Intramolecular Hydrogen Bonding in Cu II Complexes with 2,6-Pyridinedicarboxamide Ligands:...

Job/Unit: I30579 /KAP1 Date: 03-07-13 15:23:49 Pages: 13

FULL PAPER

Hydrogen-Bonding Networks

D. Wang, S. V. Lindeman,A. T. Fiedler* .................................. 1–13

Intramolecular Hydrogen Bonding in CuII

Complexes with 2,6-Pyridinedicarbox-amide Ligands: Synthesis, StructuralCharacterization, and Physical Properties

Keywords: Hydrogen bonds / Copper / Nligands / Bioinorganic chemistry / Electro-

With the goal of developing ligands that or quinoline rings). These pincer ligandschemistrysupport intramolecular hydrogen-bonding coordinate to CuII centers in a tridentatenetworks, we prepared a series of 2,6-pyrid- fashion, and the protonated heterocyclesinedicarboxamide ligands featuring outer- form hydrogen bonds with first-spheresphere heterocycles (pyridine, pyrimidine, chlorido or aqua ligands.

0000, 0–0 © 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim1


www.eurjic.org FULL PAPER

DOI:10.1002/ejic.201300579

Intramolecular Hydrogen Bonding in CuII Complexes with2,6-Pyridinedicarboxamide Ligands: Synthesis, StructuralCharacterization, and Physical Properties

Denan Wang,[a] Sergey V. Lindeman,[a] and Adam T. Fiedler*[a]

Keywords: Hydrogen bonds / Copper / N ligands / Bioinorganic chemistry / Electrochemistry

The significance of second-sphere interactions in syntheticcatalysts and metalloenzyme active sites has encouraged thedevelopment of ligand frameworks capable of supporting in-tramolecular hydrogen-bonding networks. With this ap-proach in mind, we prepared a series of 2,6-pyridinedicarb-oxamide ligands (H2LX) that feature outer-sphere heterocy-clic groups (pyridine, pyrimidine, or quinoline rings). Reac-tion with CuCl2 yields complexes with the general formula[CuCl2(LX{H}2)] (1X), in which the pincer ligands coordinateto the CuII center in a tridentate fashion through the centralpyridine and two amidato nitrogen atoms. The LX{H}2 li-gands have no net charge, however, as the pendant hetero-cycles are both protonated. X-ray crystallographic studies re-

Introduction

Multidentate ligands containing 2-pyridinecarboxamideunits have proven valuable in coordination chemistry,homogeneous catalysis, and synthetic modeling of metallo-enzyme active sites.[1] Deprotonation of these ligands gener-ates an N-amidato group with impressive σ-donating ability,which allows for the stabilization of metal centers in highoxidation states.[2] Mascharak,[3] Tolman,[4] Holm,[5] andothers[6] have successfully used the 2,6-pyridinedicarboxam-ide scaffold to mimic coordination environments found inmetallobiomolecules such as bleomycin, nitrile hydratase,acetyl coenzyme A synthase, and O2-activating copper en-zymes. Similar ligands have found application in catalystsfor asymmetric alkylations,[7] epoxide ring openings,[8] andoxidation reactions.[9]

It may be possible to enhance the reactivity and bio-logical relevance of pyridinecarboxamide-based complexesthrough the incorporation of outer-sphere (i.e., second-sphere) groups capable of forming intramolecular hydrogenbonds. Extensive studies of metalloenzymes have high-lighted the importance of such interactions in tuning redoxpotentials,[10] stabilizing intermediates,[11] and serving as

[a] Department of Chemistry, Marquette University,P. O. Box 1881, 535 N. 14th St., Milwaukee, WI 53233, USAE-mail: [email protected]: http://www.marquette.edu/chem/Fiedler.shtmlSupporting information for this article is available on theWWW under http://dx.doi.org/10.1002/ejic.201300579.

Eur. J. Inorg. Chem. 0000, 0–0 © 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim2

vealed that the two outer-sphere N–H groups form intramo-lecular hydrogen bonds with the chlorido ligands. Modifica-tion of the syntheses provided CuII complexes (i.e., 2X, 3X)that contain an aqua ligand hydrogen bonded to the outer-sphere group(s). Whereas complexes with the LX{H}2 ligandsare mononuclear, deprotonation gives rise to dimeric com-plexes (i.e., 4X, 5X) in which one of the pendant heterocyclescoordinates to a second CuII center. The complexes were alsoexamined with electronic absorption spectroscopy, cyclic vol-tammetry, and density functional theory. Our results indicatethat the LX{H}2 ligands, despite their neutrality, maintain thestrong σ-donating ability of anionic pyridine-amidato li-gands.

proton donors/acceptors.[12,13] In 1997, Redmore et al. re-ported the reaction of RuCl2(PPh3)3 with the ligand N,N�-bis(6-methyl-2-pyridyl)-2,6-pyridinedicarboxamide (H2LA

in Scheme 1).[14] As expected, this pincer ligand coordinatesto RuII in a tridentate fashion through the central pyridineand two amidato donors; however, the appended pyridinerings are both protonated, which makes the overall ligand aneutral dizwitterion [labeled LA{H}2 to indicate the protonshift]. Most significantly, the two pyridinium groups formintramolecular H-bonds with a chlorido ligand.[14] In thismanuscript, we expand upon Redmore’s initial observation

Scheme 1. Pincer ligands discussed in this study.



through the synthesis of various copper(II) complexes fea-turing intramolecular H-bonds between protonatedheterocycles and first-sphere chlorido and/or aqua ligands.

Substituted N,N�-bis(2-pyridyl)-2,6-pyridinedicarboxam-ide ligands are attractive frameworks for the incorporationsecond-sphere H-bonding interactions into transition-metalcomplexes. First, the ligand syntheses are straightforwardand the identity of the pendent group is easily modified; asshown in Scheme 1, we utilized 2,6-pyridinedicarboxamides(H2LB–D) with three different pendant moieties that modifythe H-bonding capabilities of the ligand. Pyridine, pyrimid-ine, and quinoline rings are well suited to serve as acid–base catalysts as a result of their moderate basicities. Thepyrimidinium group has the added advantage that it canfunction simultaneously as a H-bond donor and acceptor.Finally, the pendant groups in the 2,6-pyridinedicarbox-amide scaffold are generally unable to bind to the metalcenter, as the resulting ring chelate would be highlystrained.

Only one report to date has explored the coordinationchemistry of LX{H}2-type ligands with first-row transitionmetals. Gudasi et al. reacted N,N�-bis(2-pyridyl)-2,6-pyrid-inedicarboxamide (H2L; Scheme 1) with various divalentfirst-row cations to generate the putative MII/L0{H}2 com-plexes, although only the copper(II)–dichlorido complexwas structurally characterized.[15] The X-ray structure re-vealed that the pendant pyridinium groups form intramo-lecular H-bonds with the oxygen atoms of the backboneamides instead of the chlorido ligands, as intended. By con-trast, in this manuscript, we report the synthesis and X-ray structural characterization of several mononuclear CuII

complexes that display the desired H-bonding interactionsbetween first-sphere ligands and protonated outer-spheregroups (see Scheme 2 for an example). The use of methyl-substituted heterocycles was critical to our success, as it dis-courages formation of the carbonyl/pyridinium H-bondfound in the Gudasi structure. In addition, we found thatdeprotonation of one (or both) of the pendant groups re-sults in the formation of dicopper(II) complexes. Our re-sults therefore point to the possibilities and limitations ofusing 2,6-pyridinedicarboxamide-based ligands to construct

Scheme 2. Synthesis of the [CuCl2(LC{H}2)] (1C) complex.


H-bonding networks in synthetic complexes. Finally, thedonor properties of the LX{H}2-type ligands were exam-ined through the use of spectroscopic, electrochemical, andcomputational methods. These results were compared tothose obtained for a complex that instead contains ligandLE (Scheme 1) � a member of the well-known class of 2,6-bis(imino)pyridine ligands.

Results and Discussion

Synthesis and Structures of Mononuclear CuII Complexes

Whereas the H2LA and LE ligands are known com-pounds,[14,16] the pincer H2LB–D ligands in Scheme 1 werefirst prepared in our laboratory. Each was generated by thereaction of pyridine-2,6-dicarbonyl dichloride with the ap-propriate amino-substituted heterocycle under basic condi-tions. Mononuclear CuII complexes with the general for-mula [CuCl2(LX{H}2)] (1X) were then prepared by mixingequimolar amounts of CuCl2 and the desired pincer ligand(Scheme 2). X-ray-quality crystals of the 1X series were ob-tained by slow evaporation of CH2Cl2 or layering a concen-trated solution with pentane. Details concerning the X-raydiffraction (XRD) experiments included in this study areprovided in Tables 5 and 6 and in the Experimental Section.

Selected bond lengths and angles for complexes in the1X series are shown in Table 1. As displayed in Figure 1,complexes with pendant pyridyl (i.e., 1B) or pyrimidyl (i.e.,1C) groups each contain a five-coordinate (5C) CuII centerligated to two chlorido ligands and one neutral LX{H}2

bound in a meridional fashion. These complexes exhibitidealized C2 symmetry along the Cu–N1 axis, and this re-sults in nearly equivalent sets of chlorido and amidato do-nors. The coordination geometries of 1B and 1C are bestdescribed as distorted trigonal bipyramidal in which thetwo amidato nitrogen atoms (N2 and N3) occupy the axialpositions, although the N2–Cu–N3 axis is bent towards thepyridyl ring with an angle of approximately 159°. In theequatorial plane, the N1–Cu–Cl angles range between 126and 138° and the Cl1–Cu–Cl2 bonds form approximateright angles. Typical for pyridinedicarboxamide ligands,[3,17]

the shortest metal–ligand bond involves the central pyridinegroup, with a distance of (1.94�0.01) Å in the 1X series.The average copper(II)–amidato bond length of 2.05(1) Åis slightly longer than the usual distance of about 2.0 Åfound in 5C CuII complexes with dianonic 2,6-pyridine-dicarboxamide ligands.[3,17] In contrast, these bonds are sig-nificantly shorter than the corresponding Cu–N bonds indichlorido CuII complexes with neutral 2,6-bis(imino)pyr-idine ligands (i.e., Schiff bases), which display distances of(2.10�0.02) Å.[18,16] This result suggests that the amidatodonors of 1B and 1C still possess anionic character andstrong σ-donating ability, despite the overall neutrality ofthe LX{H}2 ligands.



Table 1. Selected bond lengths and bond angles for complexes 1X.

1B 1C[a] 1D

Bond length [Å]

Cu1–Cl1 2.4569(7) 2.4393(5) 2.196(1)Cu1–Cl2 2.3909(6) 2.3361(5) 3.029(1)Cu1–N1 1.946(2) 1.939(2) 1.933(3)Cu1–N2 2.053(2) 2.050(1) 2.028(3)Cu1–N3 2.050(2) 2.046(2) 2.018(3)

H-bonding parameter

N4···Cl2 [Å] 3.061(2) 3.109(2) 3.066(3)[c]

N5···Cl1 [Å] 3.119(2) 3.138(2) 3.088(3)[c]

N4–H···Cl2 [°] 152.3 161.2 151.3N5–H···Cl1 [°] 163.9 164.8 152.3

Dihedral “tilt” angle of pendant ring[b] [°]

C6–N2–C8–C10/N6 35.0(4) 22.0(3) 64.3(5)C7–N3–C9–C11/N7 23.0(4) 24.9(3) 70.5(5)

Bond angle [°]

Cl2–Cu1–Cl1 95.97(2) 97.95(2) 100.11(4)N1–Cu1–Cl1 126.22(6) 127.44(4) 175.93(9)N1–Cu1–Cl2 137.81(6) 134.58(5) 83.94(9)N1–Cu1–N2 79.07(8) 79.69(6) 79.86(1)N1–Cu1–N3 79.38(8) 79.66(6) 80.56(1)N2–Cu1–Cl1 94.47(6) 94.66(4) 99.69(8)N2–Cu1–Cl2 99.95(6) 97.03(5) 89.64(8)N3–Cu1–Cl1 99.45(6) 97.81(4) 99.79(9)N3–Cu1–Cl2 94.97(6) 97.46(5) 89.09(9)N3–Cu1–N2 158.39(8) 159.35(6) 160.40(1)

[a] Complex 1C features four symmetrically independent units thatare chemically equivalent. Metric parameters are only provided forone complex. [b] Measures the absolute angle between the pyridine-dicarboxamide plane and the plane of the pendant heterocycle.[c] In the structure of 1D, one chloride (Cl2) participates in H-bond-ing interactions with both N–H groups, see Figure 1.

As intended, the LX{H}2 framework in 1B and 1C givesrise to intramolecular H-bonds between first- and second-sphere groups (Figure 1). Each chlorido ligand participatesin a single H-bond interaction with a pendant pyridinium(or pyrimidinium) ring. The average N···Cl distance of3.11 Å and the N–H···Cl angle of approximately 160° pro-vide solid confirmation of this H-bonding assignment. Inaddition, the protonated heterocycles tilt out of the pyrid-inedicarboxamide plane unit by 22–36° to accommodatethese interactions. Thus, our complexes deviate from thestructure reported for the related [CuCl2(L{H}2)] complex,in which the pyridinium groups are H-bonded to the carb-onyl groups instead.[15] The presence of methyl substituentson the pendant rings in our systems likely discourages thisconformation, as the methyl groups would sterically clashwith the chlorido ligands. The Cu–Cl bonds in 1B and 1C

range from 2.34 to 2.46 Å with an average length of 2.40 Å.Overall, these values lie on the high end of the 2.2–2.4 Årange normally observed in copper(II)–chlorido com-plexes,[15,19,20] which suggests that the H-bonding interac-tions cause a modest lengthening of the Cu–Cl bond.

The structure of 1D, shown in Figure 1, is unique in the1X series in that the appended quinoline rings adopt a synconformation and form H-bonds with the same chlorideanion, Cl2. Relative to the pyridyl and pyrimidyl systems,


Figure 1. Thermal ellipsoid plots (50% probability) derived fromthe X-ray crystal structures of complexes 1B, 1C, and 1D. Nonco-ordinating solvent molecules and most hydrogen atoms are omittedfor clarity.

the N–H donors of the protonated quinoline groups arefurther removed from the amidato donors, which requiresCl2 to shift away from the Cu center to accommodate theH-bonding interactions. Indeed, the Cu···Cl2 distance of3.029(1) Å indicates that the CuII geometry is effectivelysquare planar. The reduction in coordination numbercauses the first-sphere Cu–N/Cl distances to decrease rela-tive to their values in 1B and 1C, and the N1–Cu–Cl1 angleapproaches 180° [the actual value is 175.9(1)°]. Thus,changes in the positions of the H-bond donating groupscan dramatically alter the coordination environment of themetal ions.

Copper(II)–triflate (OTf) complexes with the general for-mula [Cu(LX{H}2)(OTf)2(H2O)] (2X) were synthesized byusing the H2LB and H2LC ligands. These complexes wereprepared by two separate methods: (1) direct mixing of the2,6-pyridinedicarboxamide ligand with Cu(OTf)2, or(2) treatment of complex 1X with AgOTf (2 equiv.) inMeCN (Scheme 3). X-ray-quality crystals of the triflate-containing complexes were obtained by vapor diffusion ofdiethyl ether into MeCN solutions. The aqua ligand in thestructures of 2X (vide infra) presumably arises from trace



Scheme 3.

amounts of water in the solvent. The reaction of only1 equivalent of AgOTf with complex 1B gave rise to themixed chlorido/triflate [CuCl(LX{H}2)(OTf)(H2O)] (3B)complex. Given that these complexes contain easily dis-placeable triflate anions, they are well suited to serve asscaffolds that permit various ligands to coordinate to the[Cu(LX{H}2)]2+ unit.

Triflate-containing complexes 2B, 2C, and 3B exhibit con-siderable structural diversity. The six-coordinate CuII centerin 2C is bound to LC{H}2, an aqua ligand, and two triflatesin a trans orientation (Figure 2). The H2O ligand is tightlybound with a Cu–O1 distance of 1.930(2) Å, whereas theCu–N distances are comparable to those observed in the 1X

series (Table 2). The Cu–O(triflate) distances of 2.415(2)and 2.545(2) Å are indicative of weak dative bonds. Signifi-cantly, each protonated pyrimidine group participates intwo H-bonding interactions and acts as an acceptor to theaqua ligand and a donor to the carbonyl group of LC{H}2.This behavior is reminiscent of the ability of histidine imid-azoles to form H-bond networks in biological systems. Incontrast, the 4,6-dimethylpyridinium groups in 2B lack thisbifunctional nature, which results in a very different struc-ture (Figure 2). One triflate group moves to the outersphere, and this results in a 5C CuII center with square-pyramidal geometry. In the solid state, the H2O ligand of2B forms intermolecular H-bonds with both bound and un-bound triflate anions, and one pyridinium group is copla-nar with the LB{H}2 ligand owing to an intramolecularbond with a carbonyl moiety. The other pyridinium group,however, tilts out of the plane to form an intermolecularH-bond with the coordinated triflate of a nearby complex.The case of 2B is a reminder that the presence of multipleH-bond donors and acceptors in the first and second coor-dination spheres often leads to complex (and unpredictable)geometries.

The structure of 3B, shown in Figure 3, resembles that of1B described above, except that one chlorido ligand is re-placed by H2O. The 5C CuII geometry is distorted square


Figure 2. Thermal ellipsoid plots (50 % probability) derived fromthe X-ray crystal structures of complexes 2C (top) and 2B (bottom).Noncoordinating solvent molecules and most hydrogen atoms areomitted for clarity.

pyramidal (structural parameter, τ,[21] of 0.16), and theaqua ligand is in the axial position. Both the Cl and H2Oligands participate in hydrogen-bonding interactions withthe appended pyridinium groups, which display rather typi-cal N···Cl and N···O distances of 3.11 and 2.87 Å, respec-



Table 2. Selected bond lengths and bond angles for complexes 2B,2C, and 3B.

2B 2C 3B

Bond lengths [Å]

Cu1–N1 1.929(8) 1.915(2) 1.928(2)Cu1–N2 2.098(8) 2.058(2) 2.067(2)Cu1–N3 2.054(8) 2.069(2) 2.078(2)Cu1–O3 1.973(7) 1.930(2) 2.249(2)Cu1–O4 2.360(7) 2.415(2) –Cu1–O7 – 2.545(2) –Cu1–Cl1 – – 2.2900(6)

Dihedral “tilt” angle of pendant ring[a] [°]

C6–N2–C8–C10/N6 176(1) 176.2(2) 24.8(3)C7–N3–C9–C11/N7 131.6(9) 173.5(2) 14.6(3)

[a] Measures the absolute angle between the pyridinedicarboxam-ide plane and the plane of the pendant heterocycle.

tively. In the solid state, the aqua ligand forms H-bondswith two triflate counteranions, and this accounts for thelengthy Cu–O3 distance of 2.249(2) Å (Table 2).

Figure 3. Thermal ellipsoid plot (50% probability) derived from theX-ray crystal structure of 3B. The triflate counteranion, noncoordi-nating solvent molecules, and most hydrogen atoms are omitted forclarity. Each aqua ligand forms hydrogen bonds with two outer-sphere triflate anions.

Synthesis, Structures, and Magnetism of Dinuclear CuII

Complexes

Whereas the previous section demonstrated that theLX{H}2 framework yields mononuclear CuII complexeswith intramolecular H-bonding interactions, we found thatcertain conditions gave rise to dimeric structures instead.For instance, vapor diffusion of diethyl ether into a solutionof 2B in MeOH provides dark violet crystals that are clearlydistinct from the light-green crystals obtained from MeCN/Et2O. XRD analysis revealed that the violet crystals consistof the dicopper complex [Cu2(LB{H})2(OTf)2] (4B). Asshown in Figure 4, the dimeric structure of 4B arises fromthe fact that the monoanionic LB{H} ligands now possessan unprotonated pyridyl ring capable of coordinating to asecond [Cu(LB{H})]+ unit (i.e., a “ligand sharing” struc-ture[22]). Each 5C CuII center in 4B exhibits a square-pyram-idal geometry (τ = 0.17) with a weakly bound triflate anionin the axial position. The two protonated pyridyl rings form


intramolecular H-bonds with C=O groups on the oppositeLB{H}– ligands, which enhances the stability of the dimericstructure. The Cu···Cu distance was found to be 3.858(3) Å.A similar dicopper structure was previously reported byWoolins and co-workers;[6] however, this complex lacks theH-bonding interactions of 4B, and the axial positions areoccupied by aqua ligands instead of triflates.

Figure 4. Thermal ellipsoid plots (50% probability) derived fromthe X-ray crystal structures of complexes 4B and 5B. Noncoordina-ting solvent molecules and all hydrogen atoms are omitted for clar-ity. Note: Complex 5B has C2 symmetry; thus, the [Cu(LB)] unitshave identical structures.

Dicopper species were also obtained during attempts todeprotonate the LX{H}2 ligands of 1B and 1C. Reaction ofthese complexes with 2 equivalents of base (e.g., NaOMe orNEt3) in CH2Cl2 produced reddish-purple precipitates (5B

and 5C, Scheme 3) that were crystallized by vapor diffusionof pentane into dilute CH2Cl2 solutions. X-ray crystallogra-phy found that the purple material corresponds to the di-meric [Cu2(LX)2] complexes, 5B and 5C, that contain thefully deprotonated, dianionic LX ligands. As in the case of4B, one pyridyl (or pyrimidyl) donor from each LX ligandforms a link to the second Cu center, although the lack oftriflate counteranions in 5B and 5C results in four-coordi-nate centers (Figure 4, see also Figure S1, Supporting Infor-mation). The complexes are C2 symmetric with Cu···Cu dis-tances of approximately 3.2 Å. The equivalent CuII ions ex-hibit distorted square-planar geometries with short Cu–Ndistances clustered between 1.93 and 2.00 Å (Table 3).



Table 3. Selected bond lengths and bond angles for dicopper com-plexes 4B, 5B, and 5C.

4B 5B[a] 5C[a]

Cu1 Cu2

Bond length [Å]

Cu–N1 1.923(3) 1.930(3) 1.921(2) 1.926(5)Cu–N2 2.016(3) 2.008(3) 2.000(2) 2.004(5)Cu–N3 2.035(3) 2.026(3) 1.956(2) 1.974(5)Cu–N4[a] 2.005(3) 2.008(3) 1.968(2) 1.977(5)Cu–O5 2.550(3) 2.453(3) – –Cu···Cu 3.858(3) 3.109(2) 3.217(5)

Bond angle [°]

N1–Cu–N2 79.6(1) 79.6(1) 80.10(6) 80.5(2)N1–Cu–N3 80.4(1) 80.6(1) 80.74(6) 80.6(2)N1–Cu–N4 169.5(1) 168.8(1) 169.93(6) 167.4(2)N2–Cu–N3 159.5(1) 159.8(1) 160.19(6) 160.0(2)N2–Cu–N4 94.7(1) 94.8(1) 97.08(6) 95.7(2)N3–Cu–N4 104.3(1) 103.9(1) 102.65(6) 104.2(2)O5–Cu–N1 87.3(1) 88.7(1) – –O5–Cu–N2 83.4(1) 84.7(1) – –O5–Cu–N3 100.1(1) 99.0(1) – –O5–Cu–N4 100.8(1) 100.6(1) – –

[a] The copper centers in 5B and 5C are equivalent by symmetry.

UV/Vis absorption experiments in CH2Cl2 confirmedthat the dimerization process is reversible, such that sequen-tial additions of acid and base result in interconversion of1B,C and 5B,C (see Figure S2, Supporting Information).Complexes 5B and 5C were also prepared directly by thereaction of CuCl2, H2LX, and NaOMe (2 equiv.) in MeOH.

A solid-state sample of 5B provided an effective magneticmoment (μeff) of 2.4(1) μB at room temperature, which issomewhat less than the spin-only value of 2.56 μB that isexpected for a dinuclear species with two uncoupled S =1/2 spins. This result indicates the presence of modest anti-ferromagnetic coupling between the CuII centers. Indeed,broken-symmetry calculations[23] by using DFT computedan exchange coupling parameter (J) of –41 cm–1 for 5B. In-teractions of this magnitude would yield a μeff value of2.3 μB, which is reasonably close to the value measured ex-perimentally.

Interestingly, deprotonation of the quinoline-appendedH2LD ligand in the presence of CuCl2 does not yield a di-copper species. Instead, a mononuclear 5C [Cu(LD)] (6D)complex is formed in which dianionic LD serves as a pen-tadentate ligand that is coordinated through pyridyl, amid-ato, and quinoline donors. In the X-ray structure of 6D

(Figure S3 and Table S1, Supporting Information), the CuII

ion exhibits a trigonal-bipyramidal geometry with amidatogroups in the axial positions. Thus, the LD(2–) ligand iscapable of forming ring chelates involving the pairs of quin-oline and amidato donors, whereas this is not feasible inthe corresponding LB- and LC-based systems.

Electronic Absorption Spectroscopy

The electronic absorption spectra of the 1X series inCH2Cl2 are shown in Figure 5 (top). The green color of


complexes 1B and 1C arises from a broad, asymmetric ab-sorption manifold between 600 and 1100 nm that comprises(at least) two bands. These features exhibit molar absorp-tivity values (ε values) of approximately 250 m–1 cm–1, whichis characteristic of d–d transitions. The more intense fea-tures in the near-UV region are due to Cl�CuII charge-transfer excitations. The absorption spectra of 1B and 1C

are typical of 5C CuII complexes with distorted geometriesbetween the square-pyramidal and trigonal-bipyramidal li-mits.[24,25] The absorption maxima (λmax) at 700 nm and theshoulder near 900 nm are assigned to the dxy�dx2–y2 anddxz�dx2–y2 transitions, respectively (for which the z axis isperpendicular to the CuN3 plane and the x axis pointsalong the Cu–N1 bond).

Figure 5. Electronic absorption spectra of complexes 1B, 1C, 1D,and 7 in CH2Cl2 (top) and triflate-containing complexes 2B, 2D,and 3B in MeCN (bottom).

Given that the structure of 1D deviates from those of 1B

and 1C (vide supra), it is not surprising that its absorptionspectrum is also distinct; it has less overall intensity anda single band at shorter wavelength (660 nm). Indeed, thespectrum of 1D closely resembles the spectra of squareplanar CuII complexes.[24,26] Likewise, the spectra of 2B and2C in MeCN (Figure 5, bottom) are characteristic of CuII

ions in tetragonal environments with strong equatorial do-nors and weakly bound axial ligands (likely derived fromthe solvent in this case). The spectrum of 3B, however,closely resembles those reported for 5C complexes 1B and1C, which suggests that the former complex maintains a[CuCl(LB{H}2)(Y)]+ structure in solution, in which Y =H2O or MeCN.

To compare the electronic properties of the LX{H}2 li-gands with the well-established 2,6-bis(imino)pyridineframework, we generated the previously reported [(LE)-CuCl2] (7) complex, in which LE = 2,6-bis(N-2-methylphen-ylacetimino)pyridine (Scheme 1). The crystallographicstructure of 7, as elucidated by Fan et al.,[16] reveals aCuCl2N3 coordination geometry very similar to those ob-



served for complexes 1B and 1C, which makes 7 an excellentreference complex. The absorption spectrum of 7 reveals apeak with λmax = 850 nm (ε = 200 m–1 cm–1) that arises froma d–d transition (Figure 5). This feature is redshifted by ap-proximately 2100 cm–1 relative to the corresponding bandsin the 1B and 1C spectra, and this indicates that the 2,6-bis(imino)pyridine framework exerts a considerably weakerligand field than LB,C{H}2. This conclusion is supportedby electrochemical experiments, which are described in thefollowing section.

The absorption spectra of 4B and 5B in CH2Cl2 (Fig-ure S4, Supporting Information) are typical of CuII centersin square-planar geometries. However, in strongly donatingsolvents such as DMF and MeOH, our spectroscopic datasuggest that these complex exist, at least partially, as mono-mers. This is apparent when the dark violet crystals of thedicopper species are dissolved in MeOH to give blue solu-tions with absorption features similar to those observed for2B and 2C.

Electrochemistry

The redox properties of mononuclear Cu/LX{H}2 com-plexes were examined by using cyclic voltammetry (CV) inCH2Cl2 and DMF with 0.1 m [NBu4]PF6 as the supportingelectrolyte. The results are summarized in Table 4. Asshown in Figure 6, complex 1B displays an irreversible cath-odic wave at –0.43 V (vs. SCE) in CH2Cl2 that correspondsto reduction of the 5C CuII center to CuI. Previous studiesof similar [LN3Cu2+Cl2] complexes have demonstrated thatreduction often triggers dissociation of one (or both) of thechlorido ligands,[18] which likely accounts for the irrevers-ible nature of this event. The partially reversible couple ob-served at higher potential (E1/2 = +0.59 V; ΔE = 220 mV)is then assigned to the low-coordinate Cu2+/+ species de-rived from 1B reduction. The voltammograms collected for1B in CH2Cl2 and DMF share the same general form, al-

Table 4. Electrochemical data for mononuclear CuII/LX{H}2 complexes.[a]

Complex Solvent Ep,c [V][b] E1/2 [V] ΔE [mV][c] ip,c/ip,a[d]

1B CH2Cl2 –0.43 +0.59 220 0.34DMF –0.27 +0.39 170 0.35

1C CH2Cl2 –0.34 +0.61 180 0.27DMF n.o.[a] +0.40 160 0.79

1D CH2Cl2 –0.40, –0.68 Ep,a = +0.59[b] – –DMF n.o. +0.40 150 0.83

2C DMF n.o. +0.41 110 1.00

3B CH2Cl2 –0.09 +0.68 210 0.23

4B CH2Cl2 –0.28, –0.68 Ep,a = +0.42 – –

5B CH2Cl2 –0.93, –1.55 Ep,a = +0.07 – –

7 CH2Cl2 +0.02 Ep,a = +0.56[b]

[a] All potentials are reported versus SCE; n.o.: not observed. [b] Ep,c and Ep,a values are given for irreversible reduction and oxidationevents, respectively. [c] ΔE = Ep,a – Ep,c. [d] Ratio of cathodic current (ip,c) and anionic current (ip,a).


though the potentials are somewhat shifted. In contrast,complexes 1C and 1D exhibit markedly different electro-chemical behavior in the two solvents. Like 1B, both com-plexes display irreversible reduction waves at low potentials(Ep,c � –0.3 V) in CH2Cl2; however, these features disap-pear in DMF, whereas the higher-potential event near+0.4 V becomes quasireversible (Table 4 and Figure 6).

Figure 6. Cyclic voltammograms of 1B (black), 1D (red), 1E (blue),2D (grey), 3B (green), and 7 (violet) in CH2Cl2 (solid lines) andDMF (dotted lines) containing 0.1 m [NBu4]PF6 as the supportingelectrolyte ([Cu] ≈ 2.0 mm). Data were collected with a scan rateof 100 mV and each voltammogram was initiated by the cathodicsweep.

The electrochemical data are rationalized in the followingmanner: one-electron reduction of the 1X series in CH2Cl2



causes dissociation of both chlorido ligands, and this yieldsthe three-coordinate [Cu+(LX{H}2)]+ species that is oxid-ized at potentials between 0.6 and 0.8 V. This conclusionis consistent with the voltammogram measured for 3B�acomplex with only one chlorido ligand. Whereas 3B is con-siderably easier to reduce than complexes in the 1X series(as expected), the corresponding high-potential events allappear at roughly the same potential [(+0.63 �0.05) V] inCH2Cl2 (Table 4). This result indicates that reduction of 1X

and 3B gives rise to the same [Cu+(LX{H}2)]+-type species,regardless of the number of chlorido ligands in the precur-sor complex. Using similar reasoning, the two cathodicwaves observed for 1D (Figure 6) point to a prereductionequilibrium with the form [CuCl2(LD{H}2)] p

[CuCl(LD{H}2)]+ + Cl– in CH2Cl2, such that loss of theweakly bound, axial chlorido ligand shifts the Cu2+/+ po-tential of 1D from –0.68 to –0.40 V (Table 4). Yet, 1D exhib-its only one anodic feature at Ep,a = +0.59 V, and this indi-cates that reduction generates a single species, that is,[Cu+(LD{H}2)]+.

Although the similarity of the cyclic voltammogramssuggests that 1B maintains the same overall structure inCH2Cl2 and DMF, solvation of complexes 1C and 1D inDMF facilitates ionization to the [Cu2+(LX{H}2)]2+

state�a well-established process for copper–halide com-plexes in strongly donating solvents.[27] As shown in Fig-ure S5 (Supporting Information), the absorption spectra of1C and 1D in DMF are dramatically different than thosemeasured in CH2Cl2, and this provides further evidencethat solvation in DMF facilitates loss of the chlorido li-gands (by contrast, the spectra measured for 1B are nearlyidentical in the two solvents). Therefore, only one redoxcouple is observed for 1C and 1D in DMF. In support ofthis scenario, the cyclic voltammogram of[Cu2+(LC{H}2)](OTf)2 (2C), which lacks chlorido ligation,displays a single reversible couple at approximately thesame potential in DMF. Therefore, we conclude that the[Cu(LX{H}2)]2+/+ units possess E1/2 values of approximately0.6 V in CH2Cl2 and approximately 0.4 V in DMF.

The CV data measured for dicopper complexes 4B and5B in CH2Cl2 are shown in Figure S6 (Supporting Infor-mation). Both species exhibit two irreversible peaks sepa-rated by 0.4 V (for 4B) or 0.6 V (for 5B), which correspondsto stepwise reduction of the two CuII centers (Table 4).Given the different protonation states of the LB ligands in4B and 5B, it is not surprising that the former complex iseasier to reduce (by 0.65 V) than the latter. The irreversibil-ity of the cathodic waves suggests that reduction forces thedimeric structures to split apart, and the resulting CuI

monomers are then oxidized at higher potentials (Table 4).The redox properties of reference complex 7 were also

measured in CH2Cl2. The voltammogram displays anirreversible reduction wave at +0.02 V followed by anirreversible anodic event at +0.56 V. The electrochemical be-havior of 7 therefore follows the pattern described above for1B–D, in which reduction initiates chloride dissociation togive a low-coordinate CuI species. However, complex 7 iseasier to reduce than 1B or 1C by more than 0.35 V, even


though the three complexes share nearly identical CuCl2N3

coordination geometries. DFT calculations indicate that,for all the complexes in this study, the redox-active orbitalis the Cu dx2–y2 orbital that lies in the N3 plane. This orbitalis destabilized by approximately 1800 cm–1 upon substitu-tion of LE with LC{H}2 (i.e., conversion of 7 into 1C), andthis is consistent with the cathodic shift in redox potential.Collectively, the spectroscopic and electrochemical datahighlight the stronger σ-donating ability of the amidatogroups of LB,C{H}2 relative to the imine moieties of LE.Indeed, a survey of the literature reveals that the potentialsfor 1X lie at the low end of the range normally observed for5C copper complexes.[28]

Conclusions

We have described the synthesis and coordination chem-istry of tridentate pincer ligands with the 2,6-pyridinedi-carboxamide motif that feature pendant pyridine, pyrimid-ine, or quinoline rings. These ligands coordinate to CuII

centers as neutral, dizwitterions, LX{H}2, which thus pro-vides two protonated heterocycles capable of serving as H-bonding donors and/or acceptors. The [Cu(LX{H}2)] com-plexes were characterized by X-ray crystallography, and theresulting structures revealed intramolecular H-bonds be-tween the first-sphere and the outer-sphere N–H groups.The H-bonding interactions lengthen the Cu–Cl bondlengths and, in the case of 1D, facilitate the dissociation ofa chlorido ligand. However, the 2,6-pyridinedicarboxamideframework suffers from a major limitation: deprotonationof the pendant groups promotes formation of dimeric com-plexes in which appended pyridine (or pyrimidine) groupsbind to a second copper ion (Scheme 3). Thus, we are cur-rently developing modified versions of these ligands thatwill permit proton-transfer reactions without concomitantdimerization.

The electronic properties of the LX{H}2 ligand wereevaluated with absorption spectroscopy, cyclic voltammetry,and DFT calculations. Anionic pyridine–amidato ligandsare renowned as strong σ donors that are capable of stabi-lizing metal centers in high oxidation states.[1,2] Our resultsindicate that the neutral LX{H}2 ligands retain the elec-tronic features of the amidato ligands, despite the presenceof the protonated heterocycles that might diminish donorability owing to π-conjugation effects (as illustrated inScheme 4). The elevated d–d transition energies and the lowpotentials of the 1X complexes relative to those of the corre-sponding 2,6-bis(imino)pyridine-based complex (i.e., 7) sug-gest that the LX{H}2 framework exerts an unusually strongligand field for a neutral compound. The LX{H}2 ligandsare therefore reminiscent of N-heterocyclic carbenes, whichalso combine impressive σ-donating ability with neutraloverall charge. In addition, both types of ligands involve πconjugation with sp2-hybridized nitrogen atoms that carrya partial positive charge (Scheme 4). Thus, the electronicproperties of the LX{H}2 framework, in conjunction withtheir ability to participate in H-bonding interactions, sug-



gest that these ligands may have promising applications incatalyst development. We also anticipate that the presenceof outer-sphere hydrogen-bond donors will help stabilizeCu–O2 and Cu–O2R (R = H, tBu) intermediates that arerelevant to O2-activating copper enzymes.

Scheme 4. Resonance structures of LX{H}2-type ligands and N-heterocyclic carbenes.

Experimental SectionMaterials and Physical Methods: All reagents and solvents werepurchased from commercial sources and used as received unlessotherwise noted. Acetonitrile (MeCN), dichloromethane, and tetra-hydrofuran (THF) were purified and dried by using a Vacuum At-mospheres solvent purification system. Ligand LE and complex 7were prepared by using published procedures.[16] Elemental analy-ses were performed at Midwest Microlab, LLC in Indianapolis, In-diana. Infrared spectra were measured as solid powders by using aThermo Fisher Scientific Nicolet iS5 FTIR spectrometer with aniD3 ATR accessory. 1H NMR, 13C NMR, and 19F NMR spectrawere collected at room temperature with a Varian 400 MHz spec-trometer. UV/Vis spectra were collected with an Agilent 8453 diodearray spectrometer. Magnetic susceptibility measurements wereperformed at room temperature with an AUTO balance manufac-tured by Sherwood Scientific. Electrochemical measurements wereconducted in a Vacuum Atmospheres Omni-Lab glove box by usingan EC Epsilon potentiostat (iBAS) with 100 mm (NBu4)PF6 as thesupporting electrolyte. A three-electrode cell containing a Ag/AgClreference electrode, a platinum auxiliary electrode, and glass carbonworking electrode was employed for cyclic voltammetry (CV) andsquarewave (SW) measurements. The ferrocene/ferrocenium (Fc+/0)couple was employed as an internal standard. To facilitate compari-son to previously published data, redox potentials are reported ver-sus SCE. Under these conditions, Fc+/0 has an E1/2 value of 0.46 V(ΔE = 0.21 V) in CH2Cl2 and 0.45 V (ΔE = 0.12 V) in DMF.[29]

N,N-Bis(4,6-dimethylpyridin-2-yl)pyridine-2,6-dicarboxamide (H2LB):A solution of 2-amino-4,6-dimethylpyridine (1.20 g, 9.8 mmol) andtriethylamine (2.2 mL, 14.7 mmol) in CH2Cl2 (20 mL) was slowlyadded to a solution of pyridine-2,6-dicarbonyl dichloride (1.0 g,4.9 mmol) in CH2Cl2 (Scheme 5). The mixture was allowed to stirfor 2 h, after which the solvent was removed under reduced pres-sure. The solid residue was recrystallized from methanol (20 mL)to provide the product as a white solid (1.67 g, 91%). 1H NMR


(CDCl3): δ = 11.19 (s, 2 H, NH), 8.48 (d, 2 H), 8.17 (s, 2 H), 8.14(t, 1 H), 6.80 (s, 2 H), 2.50 (s, 6 H, CH3), 2.39 (s, 6 H, CH3) ppm.13C NMR (CDCl3): δ = 161.7, 156.2, 150.4, 150.5, 148.7, 139.4,125.5, 120.7, 112.0, 23.8, 21.4 ppm. C21H21N5O2·H2O (393.44):calcd. C 64.11, H 5.89, N 17.80; found C 64.46, H 5.90, N 17.80.

Scheme 5. General synthesis of pyridine-2,6-dicarboxamide li-gands.

N,N-Bis(4,6-dimethylpyrimidin-2-yl)pyridine-2,6-dicarboxamide(H2LC): To a solution of 2-amino-4,6-dimethylpyrimidine (1.21 g,9.8 mmol), triethylamine (2.2 mL, 14.7 mmol) in CH2Cl2 (20 mL),was slowly added a solution of pyridine-2,6-dicarbonyl dichloride(1.0 g, 4.9 mmol) in CH2Cl2 (15 mL). The mixture was allowed tostir for 2 h, after which the solvent was removed under reducedpressure. The solid residue was recrystallized from methanol(20 mL), providing the product as a white solid (0.76 g, 40%). 1HNMR (CDCl3): δ = 10.69 (s, 2 H, NH), 8.54 (d, 2 H), 8.13 (t, 1H), 6.83 (s, 2 H), 2.51 (s, 12 H, CH3) ppm. 13C NMR (CDCl3): δ= 168.4, 161.0, 157.1, 148.6, 139.6, 126.4, 116.2, 24.1 ppm.C19H19N7O2·H2O (395.42): calcd. C 57.71, H 5.35, N 24.80; foundC 57.91, H 5.08, N 24.63.

N,N-Bis(7-methylquinolin-8-yl)pyridine-2,6-dicarboxamide (H2LD):A solution of 2-amino-7-methylquinoline (450 mg, 2.85 mmol) andtriethylamine (0.6 mL, 4.27 mmol) in CHCl3 (20 mL) was slowlyadded to a solution of pyridine-2,6-dicarbonyl dichloride (290 mg,1.42 mmol) in CHCl3 (15 mL). The mixture was allowed to stirovernight under reflux. After cooling to room temperature andwashing with water, the organic phase was dried with MgSO4. Thesolvent was removed under reduced pressure to provide a brownsolid. Purification by column chromatography on silica (EtOAc/MeOH, 10:1) provided the final product as a white powder(427 mg, 65%). 1H NMR (CDCl3): δ = 11.23 (s, 2 H, NH), 8.57(dd, 2 H), 8.37 (d, 2 H), 8.00 (t, 1 H), 7.86 (dd, 2 H), 7.49 (d, 2H), 7.38 (d, 2 H), 7.04 (q, 2 H), 2.51 (s, 6 H, CH3) ppm. 13C NMR(CDCl3): δ = 161.7, 149.3, 149.3, 143.3, 138.9, 135.6, 135.1, 131.2,130.3, 126.6, 125.2, 125.0, 120.3, 20.2 ppm. C27H21N5O2 (447.49):calcd. C 72.47, H 4.73, N 15.65; found C 68.45, H 4.70, N 14.36.H2LX-type ligands are known to be hygroscopic;[6,14,30] thus, thedisagreement in the elemental analysis indicates that H2O(≈1.5 equiv.) was present in the sample: C27H21N5O2·1.5H2O(474.51): calcd. C 68.34, H 5.10, N 14.75.

[CuCl2(LB{H}2)] (1B): CuCl2 (74.4 mg, 0.55 mmol) and H2LB

(200 mg, 0.55 mmol) were dissolved in CH2Cl2 (10 mL) to give agreen-colored solution that was stirred overnight. The solvent wasremoved under vacuum, and the resulting green solid was washedwith Et2O to yield the product (198 mg). X-ray-quality crystalswere obtained by slow evaporation of a concentrated CH2Cl2 solu-tion (198 mg, 72%). UV/Vis (CH2Cl2): λmax (ε, m–1 cm–1) = 880(sh.), 735 nm (180). IR (solid): ν̃ = 1650, 1631, 1606, 1595 cm–1.Elemental analysis indicates that a small amount of CH2Cl2(0.2 equiv.) remained in the sample after drying.C21H21Cl2CuN5O2·0.2CH2Cl2 (526.86): calcd. C 48.33, H 4.09, N13.29; found C 48.38, H 4.17, N 13.43.



[CuCl2(LC{H}2)] (1C): A mixture of CuCl2 (89 mg, 0.66 mmol) andH2LC (250 mg, 0.66 mmol) in THF (10 mL) was stirred overnightto afford a light green solution. After removal of the solvent undervacuum, the green solid was washed with Et2O to give the product(304 mg). Green crystals suitable for crystallographic studies wereobtained by vapor diffusion of Et2O into a concentrated CH2Cl2solution (304 mg, 90 %). UV/Vis (CH2Cl2): λmax (ε, m–1 cm–1) = 870(sh.), 720 nm (220). IR (solid): ν̃ = 1673, 1622, 1582, 1498 cm–1.Elemental analysis indicates that a small amount of CH2Cl2(0.2 equiv.) remained in the sample after drying.C19H19Cl2CuN7O2·0.2CH2Cl2 (528.84): calcd. C 43.61, H 3.70, N18.54; found C 43.78, H 4.00, N 18.24.

[CuCl2(LD{H}2)] (1D): A mixture of CuCl2 (64.1 mg, 0.48 mmol)and H2LD (200 mg, 0.48 mmol) in CHCl3 (10 mL) was stirred over-night in air. The solvent was removed under vacuum to give a dark-green solid that was washed with Et2O. X-ray-quality crystals weregrown by layering a concentrated CH2Cl2 solution with pentane(212 mg, 80%). UV/Vis (CH2Cl2): λmax (ε, m–1 cm–1) = 660 nm(130). IR (solid): ν̃ = 1621, 1584, 1545, 1499 cm–1.C27H21Cl2CuN5O2 (581.94): calcd. C 55.73, H 3.64, N 12.03; foundC 55.62, H 3.70, N 12.48.

[Cu(LB{H}2)(H2O)(OTf)]OTf (2B): A mixture of Cu(OTf)2

(100 mg, 0.276 mmol) and H2LB (108 mg, 0.277 mmol) in MeCN(10 mL) was stirred overnight to provide a light green solution.After removing the solvent under vacuum, the resulting green solidwas washed with Et2O. X-ray-quality crystals were grown by vapordiffusion of Et2O into a MeCN solution (157 mg, 79%). UV/Vis(MeOH): λmax (ε, m–1 cm–1) = 610 nm (110). IR (solid): ν̃ = 1656,1626, 1604, 1592 cm–1. 19F NMR (MeCN): δ = –79.2 ppm.C23H23CuF6N5O9S2 (755.13): calcd. C 36.58, H 3.07, N 9.27; foundC 36.69, H 2.99, N 9.37.

[Cu(LC{H}2)(H2O)(OTf)2] (2C): A mixture of Cu(OTf)2 (240 mg,0.662 mmol) and H2LC (250 mg, 0.662 mmol) in THF (10 mL) wasstirred overnight in a glove box to provide a light blue solution.After removing the solvent under vacuum, the blue solid waswashed with Et2O. X-ray-quality crystals were prepared by vapordiffusion of Et2O into a concentrated MeCN solution (422 mg,86%). UV/Vis (MeOH): λmax (ε, m–1 cm–1) = 695 nm (105). IR (so-lid): ν̃ = 3434, 1765, 1623, 1582, 1501 cm–1. 19F NMR (MeCN): δ= –79.4 ppm. C21H21CuF6N7O9S2 (757.10): calcd. C 33.31, H 2.80,N 12.95; found C 33.53, H 2.88, N 12.68.

[CuCl(LB{H}2)(H2O)]OTf (3B): A mixture of CuCl2 (53.4 mg,0.40 mmol) and H2LB (150 mg, 0.40 mmol) in THF (10 mL) wasstirred for 1 h in a glove box to afford a green-colored solution.AgOTf (102.8 mg, 0.40 mmol) was then added, which resulted inthe formation of a white precipitate. The solution was stirred for1 h and then filtered, and the solvent was removed under vacuum.The resulting green solid was washed with Et2O. Green crystalswere obtained by layering a CH2Cl2 solution with pentane (125 mg,50%). UV/Vis (MeOH): λmax (ε, m–1 cm–1) = 703 nm (155). IR (so-lid): ν̃ = 1641, 1624, 1600, 1578 cm–1. 19F NMR (MeCN): δ =–79.3 ppm. The H2O ligand was removed with drying under vac-uum. C22H21ClCuF3N5O5S (623.50): calcd. C 42.38, H 3.39, N11.23; found C 42.57, H 3.30, N 11.22.

[Cu2(LB{H})2(OTf)2] (4B): A mixture of Cu(OTf)2 (100 mg,0.277 mmol) and H2LB (100 mg, 0.277 mmol) in MeCN (10 mL)was stirred overnight to provide a deep green solution. After re-moving the solvent under vacuum, the resulting green solid waswashed with Et2O. Vapor diffusion of Et2O into a MeOH solutionprovided dark violet crystals suitable for XRD analysis (110 mg,66%). UV/Vis (CH2Cl2): λmax (ε, m–1 cm–1) = 601 nm (620). IR (so-lid): ν̃ = 1671, 1614, 1590, 1579 cm–1. 19F NMR (MeCN): δ =


–79.3 ppm. C44H40Cu2F6N10O10S2 (1174.07): calcd. C 45.01, H3.43, N 11.93; found C 44.99, H 3.31, N 12.02.

[Cu2(LB)2] (5B): A mixture of CuCl2 (55.8 mg,0.415 mmol),NaOMe (56 mg, 1.04 mmol), and H2LB (150 mg, 0.415 mmol) inMeOH (10 mL) was stirred overnight to afford a deep blue solu-tion. Removal of the solvent under vacuum gave a deep blue solidthat was redissolved in CH2Cl2, filtered to remove the salt byprod-ucts, and dried under vacuum. X-ray-quality crystals were obtainedby vapor diffusion of Et2O into MeOH (106 mg, 58%). UV/Vis(CH2Cl2): λmax (ε, m–1 cm–1) = 597 nm (570). IR (solid): ν̃ = 1630,1614, 1596, 1551 cm–1. C42H38Cu2N10O4 (873.91): calcd. C 57.72,H 4.38, N 16.03; found C 57.42, H 4.78, N 14.47. The discrepancyin the N value is likely due to residual MeOH from the recrystalli-zation.

[Cu2(LC)2] (5C): A mixture of CuCl2 (107 mg,0.79 mmol), NaOMe(86 mg, 1.59 mmol), and H2LC (300 mg, 0.79 mmol) in MeOH(10 mL) was stirred for 30 min to afford a dark red solution. Re-moval of the solvent under vacuum provided a purple solid thatwas redissolved in CH2Cl2, filtered to remove the salt byproducts,and dried under vacuum. X-ray-quality crystals were obtained bylayering a CH2Cl2 solution with pentane. The insolubility of 5C innearly all solvents (except MeOH) made it difficult to eliminate theNaCl byproduct, which thus prevented isolation of the analyticallypure solid. UV/Vis (MeOH): λmax = 601 nm. IR (neat): ν̃ = 1678,1637, 1585, 1532 cm–1.

[Cu(LD)] (6D): A mixture of CuCl2 (64 mg, 0.48 mmol), NaOMe(0.96 mmol), and H2LD (200 mg, 0.48 mmol) in MeOH (10 mL)was stirred for 30 min. Removal of MeOH under vacuum provideda solid material that was redissolved with CH2Cl2. The solutionwas filtered to remove insoluble material, and the solvent was evap-orated to give a dark green solid. X-ray-quality crystals were grownby vapor diffusion of Et2O into a solution of 1,2-dicholorethane(198 mg, 79%). UV/Vis (MeCN): λmax (ε, m–1 cm–1) = 655 (210),830 nm (sh.). IR (solid): ν̃ = 1607, 1581, 1567, 1498 cm–1. Complex6D cocrystallizes with H2O, and elemental analysis indicated thatH2O (≈1.5 equiv.) remained in the sample after drying.C27H19CuN5O2·1.5H2O (536.04): calcd. C 60.49, H 4.14, N 13.07;found C 60.19, H 3.94, N 12.78.

Structure Determination: Each complex was characterized with X-ray crystallography; details concerning the data collection andanalysis are summarized in Tables 5 and 6. The X-ray diffractiondata were collected at 100 K with an Oxford Diffraction Super-Nova kappa-diffractometer equipped with dual microfocus Cu/MoX-ray sources, X-ray mirror optics, Atlas CCD detector, and a low-temperature Cryojet device. The data were processed with CrysAli-sPro program package (Oxford Diffraction Ltd., 2010) typically byusing a numerical Gaussian absorption correction (based on thereal shape of the crystal) followed by an empirical multi-scan cor-rection using SCALE3 ABSPACK routine. The structures weresolved by using the SHELXS program and refined with theSHELXL program[31] within the Olex2 crystallographic package.[32]

All computations were performed on an Intel PC computer withWindows 7 OS. Some structures contain disorder that was detectedin difference Fourier syntheses of electron density and accountedfor by using capabilities of the SHELX package. In most cases,hydrogen atoms were localized in difference syntheses of electrondensity but were refined by using appropriate geometric restrictionson the corresponding bond lengths and bond angles within a ri-ding/rotating model (torsion angles of methyl hydrogen atoms wereoptimized to better fit the residual electron density).

CCDC-927711 (for 1B), -927712 (for 1C), -927713 (for 1D), -927714(for 2B), -927715 (for 2C), -927716 (for 3B), -927717 (for 4B),



Table 5. Summary of X-ray crystallographic data collection and structure refinement.

1B 1C·0.5CH2Cl2 1D·CH2Cl2 2B 2C

Empirical formula C21H21Cl2CuN5O2 C19.5H20Cl3CuN7O2 C28H23Cl4CuN5O2 C23H23CuF6N5O9S2 C21H21CuF6N7O9S2

Formula weight 509.88 554.32 666.85 755.12 757.11Crystal system triclinic triclinic triclinic orthorhombic triclinicSpace group P1̄ P1̄ P1̄ Pna21 P1̄a [Å] 7.1019(3) 12.7826(3) 9.4544(7) 28.0304(9) 8.3766(3)b [Å] 9.8288(4) 15.0326(3) 10.3822(8) 13.2373(5) 9.8949(4)c [Å] 15.9843(6) 25.2949(4) 15.901(1) 7.9327(2) 19.0738(8)α [°] 72.240(4) 89.226(1) 106.402(7) 90 98.076(3)β [°] 82.180(3) 86.712(2) 102.062(7) 90 95.700(3)γ [°] 87.994(3) 77.779(2) 99.764(6) 90 113.151(4)Volume [Å3] 1052.66(7) 4742.6(2) 1420.1(2) 2943.40(16) 1418.3(1)Z 2 8 2 4 2ρcalcd. [g cm–3] 1.609 1.553 1.56 1.704 1.773μ [mm–1] 4.052 4.685 4.85 3.273 3.419Θ range [°] 9.4 to 147.7 7.0 to 147.8 9.2 to 148.1 7.4 to 147.4 9.5 to 147.6Reflections collected 15887 95897 13911 15571 21838Independent reflections (Rint) 4178 (0.0298) 18936 (0.0330) 5595 (0.0630) 5651 (0.0422) 5646 (0.0326)Data/restraints/parameters 4178/0/284 18936/3/1209 5595/15/418 5651/1/421 5646/0/435Goodness-of-fit on F2 1.032 1.016 1.013 1.112 1.022R1/wR2 indexes [I � 2σ(I)][a] 0.0353/0.0919 0.0325/0.0871 0.0467/0.1044 0.1007/0.2383 0.0333/0.0831R1/wR2 indexes [all data][a] 0.0412/0.0965 0.0372/0.0913 0.0769/0.1207 0.1026/0.2395 0.0408/0.0891

[a] R1 = Σ||Fo| – |Fc||/Σ|Fo|; wR2 = [Σw(Fo2 – Fc

2)2/Σw(Fo2)2]1/2.

Table 6. Summary of X-ray crystallographic data collection and structure refinement.

3B 4B·MeOH 5B 5C 6D·H2O

Empirical formula C22H23ClCuF3N5O6S C45H44Cu2F6N10O11S2 C42H38Cu2N10O4 C38H34Cu2N14O4 C27H21CuN5O3

Formula weight 641.50 1206.11 873.90 877.87 527.04Crystal system triclinic triclinic monoclinic monoclinic monoclinicSpace group P1̄ P1̄ C2/c C2/c C2/ca [Å] 10.3358(6) 11.1489(3) 20.1899(4) 19.7551(9) 20.3092(13)b [Å] 11.0707(6) 15.5689(5) 10.1779(1) 9.8198(4) 10.0074(4)c [Å] 13.0014(8) 15.6873(6) 19.1402(4) 19.7188(9) 12.8234(9)α [°] 67.749(5) 98.963(3) 90 90 90β [°] 69.068(5) 93.322(3) 102.390(2) 103.908(5) 118.907(9)γ [°] 76.349(5) 105.789(3) 90 90 90Volume [Å3] 1277.3(1) 2573.8(2) 3841.5(1) 3713.1(3) 2281.5(2)Z 2 2 4 4 4ρcalcd. [g cm–3] 1.668 1.554 1.511 1.57 1.521μ [mm–1] 3.579 2.564 1.165 1.941 1.690Θ range [°] 7.7 to 147.26 7.52 to 147.52 5.96 to 58.22 9.22 to 147.66 10 to 148Reflections collected 11783 16501 21840 14296 11433Independent reflections (Rint) 4998 (0.0217) 16501 (0.0000) 4672 (0.0337) 3688 (0.0380) 2274 (0.0364)Data/restraints/parameters 4998/0/364 16501/18/704 4672/0/266 3688/0/266 2274/36/151Goodness-of-fit on F2 1.040 1.076 1.049 1.166 1.041R1/wR2 indexes [I � 2σ(I)][a] 0.0374/0.0972 0.0917/0.2407 0.0316/0.0717 0.0944/0.2205 0.0467/0.1217R1/wR2 indexes [all data][a] 0.0420/0.1017 0.0953/0.2474 0.0383/0.0753 0.0977/0.2222 0.0550/0.1295

[a] R1 = Σ||Fo| – |Fc||/Σ|Fo|; wR2 = [Σw(Fo2 – Fc

2)2/Σw(Fo2)2]1/2.

-927718 (for 5B), -927719 (for 5C), and 928161 (for 6D) contain thesupplementary crystallographic data for this paper. These data canbe obtained free of charge from The Cambridge CrystallographicData Centre via www.ccdc.cam.ac.uk/data_request/cif.

Supporting Information (see footnote on the first page of this arti-cle): Ellipsoid plots of 5C and 6D; metric parameters for 6D; elec-tronic absorption spectra of 1X, 4B, and 5B in various solvents;cyclic voltammograms of 4B and 5B.

DFT Calculations: DFT computations were performed by using theORCA 2.7 software package developed by Neese.[33] In each case,the corresponding X-ray structure provided the starting point forgeometry optimizations and the computational model included theentire complex (excluding uncoordinated solvent molecules). Thecalculations employed the Becke–Perdew (BP86) functional.[34]


Ahlrichs’ valence triple-ζ basis set (TZV) was used for all atoms,in conjunction with the TZV/J auxiliary basis set.[35] Extra polar-ization functions were included on non-hydrogen atoms.

Acknowledgments

This research is generously supported by Marquette University andthe U. S. National Science Foundation (NSF) (CHE-1056845)

[1] a) A. Rajput, R. Mukherjee, Coord. Chem. Rev. 2013, 257, 350–368; b) H. L. Kwong, H. L. Yeung, C. T. Yeung, W. S. Lee, C. S.Lee, W. L. Wong, Coord. Chem. Rev. 2007, 251, 2188–2222; c)O. Belda, C. Moberg, Coord. Chem. Rev. 2005, 249, 727–740.

[2] A. K. Patra, R. Mukherjee, Inorg. Chem. 1999, 38, 1388–1393.



[3] a) D. S. Marlin, M. M. Olmstead, P. K. Mascharak, Inorg.Chem. 2001, 40, 7003–7008; b) D. S. Marlin, M. M. Olmstead,P. K. Mascharak, Angew. Chem. 2001, 113, 4888–4890; Angew.Chem. Int. Ed. 2001, 40, 4752–4753; c) J. C. Noveron, M. M.Olmstead, P. K. Mascharak, J. Am. Chem. Soc. 2001, 123,3247–3259; d) D. S. Marlin, P. K. Mascharak, Chem. Soc. Rev.2000, 29, 69–74; e) F. A. Chavez, M. M. Olmstead, P. K. Mas-charak, Inorg. Chim. Acta 1998, 269, 269–273.

[4] a) P. J. Donoghue, A. K. Gupta, D. W. Boyce, C. J. Cramer,W. B. Tolman, J. Am. Chem. Soc. 2010, 132, 15869–15871; b)P. J. Donoghue, J. Tehranchi, C. J. Cramer, R. Sarangi, E. I.Solomon, W. B. Tolman, J. Am. Chem. Soc. 2011, 133, 17602–17605.

[5] a) D. Huang, O. V. Makhlynets, L. L. Tan, S. C. Lee, E. V.Rybak-Akimova, R. H. Holm, Inorg. Chem. 2011, 50, 10070–10081; b) D. G. Huang, R. H. Holm, J. Am. Chem. Soc. 2010,132, 4693–4701.

[6] S. L. Jain, P. Bhattacharyya, H. L. Milton, A. M. Z. Slawin,J. A. Crayston, J. D. Woollins, Dalton Trans. 2004, 862–871.

[7] a) B. M. Trost, I. Hachiya, J. Am. Chem. Soc. 1998, 120, 1104–1105; b) O. Belda, C. Moberg, Acc. Chem. Res. 2004, 37, 159–167.

[8] H. Adolfsson, C. Moberg, Tetrahedron: Asymmetry 1995, 6,2023–2031.

[9] a) C. M. Che, W. K. Cheng, J. Chem. Soc., Chem. Commun.1986, 1443–1444; b) L. Yang, Z. Q. Wu, L. Liang, X. G. Zhou,J. Organomet. Chem. 2009, 694, 2421–2426; c) L. Yang, R. N.Wei, R. Li, X. G. Zhou, J. L. Zuo, J. Mol. Catal. A 2007, 266,284–289; d) S. H. Lee, J. H. Han, H. Kwak, S. J. Lee, E. Y. Lee,H. J. Kim, J. H. Lee, C. Bae, S. N. Lee, Y. Kim, C. Kimal,Chem. Eur. J. 2007, 13, 9393–9398.

[10] a) C. K. Vance, A.-F. Miller, Biochemistry 2001, 40, 13079–13087; b) R. G. Hadt, N. Sun, N. M. Marshall, K. O. Hodgson,B. Hedman, Y. Lu, E. I. Solomon, J. Am. Chem. Soc. 2012,134, 16701–16716; c) N. M. Marshall, D. K. Garner, T. D. Wil-son, Y. G. Gao, H. Robinson, M. J. Nilges, Y. Lu, Nature 2009,462, 113–U127.

[11] a) I. Schlichting, J. Berendzen, K. Chu, A. M. Stock, S. A.Maves, D. E. Benson, R. M. Sweet, D. Ringe, G. A. Petsko,S. G. Sligar, Science 2000, 287, 1615–1622; b) S. Nagano, T. L.Poulos, J. Biol. Chem. 2005, 280, 31659–31663.

[12] a) D. W. Christianson, W. N. Lipscomb, Acc. Chem. Res. 1989,22, 62–69; b) M. Hernick, C. A. Fierke, Arch. Biochem. Bio-phys. 2005, 433, 71–84; c) J. D. Lipscomb, Curr. Opin. Struct.Biol. 2008, 18, 644–649.

[13] a) A. S. Borovik, Acc. Chem. Res. 2005, 38, 54–61; b) R. L.Shook, A. S. Borovik, Inorg. Chem. 2010, 49, 3646–3660.

[14] S. M. Redmore, C. E. F. Rickard, S. J. Webb, L. J. Wright, In-org. Chem. 1997, 36, 4743–4748.

[15] K. Gudasi, R. Vadavi, R. Shenoy, M. Patil, S. A. Patil, M.Nethaji, Inorg. Chim. Acta 2005, 358, 3799–3806.


[16] R. Q. Fan, P. Wang, Y. L. Yang, Y. J. Zhang, Y. B. Yin, W.Hasi, Polyhedron 2010, 29, 2862–2866.

[17] a) T. Kawamoto, O. Prakash, R. Ostrander, A. L. Rheingold,A. S. Borovik, Inorg. Chem. 1995, 34, 4294–4295; b) T. Kawam-oto, B. S. Hammes, R. Ostrander, A. L. Rheingold, A. S. Boro-vik, Inorg. Chem. 1998, 37, 3424–3427; c) Z. Shirin, J. Thomp-son, L. Liable-Sands, G. P. A. Yap, A. L. Rheingold, A. S.Borovik, J. Chem. Soc., Dalton Trans. 2002, 1714–1720.

[18] a) S. Y. Shaban, F. W. Heinemann, R. van Eldik, Eur. J. Inorg.Chem. 2009, 3111–3118; b) R. Balamurugan, M. Palanianda-var, M. A. Halcrow, Polyhedron 2006, 25, 1077–1088.

[19] a) U. Geiser, R. M. Gaura, R. D. Willett, D. X. West, Inorg.Chem. 1986, 25, 4203–4212; b) L. P. Battaglia, A. B. Corradi,U. Geiser, R. D. Willett, A. Motori, F. Sandrolini, L. Antolini,T. Manfredini, L. Menabue, G. C. Pellacani, J. Chem. Soc.,Dalton Trans. 1988, 265–271; c) F. R. Fronczek, A. Mamo, S.Pappalardo, Inorg. Chem. 1989, 28, 1419–1422.

[20] Analysis of five-coordinate CuCl2N3 structures in the Cam-bridge Structural Database revealed a median Cu–Cl distanceof 2.31 Å.

[21] A. W. Addison, T. N. Rao, J. Reedijk, J. Vanrijn, G. C. Ver-schoor, J. Chem. Soc., Dalton Trans. 1984, 1349–1356.

[22] J. M. Rowland, M. M. Olmstead, P. K. Mascharak, Inorg.Chem. 2000, 39, 5326–5332.

[23] S. Sinnecker, F. Neese, L. Noodleman, W. Lubitz, J. Am. Chem.Soc. 2004, 126, 2613–2622.

[24] A. B. P. Lever, Inorganic Electronic Spectroscopy, Elsevier, Am-sterdam, 1968.

[25] a) A. A. G. Tomlinson, B. J. Hathaway, J. Chem. Soc. A 1968,2578–2583; b) A. A. G. Tomlinson, B. J. Hathaway, J. Chem.Soc. A 1968, 1905–1909.

[26] I. M. Procter, B. J. Hathaway, P. Nicholls, J. Chem. Soc. A 1968,1678–1684.

[27] F. B. Tamboura, M. Gaye, A. S. Sall, A. H. Barry, T. Jouini,Inorg. Chem. Commun. 2002, 5, 235–238.

[28] S. Flanagan, J. Dong, K. Haller, S. K. Wang, W. R. Scheidt,R. A. Scott, T. R. Webb, D. M. Stanbury, L. J. Wilson, J. Am.Chem. Soc. 1997, 119, 8857–8868.

[29] N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96, 877–910.[30] Q. Yu, T. E. Baroni, L. Liable-Sands, G. P. A. Yap, A. L.

Rheingold, A. S. Borovik, Chem. Commun. 1999, 1467–1468.[31] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.[32] O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard,

H. Puschmann, J. Appl. Crystallogr. 2009, 42, 339–341.[33] F. Neese, ORCA: An Ab Initio, Density Functional and Semiem-

pirical Program Package, v. 2.8, University of Bonn, Germany,2010.

[34] a) A. D. Becke, J. Chem. Phys. 1986, 84, 4524–4529; b) J. P.Perdew, Phys. Rev. B 1986, 33, 8822–8824.

[35] a) A. Schafer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100,5829–5835; b) A. Schafer, H. Horn, R. Ahlrichs, J. Chem. Phys.1992, 97, 2571–2577.

Received: May 3, 2013Published Online: �

Intramolecular Hydrogen Bonding in Cu II Complexes with 2,6-Pyridinedicarboxamide Ligands:...

Documents

Transcript of Intramolecular Hydrogen Bonding in Cu II Complexes with 2,6-Pyridinedicarboxamide Ligands:...