wall, leaving its coordination sphere occupied partially by solvent molecules and available to bind a guest. The first-sphere coordination between guests and the metal centers can enhance the binding affinity of the helix, while the second-sphere interaction between the guest and the inner wall of the capsule leads to shape and size selectivity. Moreover, interaction of the oligomer with a hydrated Ca2+ or Ba2+ ion occurs entirely in the second sphere, via hydrogen bonding between ion-coordinated water and oligomeric N atoms, but still results95 in a helical geometry for the oligomeric complex. Figure 5 | (a) Schematic illustration of molecular recognition through second-sphere coordination using helical molecular capsules formed by metal-coordination-directed folding. (b) Stick-and-sphere representation of X-ray crystal superstructure of an adduct between a β-d-mannopyranose and a helical foldamer ligand associated with Cu2+. (c) A zoomed-in view of the substrate associated with the capsule through second-sphere coordination, which involves hydrogen-bond (dashed lines) formation between the foldamer ligand and the hydroxyl groups of the β-d-mannopyranose. The Cu2+ center adopts a square pyramidal geometry, coordinating with three nitrogen atoms of the pyz–pyr–pyz segment and two O atoms of two H2O molecules. Adapted with permission from ref 97. Copyright 2018 Royal Society of Chemistry. Download figure Download PowerPoint As an extension of these earlier reports, the same authors demonstrated molecular recognition of carbohydrates97 in organic solvents using foldamer-type molecular capsules, by taking advantage of first- and second-sphere coordination. Depending on the nature of the metal center, the molecular capsules show different binding affinities and diastereoselectivities toward a range of carbohydrates, including d/l-threitol, xylitol, d/l-mannopyranose, d/l-glucopyranose, d/l-galactopyranose, and d/l-fructopyranose. For example, the Cu2+ complex of the foldamer binds d-threitol 14 times more strongly than its K+ counterpart. X-Ray crystal superstructure analysis of the adducts between a Cu2+-coordinated capsule and d/l-mannopyranose reveals (Figures 5b and 5c) that the substrate is associated with the capsule through second-sphere coordination, which involves hydrogen bond formation between the foldamer ligand and the hydroxy groups of the mannopyranose. The Cu2+ center adopts a square pyramidal geometry, coordinating with three N atoms of the pyz–pyr–pyz segment and the O atoms of two H2O molecules. The d-mannopyranose assumes the β-pyranose configuration in the solid state. In contrast, this molecule exists predominantly (97%) in the α-pyranose form in CHCl3/Me2SO (4∶1, v/v) solution. The authors also reported the single-crystal superstructure of an adduct between d/l-threitol and a smaller analog of the molecular capsule coordinated with a Cu2+ ion. Here, the Cu2+ center interacts with the substrate through second-sphere coordination involving solvents—a H2O and a MeOH molecule—that are coordinated directly to the metal ion. As an extension of such molecular recognition, the authors also developed a new strategy98 to prepare the foldamer capsule shells around a [2Fe–2S] cluster. The foldamer shell influences the structural and spectroscopic properties of the metal cluster, including desymmetrization and confinement of part of its first coordination sphere within the foldamer cavity. Leigh et al.99–102 have reported molecular recognition of chloride anions through second-sphere coordination using molecular knots. These molecular knots were synthesized (Figure 6a) by transmetallation of a Zn(II) infused pentafoil (51) knot with tetrafluoroborate salts of Co(II), Cu(II), and Ni(II).99 The resulting metallated knots exhibit second-sphere coordination of a single chloride anion within the central cavity of the knot by utilizing both multiple [CH⋯Cl−] hydrogen bonds and electrostatic interactions. X-ray crystal superstructures of these transmetallated pentafoil knots reveal (Figures 6b–6d) that five of the 15 bipyridine groups form an inner cavity lined with 10 electron-poor H atoms that form an array of hydrogen bonds with the chloride anion located inside the cavity. This type of second-sphere coordination is the first example exhibited in molecular-knot geometry. The diameter of the central cavity varies depending on the metal cation: 3.3 Å for Fe(II), Zn(II), and Ni(II), and 3.5 Å for Co(II). The metal-to-metal distance in the distorted cobalt knot is, however, smaller than those in the case of iron, nickel, and zinc pentafoil knots. In contrast, the metal–metal–metal angle is higher in cobalt compared with those in the other three metallic pentafoil knots. These conformational changes in the knotted ligands lead to different distances between the chloride anion and the metal ions in these knot complexes. As a result, the binding affinities for chloride anion through second-sphere coordination and electrostatic interaction vary with different M(II) ions over nearly three orders of magnitude, from Ka = 8 × 104 M−1 (Cu2+ knot) to Ka = 3.3 × 107 M−1 (Fe2+ knot) in MeCN. Figure 6 | (a) Graphical illustration and structural formulas of various pentafoil knotted complexes synthesized by transmetalation of Zn(II) infused pentafoil (51) knot with tetrafluoroborate salts of Co(II), Cu(II), and Ni(II). The resulting metallated knots exhibit second-sphere coordination of a single chloride ion within the central cavity of the knot, providing [CH⋯Cl−] hydrogen bonding and electrostatic interactions. (b) Ball-and-stick representations of the X-ray crystal superstructures of chloride complex using the pentafoil knots sustained by the coordination with (b) Co2+, (c) Ni2+, and (d) Zn2+. The binding affinities for Cl− anions are determined by the second-sphere coordination and electrostatic interactions with different M(II) ions. Download figure Download PowerPoint Construction of Mechanically Interlocked Molecules The synthesis of MIMs103–108 such as catenanes,109–122 rotaxanes,123–132 and suitanes133–139 has become a field with intense research activity on account of the potential of such molecules to act as molecular machines.140–153 Several effective synthetic strategies for synthesizing MIMs have been achieved, based on the template effect.154–156 One of these strategies involves the coordination of organic ligands to transition metals to promote the formation of mechanically interlocked structures.157–168 Second-sphere coordination, as an extension of this strategy, can also be employed to advance the effective synthesis of MIMs. Wisner et al.169–173 demonstrated the use of second-sphere coordination to prepare several MIMs such as rotaxanes and catenanes. For example, this group described the synthesis of [2]pseudorotaxanes in a single step employing second-sphere coordination.169 An isophthalamide-based tetralactam macrocycle, serving as the second-sphere ligand, establishes four sets of hydrogen bonds with the two chloride first-sphere ligands in a trans-palladium dichloride complex, leading to formation of a pseudorotaxane. It was found that the stability of the pseudorotaxane diminishes markedly as the size of the halide ligand increases from Cl < Br < I. In a subsequent investigation, these authors reported170 the synthesis of a [2]rotaxane based on the same system. To produce the mechanically interlocked architecture shown in Figure 7a, 4-(3,5-di-tert-butylbenzyloxy)-pyridine was employed as the ligand, on the basis that its terminal groups are sterically demanding and function as stoppers to prevent the dethreading of the tetralactam macrocycle. The rotaxane was obtained in 89% isolated yield simply by mixing the tetralactam macrocycle, the ligands, and trans-bis(benzonitrile)palladium(II) dichloride in CHCl3 at room temperature. The resulting rotaxane is soluble in CHCl3 and remains stable over time as well as under column chromatographic conditions. The structure of the rotaxane was confirmed by X-ray crystallographic analysis, which revealed (Figure 7b) that the amide groups of the isophthalamide subunits in the macrocycle establish four sets of hydrogen bonds with the chloride ligands of the palladium(II) metal complex, providing the driving force for rotaxane formation. Following these investigations, these same authors explored172 the stabilities of a series of pseudorotaxanes by varying the metals and the para-substituted pyridine ligand. They found that the hydrogen bond-accepting ability of the chloride ligands can be tuned by varying the electron-donating/-withdrawing nature of the para-substituted pyridine coligands. Moreover, the stability of the pseudorotaxane in CHCl3 solution decreases slightly when the Pd(II) ion is replaced by Pt(II). Replacing chloride by thiocyanate in the first coordination sphere led to the formation (Figure 7c) of a series of (pseudo)rotaxanes by second-sphere coordination (hydrogen bonding) between the thiocyanate ligands and the amide units of the macrocycle.171 The directional nature of the thiocyanate ligands affords a doubly degenerate binding geometry, leading to a rotaxane (Figure 7c) in which the ring shuttles back and forth 3300 times a second in solution at 15 °C. Figure 7 | (a) Structural formula and (b) tubular and space-filling representations of X-ray solid-state superstructure of a rotaxane constructed by second-sphere coordination. The amide groups of the isophthalamide subunits in the macrocycle establish four sets of hydrogen bonds with the chloride ligands of the palladium(II) metal complex, providing the driving force for the [2]rotaxane formation. (c) X-ray solid-state structure of a [2]rotaxane formed by second-sphere coordination between the thiocyanate ligands and the amide units of the macrocycle. The directional nature of the thiocyanate ligands affords a degenerate binding geometry, leading to a degenerate [2]rotaxane shuttle where the ring moves back and forth along the dumb-bell shaped axle 3300 times a second in CDCl3 solution at 15 °C. Adapted with permission from ref 170. Copyright 2006 Royal Society of Chemistry. Download figure Download PowerPoint In another example, Wisner et al.173 demonstrated the construction of catenanes (Figure 8a) using simultaneous first- and second-sphere coordination. An acyclic bidentate ligand that resembles a three-quarter unit of a previously employed tetralactam macrocycle was designed and synthesized. This ligand coordinates directly to a PdCl2 unit in a trans arrangement, and the resulting product comprises (Figure 8a) a pair of catenated macrocycles, each incorporating the metal subunit trans-PdCl2L2 in its scaffold. The catenation is driven by mutual recognition of the PdCl2L2 subunit in one ring by its orthogonally positioned partner. The catenane was synthesized in 87% isolated yield simply by heating a solution of an equimolar amount of Pd(PhCN)2Cl2 and the trans-bidentate ligand. X-ray crystallographic analysis of the catenane reveals (Figure 8b) that a template effect is manifested in the interaction between both the PdCl2 subunits and their opposing macrocyclic cavities. Eight pairs of [NH⋯Cl] hydrogen bonds were found between the first-sphere Cl− ligands and the amide units of the second-sphere ligand. A reversible transformation between catenane and macrocycle was also demonstrated by varying the solvent polarity. Transformation of the catenane into two individual macrocycles was observed when the sample was dissolved in the mixed solvent system (CD3)2SO/CDCl3 as a result of competitive binding of the polar solvent with the NH groups of the macrocycle. The catenane re-forms when the macrocycle is redissolved in CDCl3. Figure 8 | (a) Structural formula and (b) tubular and space-filling representations of X-ray solid-state structure of a catenane constructed by second-sphere coordination. The catenation formation is driven by the mutual recognition of the PdCl2L2 subunit in one ring by its orthogonally disposed partner. Eight pairs of [NH⋯Cl] hydrogen bonds were found between the first-sphere Cl ligands and the amide units of the second-sphere ligand. Adapted with permission from ref 173. Copyright 2007 John Wiley and Sons. Download figure Download PowerPoint Metal Separation Second-sphere coordination of transition-metal complexes could lead to changes in solubility of the resulting adducts, which can be utilized for the separation of metal ions. We demonstrated (Figure 9c) the separation of copper from cobalt using35 this principle in 1983. We found that 18C6 formed (Figure 9a) a 1∶1 polymeric adduct with [Cu(NH3)4(H2O)][PF6]2, leading to precipitation from aqueous solution. X-ray structural investigation revealed