Synthesis, characterization, and reactivity studies of heterodinuclear complexes modeling active sites in purple acid phospatases

ARTICLE

pubs.acs.org/IC

Synthesis, Characterization, and Reactivity Studies of Heterodinuclear

Complexes Modeling Active Sites in Purple Acid Phospatases

Martin Jarenmark,*^,†Matti Haukka,^‡Serhiy Demeshko,^§Felix Tuczek,^||Luca Zuppiroli,^{^}Franc Meyer,^§and

Ebbe Nordlander*^,†

^†Inorganic Chemistry Research Group, Chemical Physics, Center for Chemistry and Chemical Engineering, Lund University, Box 124,

SE-221 00 Lund, Sweden

^‡Department of Chemistry, University of Eastern Finland, P.O. Box 111, FI-80101 Joensuu, Finland

^§Institute for Inorganic Chemistry, Georg-August-University G€ottingen, Tammanstrasse 4, D-37077 G€ottingen, Germany

Institute for Inorganic Chemistry, Christian-Albrechts-University Kiel, Olshausenstrasse 40, D-24098 Kiel, Germany

^{^}Dipartimento di Chimica “A. Mangini”, Universita di Bologna, Viale Risorgimento 4, I-40136 Bologna, Italy

S Supporting Information

b

ABSTRACT: To model the heterodinuclear active sites in plant

purple acid phosphatases, a mononuclear synthon, [Fe(III)-

(H₂IPCPMP)(Cl₂)][PF₆] (1), has been generated in situ from

the ligand 2-(N-isopropyl-N-((2-pyridyl)methyl)aminomethyl)-

6-(N-(carboxylmethyl)-N-((2-pyridyl)methyl)amino methyl)-4-

methylphenol (IPCPMP) and used to synthesize heterodinuclear

complexes of the formulas [Fe(III)M(II)(IPCPMP)(OAc)₂-

(CH₃OH)][PF₆] (M = Zn (2), Co (3), Ni (4), Mn (5)),

[Fe(III)Zn(II)(IPCPMP)(mpdp)][PF₆] (6) (mpdp = meta-

phenylene-dipropionate),

and

[Fe(III)Cu(II)(IPCPMP)

(OAc)}₂(μ-O)][PF₆] (7). Complexes 2ꢀ4, 6, and 7 have been

crystallographically characterized. The structure of 6 is a solid

state coordination polymer with heterodinuclear monomeric units, and 7 is a tetranuclear complex consisting of two heterodinuclear

phenolate-bridged Fe(III)Cu(II) units bridged through a μ-oxido group between the two Fe(III) ions. M€ossbauer spectra conﬁrm the

presence of high spin Fe(III) in an octahedral environment for 1, 3, and 5 while 2 and 4 display relaxation eﬀects. Magnetic

susceptibilitymeasurementsindicate weak antiferromagneticcoupling for3, 4, and 5 and conﬁrm theassignmentof the metalcentersin

2ꢀ5 as high spin Fe(III)-M(II) (M = Zn, Co (high spin), Ni (high spin), Mn (high spin)). Complexes 2ꢀ5 are intact in acetonitrile

solution as indicated by IR spectroscopy (for 2ꢀ4) and electrospray ionization mass spectrometry (ESI-MS) but partly dissociate to

hydroxide species and a mononuclear complex in water/acetonitrile solutions. UVꢀvis spectroscopy reveal pH-dependent behavior,

and species that form upon increasing the pH have been assigned to μ-hydroxido-bridged Fe(III)M(II) complexes for 2ꢀ5 although 2

and 3 is further transformed into what is propsed to be a μ-oxido-bridged tetranuclear complex similar to 7. Complexes 2ꢀ5 enhance

phosphodiester cleavage of 2-hydroxy-propyl-p-nitrophenyl phosphate (HPNP) and bis(2,4-dinitrophenyl)phosphate (BDNPP), but

the reactivities are diﬀerent for diﬀerent complexes and generally show strong pH dependence.

’ INTRODUCTION

are very similar.^13,14They are employed to hydrolyze substrates

such as urea (urease),^13,14peptide bonds in proteins (e.g., leucine

aminopeptidase¹⁵), phosphotriesters (phosphotriesterase¹⁶),

DNA and RNA (e.g., DNA polymerase and RNase H¹⁷). These

enzymes have active sites where two metals of the same kind bind

in close proximity to each other (about 2.9ꢀ4 Å apart).¹²Two

other members of the hydrolase superfamily, the purple acid

phosphatases (PAPs)^18ꢀ20and Ser/Thr Protein Phosphatases

(PPs),^12,18are special in that they have been shown to bind two

diﬀerent metal ions in their active sites. While the true (native)

Transition metal ions are extensively used by nature to

catalyze a broad range of reactions.¹While the presence of one

metal in the active site is suﬃcient for the activity of many

metalloenzymes, others have evolved to make use of a multi-

nuclear active site where the metals work in a cooperative way

with similar or very diﬀerent roles. Important examples of

protein/enzyme classes containing polynuclear metal sites in-

clude nitrogenases,^2,3oxygenases,^4,5hydrogenases,⁶ironꢀsulfur

proteins,⁷oxygen transport proteins,^8,9the oxygen evolving

complex of photosystem II,^10,11and di- and trinuclear

hydrolases.¹²The latter is a diverse group regarding structures,

substrate selectivity, and metal content but their modes of action

Received: October 6, 2010

Published: March 31, 2011

r

2011 American Chemical Society

3866

dx.doi.org/10.1021/ic1020324 Inorg. Chem. 2011, 50, 3866–3887

|

Inorganic Chemistry

ARTICLE

metal content of PPs has not been completely established, the

PAPs are known to have a requirement of one Fe(III) and a

divalent metal (Fe(II), Zn(II), or Mn(II)) in their active sites.

Purple acid phosphatases are nonspeciﬁc phosphomonoes-

terases isolated from various mammals and plants. The speciﬁc

functions of these enzymes appear to be dependent on the

organism. In mammals they have been associated with regulation

of bone resorption in osteoclasts²⁰and iron transport during

pregnancy.²¹Plant PAPs might have a role in recruiting phos-

phate in developing plants.²²Typical characteristics of PAP

enzymes are their intense pink/purple color, the low pH

optimum (∼pH 5ꢀ6) of their activity, and their resistance to

tartrate inhibition.^12,18,19Several PAPs from mammalian and

plant sources (human,²³pig,²⁴rat,^25,26kidney bean,^27,28and

sweet potato²⁹) have been structurally characterized by crystal-

lography. The mammalian enzymes have a heterovalent Fe-

(III)Fe(II) metal site in its active form, but only structures of

the inactive oxidized di-Fe(III) state have been determined so far.

The plant PAPs, which have very low sequence homology with

the mammalian enzymes, have two diﬀerent metal ions in the

active site, Fe(III) and either Zn(II)^30ꢀ32or Mn(II).^32,33Despite

the relatively low sequence homology between PAPs from

diﬀerent sources, the amino acid residues forming the coordina-

tion environment of the metals are conserved in all PAPs. A

notable feature of the active sites (Figure 1) is the unsymmetric

coordination of the two metals. The Fe(III) resides in an oxygen-

rich environment with the relatively hard donors aspartate and

phenolate; the latter amino acid residue is involved in a ligand to

metal charge transfer (LMCT) interaction with the Fe(III)

responsible for the intense color of the enzyme. The divalent

metal occupies a site with one more nitrogen donor and an

asparagine but no anionic ligands except the bridging hydroxido

moiety and a μ_1,1-κ¹(O) aspartate group. In sweet potato PAP

(spPAP), which contains an Fe(III)Mn(II) active site, the

observed strong antiferromagnetic coupling between the two

metal ions indicates a bridging oxido rather than a μ-hydroxido

ligand.³³In the general PAP active site, the divalent metal has a

water molecule coordinated to it, and the Fe(III) has been

modeled with a terminally coordinated hydroxido group,^25,27,28

although some spectroscopic evidence speak against this.³⁴This

discrepancy is also one of the issues that are discussed when it

comes to the mechanism(s) of these enzymes. Investigations

have led to the conclusion that a solvent-derived nucleophile

coordinated to (at least) the Fe(III) makes a direct attack on the

phosphorus atom of the substrate with inversion of its conﬁg-

uration, but whether this nucleophile is terminally coordinated or

bridging between the metals has not been established with

certainty. Also, the exact binding of the substrate prior to

hydrolysis is under debate and might be pH-dependent.¹⁸The

fact that the pig PAP uteroferrin, which in its native form contains

a Fe(III)Fe(II) core, and the kidney bean PAP (with an Fe-

(III)Zn(II) core) both exhibit diesterase activity, that is, hydro-

lyze both an organophosphate diester and the monoester that is

the product of the initial hydroslysis, has been interpreted to

mean that the phosphate ester coordinates in a terminal fashion

to the divalent metal of the enzyme active site, and that hydrolysis

is eﬀected by a hydroxyl nucleophile that is generated at and

terminally coordinated to the ferric ion,³⁵a mechanism that is

corroborated by reactivity studies of a dinuclear Fe(III)Zn(II)

model complex.³⁶In a recent investigation on uteroferrin and its

Fe(III)Mn(II) derivative, Schenk and co-workers³⁷studied the

hydrolysis of phenyl phosphate by the enzyme, using stopped-

Figure 1. Structure of the active site in kidney bean purple acid

phosphatase in the resting state based on crystallographic and spectro-

scopic data. Adapted from ref 18.

ﬂow spectrophotometry. The investigators found that substrate

hydrolysis occurred at a faster rate than the formation of a μ-1,3-

phosphate complex for both forms of the metalloenzyme. A

slower rate of interaction between the substrate and the ferric ion

relative to the catalytic rate was interpreted to mean that the

substrate is hydrolyzed when coordinated in a monodentate

fashion to the divalent metal, and the authors suggested that the

active hydroxide nucleophile is generated by the action of a

Fe(III)-coordinated hydroxide ion on a water molecule in the

second coordination sphere of the metal core, aided by histidine

residues capable of acting as Br€onsted bases (vide infra).

Small dinuclear metal complexes have been extensively used

to model metal sites in enzymes.^38ꢀ41The synthesis of heterodi-

nuclear complexes presents a special challenge because of the

possibility of disproportionation and selectivity problems when

the metals are coordinated to a dinucleating ligand; such

problems may result in the formation of mixtures of homo-

and heterodinuclear complexes, as well as metal site isomers.

These diﬃculties have been circumvented by designing ligands

that bind metals selectively in one pocket or strongly enough to

prevent disproportionation and metal mixing. Diﬀerent types of

ligands that have been successfully used to model PAPs are

illustrated in Figure 2. Symmetric ligands (BPMP,⁴²BPBP,⁴³

BPMOP⁴⁴etc., type (a) Figure 2) do not have the problem of

metal site selectivity and since they bind the metals strongly

because of the chelate eﬀect, selective formation of heterodi-

nuclear complexes is possible although investigation of true

species distribution in various solvents have not always been

reported. Unsymmetric ligands frequently utilize a hard and

anionic oxygen donor on one side to selectively bind the Fe(III),

and only nitrogen donors on the side that preferentially binds the

divalent metal (e.g., BPBPMP,⁴⁵BHBBPMP,⁴⁶Figure 2).

We have previously reported the two unsymmetric ligands

IPCPMP⁴⁷and ICIMP⁴⁸(Figure 2, type a) that incorporate a

terminal carboxylate donor as well as pyridyl and 1-methylimi-

dazolyl groups, respectively. The terminal carboxylate moiety

models the oxygen-rich environment often found in active sites

of metalloenzymes.¹On the opposite side, a noncordinating

group has been incorporated to further model the unsymmetric

coordination of the metal sites and to provide for a site on the

bound metal that is either vacant or ﬁlled by a (loosely)

coordinated ligand. The two ligands have been shown to

selectively bind Fe(III) on one side of the phenolate, forming

mononuclear Fe(III) complexes that have been isolated and

characterized, and that have been used as synthons for heterodi-

nuclear Fe(III)Zn(II) complexes that are capable of enhancing

the rate of transesteriﬁcation of 2-hydroxypropyl-p-nitrophenyl-

phosphate (HPNP) relative to the noncatalyzed reaction.⁴⁹Here

3867

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 2. Schematic depiction of ligands that have been used to model active sites in PAPs. Type (a) includes ICIMP (R1 = R4 = 1-metylimidazole, R2 =

CO₂^ꢀ, R3 = iso-propyl; R5 = CH₃) BPBPMP (R1 = R2 = R3 = 2-pyridyl, R4 = 2-phenolate, R5 = CH₃), BPMP (R1 = R2 = R3 = R4 = 2-pyridyl, R5 =

CH₃), BPBP (R1 = R2 = R3 = R4 = 2-pyridyl, R5 = t-butyl), BHBBPMP (R1 = R2 = 2-pyridyl, R3 = R4 = 2-phenol, R5 = CH₃), BPMOP (R1 = R2 = R3 =

R4 = 2-pyridyl, R5 = methoxy). Type (b) includes tdmbpo (R7 = R8 = 4,5-dimethyl)benzimidazolyl); thebpo (R7 = R8 = (2-hydro-

xyethyl)benzimidazolyl); bdmbbppo (R7 = (4,5-dimethyl)benzimidazolyl, R8 = 2-pyridyl. (c) Macrocyclic bis-phenolato ligand. The present study

involves type (a) IPCPMP (R1 = R4 = 2-pyridyl, R2 = CO₂^ꢀ, R3 = iso-propyl, R5 = CH₃) (d) 1,3-meta-phenylenedipropionate (mpdp).

Chart 1. Schematic Depiction of the New Complexes Presented in This Work

we report a detailed study of other heterodinuclear complexes of

the type Fe(III)M(II) (M(II) = Zn(II), Co(II), Ni(II), Cu(II),

and Mn(II)) that are based on IPCPMP, and further solution and

reactivity investigations to elucidate the mechanism(s) and

relevance of these complexes as models of the active sites in

purple acid phosphatases.

this synthetic method for 2 (FeZn) and 4 (FeNi), contamination

by tributyl ammonium hexafluorophoshate often interfered with

the isolation of the product. To circumvent this problem, a second

method (method B) was devised, in which an excess of sodium

acetate (6 equiv) dissolved in methanol was added dropwise.

Filtration through Celite removed the precipitated sodium salts,

and crystalline material of 2 and 4 could be grown in good yields

from methanol/2-propanol solvent mixtures. All analyses favor the

formulation of these complexes as having the same composition as

those obtained using the first method, except for the case of the

Fe(III)Cu(II) complex 7, where method B did not yield a pure

product.

Solid State Structures. The structures of 1 and 2 have been

reported previously,⁴⁹and ORTEP illustrations are shown in

Figure 3. The crystal structures of 3 and 4 [Fe(III)M(II)-

(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Co (3) and Ni (4),

Figure 4) are very similar to that of 2 (FeZn); crystallographic

data for 3 and 4 are summarized in Table 1, and relevant bond

distances and angles for 2ꢀ4 are listed in Table 2. In the case of 3

and 4, both metals are octahedrally coordinated in an N₂O₄

environment with one phenolate and two syn-syn-μ_1,3-carboxy-

late bridges. The Fe(III) ion is in all cases found on the side with

the hard terminal carboxylate donor of IPCPMP, and this

carboxylate coordinates in the anti mode while the usual

mode for carboxylate donors in biological systems is syn-

coordination.⁵¹A methanol molecule derived from the solvent

coordinates at the site on the divalent metal ion that is left open

by the noncoordinating isopropyl group. The bonds to the

Fe(III) ion are on average the same for 2ꢀ4 but shorter than

those to the divalent metal (FeꢀX_ave2ꢀ4: 2.04 Å; ZnꢀX_ave2:

2.12; 3: 2.12; 4: 2.08 Å) although the bond to the pyridyl.

’ RESULTS AND DISCUSSION

Syntheses. The complexes in this study have been synthesized

by two different methods, both of which proceed via the in situ

formation of the mononuclear synthon [Fe(III)(H₂IPCPMP)-

(Cl₂)][PF₆] (1)⁴⁹by addition of 1 equiv of FeCl₃to a solution of

the ligand. Addition of divalent metals (Zn, Co, Ni, Mn, Cu) with

appropriate carboxylate donors to this solution have made it

possible to isolate several different heterodinuclear complexes

(Chart 1) of the formula [Fe(III)M(II)(IPCPMP)(OAc)₂-

(CH₃OH)][PF₆] (M = Zn (2),⁴⁹Co (3), Ni (4), Mn (5)) as

wellastwootheroftheformulas[Fe(III)Zn(II)(IPCPMP)(mpdp)]-

[PF₆] (6) (mpdp = meta-phenylene-dipropionate, Figure 2d) and

[Fe(III)Cu(II)(IPCPMP)(OAc)}₂(μ-O)][PF₆] (7). To achieve

coordination of the divalent metal, the mononuclear synthon (1)

needs to be deprotonated. The first method (method A) used a

dropwise addition of 6 equiv of tributyl amine, during which the

intense blue color of 1 gradually changed to less intense red-purple

for 2ꢀ6 while for 7 (FeCu) the color turned red-brown. From

these solutions, X-ray quality crystals were grown for 2 (FeZn), 3

(FeCo), 4 (FeNi), 6 (FeZn-mpdp), and 7 (FeCu), but so far no

suitable crystals have been grown for 5 (FeMn) (the structure of

the analogous compound [FeMn(ICIMP)(OAc)₂Cl] has been

determined by X-ray crystallography⁵⁰). During efforts to scale up

3868

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 3. ORTEP⁹⁶representations of [Fe(III)(H₂IPCPMP)(Cl₂)][PF₆] (1) (a) and [Fe(III)Zn(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (2) (b)

from ref 49.

Figure 4. ORTEP representations of the crystal structures of [Fe(III)Co(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (3) (a) and

[Fe(III)Ni(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (4) (b) along with a Mercury⁹⁷overlay structure of complexes 2, 3, and 4 (c), where M(II) =

Zn(II), Co(II), or Ni(II), respectively. All solvent molecules and counterions have been omitted for clarity, and the ellipsoids are drawn at 50%

probability. The (R,R) enantiomers are shown in (a) and (b) while in (c) only the (S,S) enantiomers were overlaid.

nitrogen is actually significantly longer (Fe1ꢀN2 2: 2.165(3); 3:

2.162(2); 4: 2.258(3) Å) than the corresponding one to the

divalent metal ion (Zn1ꢀN4 2: 2.115(3); 3: 2.100(2); 4:

2.048(3) Å. Shorter bonds are to be expected for Fe(III) because

of its small size and high charge-to-radius ratio. Divalent metal

ions are considerably softer than the hard Fe(III) ion and hence

have stronger interactions with pyridyl groups that are relatively

soft compared to alkyl amines and oxygen groups. This effect is

also seen for 2 and 3 when comparing distances between the

metals and the relatively hard tertiary nitrogens, where the

metalꢀnitrogen distances are shorter for Fe(III) (Fe1ꢀN1 2,

3: 2.210(3), 2.10(2)Å) than for Zn(II) (2: Zn1ꢀN3 2.282(3) Å)

and Co(II) (3: Co1ꢀN3 2.268(3) Å). The same distance to

Ni(II) in 4 (Ni1ꢀN3 2.210(3) Å) is identical to the correspond-

ing FeꢀN distance (Fe1ꢀN1 2,209(3) Å) which might reflect

the higher charge-to-radius ratio of Ni(II) as compared to Zn(II)

and Co(II) (vide infra). The shortened bonds to Fe(III) also

make all three bridging moieties unsymmetrically coordinated

between the two metals with significantly longer bonds to the

divalent metal ion (Average FeꢀO_bridge2: 1.95, 3: 1.96; 4:

1.96 Å; M(II)ꢀO_bridge2(Zn): 2.10; 3(Co): 2.08; 4(Ni): 2.06 Å).

Also, the two carboxylate bridges coordinate at slightly different

distances to the Fe(III) with the one trans to the tertiary amine at

shorter distance, as observed by others.⁵²It is a bit surprising to

note that the distances between metals and tertiary nitrogens are

longer (average for 2ꢀ4: Fe1ꢀN1 2.21 Å, Zn1ꢀN3 2.25 Å) than

the corresponding ones to pyridyl nitrogens (average for 2ꢀ4:

Fe1ꢀN2 2.16 Å, Zn1ꢀN4 2.09 Å). This has previously been

ascribed to the strain introduced by the limited length of the one-

carbon linker between the tertiary nitrogen and the three

attached donor groups⁵³(e.g., phenolate, pyridyl, carboxylate).

Also, the clamping effect of the bridges will pull the metals away

from the tertiary nitrogen that has limited flexibility because of its

linkage to the phenol ring.

To facilitate the comparison of complexes 2 (FeZn), 3

(FeCo), and 4 (FeNi), an overlay plot of their structures is

shown in Figure 4c. It is apparent that they are virtually identical,

with root-mean-square deviations varying from 0.039 to 0.054

depending on which two structures are compared. A detailed

comparison of the bond lengths (Table 2) shows that the only

signiﬁcant diﬀerences are observed for the bonds to the divalent

metal ion. The Fe(III)ꢀX distances (X = N- or O-donor)

average to 2.038, 2.040, and 2.040 Å for 2, 3, and 4, respectively.

All bonds to Ni(II) in 4 (NiꢀX_ave= 2.083 Å) are shorter than the

corresponding ones to Zn(II) and Co(II) in 2 and 3, the latter

ones being quite similar (ZnꢀX_ave= 2.129 Å; CoꢀX_ave= 2.116 Å).

This is likely an eﬀect of the smaller ionic radius of Ni(II)

(0.83 Å vs 0.88 and 0.885 Å for Zn(II) and high spin Co(II),

3869

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Table 1. Crystallographic Data for Complex 2, 3, 4, 6, and 7 (L = IPCPMP)

[FeZnL(OAc)₂

(CH₃OH)]PF₆

[FeCoL(OAc)₂

(CH₃OH)]PF₆

[FeNiL(OAc)₂

(CH₃OH)]PF₆

[FeZnL(mpdp)]

PF₆

[{FeCuL(OAc)}₂

(μ-O)]PF₆(CH₃OH)₂

3

2

3

4

6

7

empirical formula

M_r(g/mol)

crystal size (mm³)

crystal system

space group

a (Å)

C

₃₁H₄₀F₆FeN₄O₈PZn

C₃₁H₄₀CoF₆FeN₄O₈P

C₃₁H₄₀F₆FeN₄NiO₈P

C₃₈H₄₂F₆FeN₄O₇PZn C₆₀H₈₂Cu₂F₁₂Fe₂N₈O₁₅P₂

862.86

856.42

856.20

932.95

1684.06

0.14 ꢁ 0.08 ꢁ 0.07

monoclinic

P2₁/n

0.24 ꢁ 0.08 ꢁ 0.06

monoclinic

Pn

0.29 ꢁ 0.14 ꢁ 0.12

monoclinic

Pn

0.30 ꢁ 0.29 ꢁ 0.20

monoclinic

Pn

0.20 ꢁ 0.13 ꢁ 0.13

monoclinic

P2₁/c

10.4594(5)

13.6060(9)

12.7267(7)

90

10.4714(2)

13.6885(4)

12.6895(4)

90

10.4533(3)

13.6671(4)

12.6719(2)

90

18.9033(6)

13.7500(7)

14.9602(5)

90

12.4979(6)

22.0221(12)

12.9244(4)

90

b (Å)

c (Å)

R (deg)

β (deg)

99.665(3)

90

99.067(2)

90

98.781(2)

90

90.290(2)

90

103.126(3)

90

γ (deg)

volume (Å³)

1785.44(18)

2

1796.16(9)

2

1789.17(8)

2

3888.4(3)

4

3464.2(3)

2

Z

F

_calc.(g/cm³)

abs. coeﬃcient

(mm^ꢀ1

F(000)

1.605

1.584

1.589

1.594

1.614

1.209

0.995

1.062

1.115

1.164

)

886

880

882

1916

1732

wavelength (Å)

temperature (K)

θ range (deg)

index ranges

0.71073

120(2)

3.58 to 27.53

ꢀ13 e h e 12

ꢀ17 e k e 17

ꢀ16 e l e 16

15909

2.20 to 27.49

ꢀ13 e h e 13

ꢀ17 e k e 17

ꢀ16 e l e 16

41179

3.57 to 27.52

ꢀ13 e h e 13

ꢀ17 e k e 17

ꢀ16 e l e 16

30688

2.01 to 25.01

ꢀ20 e h e 22

ꢀ15 e k e 16

ꢀ17 e l e 17

22376

1.86 to 25.35

ꢀ14 e h e 15

ꢀ26 e k e 26

ꢀ15 e l e 15

27601

reﬂns collected

indep. reﬂns [R_int

compltnss to

θ = 27.53ꢀ

]

7003 [0.0379]

99.2%

8155 [0.0439]

99.9%

8007 [0.0375]

99.5%

6768 [0.0635]

98.7%

6330 [0.0971]

99.8%

absorpn correction

semiempirical

from equivalents

0.9267 and 0.7610

from equivalents

0.8941 and 0.7598

from equivalents

0.8181 and 0.7433

from equivalents

0.8668 and 0.8045

from equivalents

0.9240 and 0.8521

max. and min

transmisn

reﬁnement method

full-matrix least-

squares on F²

7003/2/495

full-matrix least-

squares on F²

8155/51/488

full-matrix least-

squares on F²

8007/50/488

full-matrix least-

squares on F²

6768/0/526

full-matrix least-

squares on F²

6330/0/466

data/restrnts/

paramtrs

goodness-of-

ﬁt on F²

1.015

1.035

1.029

1.005

1.026

ﬁnal R indices

[I > 2σ(I)]

R1 = 0.0383

R1 = 0.0367

R1 = 0.0393

R1 = 0.0352

R1 = 0.0515

wR2 = 0.0846

R1 = 0.0509

wR2 = 0.0868

R1 = 0.0456

wR2 = 0.0931

R1 = 0.0497

wR2 = 0.0811

R1 = 0.0590

wR2 = 0.1190

R1 = 0.0948

R indices (all data)

wR2 = 0.0903

0.568/ꢀ0.445

wR2 = 0.0916

0.627/ꢀ0.564

wR2 = 0.0976

0.624/ꢀ0.456

wR2 = 0.0908

0.408/ꢀ0.403

wR2 = 0.1377

0.929/ꢀ0.707

resid electrn dens

(e Å^ꢀ3

)

respectively)⁵⁴as noted previously for similar FeNi and FeZn

complexes of the ligand BPMP.⁵⁵The coordination environ-

ments of the divalent metals in 2 (Zn(II)) and 3 (Co(II)) diﬀer

only with regard to the bond to the phenolic oxygen, which is

longest in 2 (2.137(2) Å) and of similar lengths in 3 and 4

(2.075(2) and 2.068(2) Å, respectively). This diﬀerence is

probably due to the π-donation from the p-orbitals of the

phenolic oxygen to suitable metal d-orbitals, which is absent

for Zn(II) because of its d¹⁰conﬁguration but present for the

other two metals. The metalꢀmetal distances are quite similar,

namely, 3.4556(6)Å, 3.4351(5), and 3.4273(5) Å for 2, 3, and 4,

respectively, and the small diﬀerences that are observed may be

3870

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Table 2. Selected Bond Distances (Å) and Bond Angles (deg) for [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2),

Co (3), Ni (4)), [Fe(III)Zn(II)(IPCPMP)(mpdp)][PF₆] (6), and [Fe(III)Cu(II)(IPCPMP)(OAc)}₂(μ-O)][PF₆] (7)

2

3

4

6

7

Zn1ꢀO1

Zn1ꢀO3

Zn1ꢀO5

Zn1ꢀO8

Zn1ꢀN3

Zn1ꢀN4

Fe1ꢀO1

Fe1ꢀO2

Fe1ꢀO4

Fe1ꢀO6

Fe1ꢀN1

Fe1ꢀN2

Fe1ꢀZn1^a

2.137(2)

2.088(3)

2.061(3)

2.089(3)

2.282(3)

2.115(3)

1.952(2)

1.971(3)

1.932(3)

1.999(3)

2.210(3)

2.165(3)

3.4556(6)

Co1ꢀO1

Co1ꢀO3

Co1ꢀO5

Co1ꢀO8

Co1ꢀN3

Co1ꢀN4

Fe1ꢀO1

Fe1ꢀO2

Fe1ꢀO4

Fe1ꢀO6

Fe1ꢀN1

Fe1ꢀN2

Fe1ꢀCo1^a

2.075(2)

2.085(2)

2.088(2)

2.079(2)

2.268(3)

2.100(2)

1.972(2)

1.971(2)

1.932(2)

1.994(2)

2.210(2)

2.162(2)

3.4351(5)

Ni1ꢀO1

Ni1ꢀO3

Ni1ꢀO5

Ni1ꢀO8

Ni1ꢀN3

Ni1ꢀN4

Fe1ꢀO1

Fe1ꢀO2

Fe1ꢀO4

Fe1ꢀO6

Fe1ꢀN1

Fe1ꢀN2

Fe1ꢀNi1^a

2.068(2)

2.055(2)

2.048(2)

2.069(2)

2.210(3)

2.048(3)

1.964(2)

1.976(2)

1.934(2)

1.997(2)

2.209(3)

2.158(3)

3.4273(5)

Zn1ꢀO1

Zn1ꢀO3

Zn1ꢀO5

Zn1ꢀO7

Zn1ꢀN3

Zn1ꢀN4

Fe1ꢀO1

Fe1ꢀO2

Fe1ꢀO4

Fe1ꢀO6

Fe1ꢀN1

Fe1ꢀN2

Fe1ꢀZn1^a

Fe1ꢀZn#1

2.1070(18) Cu1ꢀO1

2.1149(19) Cu1ꢀO3

2.088(2)

1.909(3)

1.935(3)

2.0270(19)

2.319(2)

2.120(2)

Cu1ꢀN3

Cu1ꢀN4

2.022(4)

1.946(4)

2.047(3)

2.072(3)

2.000(3)

1.7934(6)

2.255(4)

2.156(4)

3.2981(9)

3.5868(8)

1.9360(18) Fe1ꢀO1

1.953(2)

Fe1ꢀO2

1.9961(18) Fe1ꢀO6

2.0265(18) Fe1ꢀO8

2.198(2)

2.144(2)

3.4695(5)

5.5799(5)

Fe1ꢀN1

Fe1ꢀN2

Fe1ꢀCu1^a

Fe1ꢀFe1#1^a

Fe1ꢀO1ꢀZn1 115.27(11) Fe1ꢀO1ꢀCo1 116.10(9)

Fe1ꢀO1ꢀNi1 116.43(10) Fe1ꢀO1ꢀZn1 118.16(9)

Cu1ꢀO1ꢀFe1

112.93(14)

N4ꢀZn1ꢀN3 80.54(12)

N4ꢀZn1ꢀO1 87.09(10)

O1ꢀZn1ꢀN3 88.00(10)

N4ꢀCo1ꢀN3 80.39(9)

O1ꢀCo1ꢀN3 89.03(8)

O1ꢀCo1ꢀN4 88.01(8)

N4ꢀNi1ꢀN3 82.37(10)

N4ꢀNi1ꢀO1 88.26(10)

N4ꢀNi1ꢀO3 90.05(11)

N4ꢀNi1ꢀO8 89.78(10)

O1ꢀNi1ꢀN3 89.77(9)

N4ꢀZn1ꢀN3 79.68(9)

O1ꢀZn1ꢀN3 87.67(8)

O1ꢀZn1ꢀN4 89.62(8)

O1ꢀZn1ꢀO3 89.74(7)

O3ꢀZn1ꢀN3 102.80(8)

Fe1ꢀO8ꢀFe1#1 180.0

85.73(15)

N4ꢀCu1ꢀN3

O1ꢀCu1ꢀN3

O1ꢀCu1ꢀN4

O1ꢀCu1ꢀO3

O3ꢀCu1ꢀN3

O3ꢀCu1ꢀN4

N2ꢀFe1ꢀN1

O1ꢀFe1ꢀN1

O1ꢀFe1ꢀN2

O1ꢀFe1ꢀO2

O2ꢀFe1ꢀN1

O2ꢀFe1ꢀN2

O6ꢀFe1ꢀN1

O6ꢀFe1ꢀN2

O6ꢀFe1ꢀO1

O6ꢀFe1ꢀO2

O8ꢀFe1ꢀN1

O8ꢀFe1ꢀN2

O8ꢀFe1ꢀO1

O8ꢀFe1ꢀO2

O8ꢀFe1ꢀO6

95.92(14)

163.85(14)

92.88(13

O3ꢀZn1ꢀN3 169.83(10) O1ꢀCo1ꢀO3 92.89(8)

O3ꢀZn1ꢀN4 89.38(12)

O3ꢀZn1ꢀO1 92.83(10)

O3ꢀZn1ꢀO8 89.86(11)

O1ꢀCo1ꢀO5 92.51(8)

O1ꢀCo1ꢀO8 177.31(9)

O3ꢀCo1ꢀN3 170.77(9)

O1ꢀNi1ꢀO8 177.56(10) O3ꢀZn1ꢀN4 177.41(9)

O3ꢀNi1ꢀN3 171.83(10) O5ꢀZn1ꢀN3 167.58(8)

156.95(15)

91.68(15)

76.33(14)

86.23(13)

88.89(13)

83.22(12)

87.60(14)

162.54(13)

75.69(13)

98.72(14)

157.99(12)

83.57(13)

168.15(11)

93.57(10)

99.92(9)

O5ꢀZn1ꢀN3 102.87(11) O3ꢀCo1ꢀN4 90.64(10)

O5ꢀZn1ꢀN4 176.48(12) O3ꢀCo1ꢀO5 85.75(10)

O3ꢀNi1ꢀO1 93.03(9)

O3ꢀNi1ꢀO8 88.41(9)

O5ꢀNi1ꢀN3 101.07(10) O5ꢀZn1ꢀO3 88.72(8)

O5ꢀZn1ꢀN4 88.75(9)

O5ꢀZn1ꢀO1 87.76(7)

O5ꢀZn1ꢀO1 92.15(10)

O5ꢀZn1ꢀO3 87.23(12)

O5ꢀZn1ꢀO8 91.17(11)

O8ꢀZn1ꢀN3 88.79(11)

O8ꢀZn1ꢀN4 89.75(11)

O5ꢀCo1ꢀN3 103.20(9)

O5ꢀCo1ꢀN4 176.37(10) O5ꢀNi1ꢀN4 176.33(11) O7ꢀZn1ꢀN3 85.17(8)

O8ꢀCo1ꢀN3 88.92(9)

O8ꢀCo1ꢀN4 89.93(9)

O8ꢀCo1ꢀO3 88.86(8)

O5ꢀNi1ꢀO1 93.02(9)

O5ꢀNi1ꢀO3 86.45(11)

O5ꢀNi1ꢀO8 89.03(10)

O8ꢀNi1ꢀN3 88.55(10)

N2ꢀFe1ꢀN1 78.32(10)

O1ꢀFe1ꢀN1 87.94(9)

O1ꢀFe1ꢀN2 84.78(9)

O1ꢀFe1ꢀO2 93.38(9)

O1ꢀFe1ꢀO6 164.44(10) O1ꢀFe1ꢀO4

O1ꢀFe1ꢀO6

O2ꢀFe1ꢀN2 171.05(11) O2ꢀFe1ꢀN1

O2ꢀFe1ꢀO6 90.19(10) O2ꢀFe1ꢀN2

O4ꢀFe1ꢀN1 167.87(10) O2ꢀFe1ꢀO4

O4ꢀFe1ꢀN2 91.73(11)

O4ꢀFe1ꢀO1 98.14(10)

O4ꢀFe1ꢀO2 97.20(11)

O4ꢀFe1ꢀO6 96.43(10)

O6ꢀFe1ꢀN1 76.75(10)

O6ꢀFe1ꢀN2 89.36(10)

O7ꢀZn1ꢀN4 93.43(8)

O7ꢀZn1ꢀO1 171.58(8)

O7ꢀZn1ꢀO3 87.54(8)

O7ꢀZn1ꢀO5 100.13(8)

O8ꢀZn1ꢀO1 175.82(10) O8ꢀCo1ꢀO5 89.66(9)

N2ꢀFe1ꢀN1

O1ꢀFe1ꢀN1

O1ꢀFe1ꢀN2

O1ꢀFe1ꢀO2

O1ꢀFe1ꢀO6

O2ꢀFe1ꢀN1

O2ꢀFe1ꢀN2

O2ꢀFe1ꢀO6

O4ꢀFe1ꢀN1

O4ꢀFe1ꢀN2

O4ꢀFe1ꢀO1

O4ꢀFe1ꢀO2

O4ꢀFe1ꢀO6

O6ꢀFe1ꢀN1

O6ꢀFe1ꢀN2

78.09(11)

87.90(10)

84.81(10)

94.22(10)

N2ꢀFe1ꢀN1

O1ꢀFe1ꢀN1

O1ꢀFe1ꢀN2

O1ꢀFe1ꢀO6

78.07(9)

87.94(8)

84.63(8)

164.36(9)

92.42(9)

N2ꢀFe1ꢀN1

O1ꢀFe1ꢀN1

O1ꢀFe1ꢀN2

O1ꢀFe1ꢀO2

74.59(8)

90.86(8)

165.45(9)

103.06(8)

93.85(7)

88.32(7)

165.12(8)

91.41(8)

92.91(8)

95.07(8)

91.37(8)

86.86(8)

171.04(8)

79.89(8)

88.87(8)

164.32(11) O2ꢀFe1ꢀN1

92.90(11)

O2ꢀFe1ꢀN2

170.97(12) O2ꢀFe1ꢀO1

89.89(11)

O2ꢀFe1ꢀO6

167.80(10) O4ꢀFe1ꢀN1

91.73(11)

97.96(11)

97.30(11)

96.51(10)

76.78(10)

88.76(10)

170.42(10) O2ꢀFe1ꢀN1 92.88(10)

103.10(9)

100.18(9)

93.94(8)

89.90(9)

167.46(9)

91.63(9)

98.30(9)

97.95(9)

96.17(9)

76.74(9)

89.06(9)

O4ꢀFe1ꢀN2

O4ꢀFe1ꢀO1

O4ꢀFe1ꢀO2

O4ꢀFe1ꢀO6

O6ꢀFe1ꢀN1

O6ꢀFe1ꢀN2

O2ꢀFe1ꢀO6

O4ꢀFe1ꢀN1

O4ꢀFe1ꢀN2

O4ꢀFe1ꢀO6

O6ꢀFe1ꢀN1

O6ꢀFe1ꢀN2

^aDetermined using DIAMOND v. 3.1.

ascribed to the above-mentioned diﬀerences in bond lengths to

the phenolic oxygen. These intermetal distances are slightly

longer, but comparable to the Fe(III)ꢀM(II) distances crystal-

lographically determined for heterodinuclear plant PAP enzymes

(3.1ꢀ3.3 Å).^27ꢀ29These results can be compared to other

heterodinuclear complexes of similar ligands with one phenolate

and two acetate bridges that have been studied by other groups. The

metalꢀmetal distances in 2ꢀ4 are slightly shorter than in com-

plexes containing terminal phenolate groups coordinated to Fe(III)

(the BPBPMP ligand; MꢀM = 3.470(1)ꢀ3.510(9) Å)^56ꢀ60

3871

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 5. (a) ORTEP representation of the crystal structure of one monomer of [Fe(III)Zn(II)(IPCPMP)(mpdp)][PF₆] (6). (b) A Mercury plot of

two repeating units in the polymeric network of 6 with only metals and atoms that are bridging and part of the polymeric backbone. All solvent and

counterions have been omitted for clarity, and ellipsoids are drawn at 50% probability.

but similar or slightly longer than complexes with only pyridyl

donor moieties (and the two tertiary amines) in the dinucleating

ligand (3.378(8)ꢀ3.437(1) Å).^53,55,61Further comparison to

heterodinuclear complexes of BPBPMP^56ꢀ60show that the

terminal phenolate-Fe(III) bond in the BPBPMP complexes

are approximately 0.1 Å shorter than the corresponding carbox-

ylate-Fe(III) bonds in the IPCPMP complexes (O6ꢀFe1 in 2, 3,

and 4, cf. Table 2), that occupy the same position trans to the

bridging phenolate. The bond between Fe(III) and the bridging

phenolate is accordingly slightly shorter (0.05ꢀ0.03 Å) in 2, 3,

and 4 as compared to the BPBPMP complexes. It is obvious,

however, that the nature of the exogenous bridge(s) strongly

inﬂuences the intermetal distance between the metals when

comparisons are made to complexes with similar phenolate-

bridging ligands as used here but diﬀerent additional bridging

moieties, for example, a hydroxide-bridged complex (Fe(III)ꢀ

Zn(II) 3.040(1) Å),³⁶diphenylphosphate-bridged complexes

(Fe(III)ꢀZn(II) 3.695(1) to 3.7402(7) Å),^44,46,62a molyb-

date-bridged complex (Fe(III)ꢀZn(II) 3.819(4) Å)⁴³or com-

plexes only containing the phenolate bridge and no exogenous

bridge (3.5821(12);⁶³3.828(1)⁶⁴Å).

It is clear from these structural comparisons that the mono-

nuclear “synthon” 1 functions as a scaﬀold, inside which a divalent

ion and two bridging carboxylates are easily ﬁtted.⁴⁹In 2ꢀ4, the

sixth coordination site of the divalent metal is ﬁlled by a methanol

ligand originating from the solvent. However, in the case of the

formation of a dinuclear Fe(III)Zn(II) complex by reaction of 1

with ZnCl₂and meta-phenylene dipropionate (mpdp, Figure 2d), the

crystal structure of the resultant complex [Fe(III)Zn(II)(IPCPMP)-

(mpdp)][PF₆] (6) reveals a heterodinuclear solid state polymer as

shown in Figure 5. Crystallographic data are summarized in

Table 1, and relevant bond distances and angles are listed in

Table 2. The repeating dinuclear units are connected via the

carboxylate functionality of the IPCPMP ligandthat iscoordinated

totheFe(III) and that bridgesin asyn,anti-μ_1,3mannertotheZnat

the “open” site that is occupied by solvent in 2ꢀ4. The bond

distances are only slightly changed as compared to 2, and the

metalꢀmetal distance within a repeating dinuclear unit is

Fe1ꢀZn1 3.4695(5) Å and between two repeating units

Fe1ꢀZn#1 5.5799(5) Å. In fast atom bombardment (FAB) mass

spectrometry of this complex only the dinuclear complex

[Fe(III)Zn(II)(IPCPMP)(mpdp)]^þis observed.

The above-mentioned divalent metals readily accommodate

(distorted) octahedral geometries, but Cu(II) is well-known to

prefer strong tetragonal distortion from an octahedral geometry

or square planar coordination. Accordingly, the reaction of the

mononuclear “synthon” 1 with Cu(II) acetate monohydrate,

using method A (vide supra and Experimental Section), gave

[Fe(III)Cu(II)(IPCPMP)(OAc)}₂(μ-O)][PF₆] (7). The crys-

tal structure of 7 was determined, and the molecular structure of

the complex is shown in Figure 6; crystallographic data are

summarized in Table 1, and relevant bond distances and angles

are listed in Table 2. Complex 7 is tetranuclear, consisting of two

dinuclear moieties that contains two μ-oxido bridged octahedral

Fe(III) ions and two Cu(II) ions in a distorted square planar

geometry. The two heterodinuclear parts are related by inversion

symmetry with the center of inversion located on the bridging

oxido group. Coordination of the copper by only four donors

nominally leaves one vacant coordination site on the iron ion

because of the absence of a (second) carboxylate bridge in the

dinuclear moiety; this coordination site is ﬁlled by the μ-oxido

ligand that most likely is formed from residual water in the

solvent (methanol). Each Fe(III) ion is thus in an octahedral

N₂O₄environment bridged to a Cu(II) ion through the pheno-

late of the ligand and one syn,syn-μ_1,3-acetate bridge. With only

four donors (N₂O₂), the bonds to the Cu(II) ions are shorter by

more than 0.1 Å (average CuꢀX = 1.953 Å) compared to those

for the divalent metals in 2ꢀ4. In contrast, all bonds to the

Fe(III) are longer (average FeꢀX = 2.054 Å) than in 2ꢀ4 with

the exception of the FeꢀO_oxidobond (1.7934(6) Å). The

Fe(III)ꢀOꢀFe(III) angle is 180ꢀ, and the distance between

them is 3.5868(8) Å. The Fe(III)ꢀCu(II) distance (3.2981(9) Å)

is shorter than the Fe(III)ꢀM(II) distance in either 2, 3, or 4

mainly because of the very short bond between the Cu(II) and

the phenolic oxygen, since the corresponding Fe(III)ꢀO_phenolate

distance (2.047(3) Å) is actually longer in 7 than in 2ꢀ4 and

analogous complexes of the BPBPMP ligand (vide supra).

Similar FeꢀOꢀFe-bridged tetranuclear (dimer of dimer) struc-

tures, as in 7 (FeCu), has been observed previously with

heterodinuclear complexes of macrocyclic bis-phenolato-brid-

ging ligands (Figure 2c)^65,66and oxido-bridged dinuclear Fe(III)

complexes are quite common.⁶⁷Typical FeꢀO_oxidoand FeꢀFe

distances are 1.734(1)ꢀ1.7816(7) Å and 3.468ꢀ3.563(1) Å,

respectively, for monobridged μ-oxido complexes with an

3872

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 6. ORTEP representations of the crystal structure of [Fe(III)Cu(II)(IPCPMP)(OAc)}₂(μ-O)][PF₆] (7). All solvent molecules and

counterions have been omitted for clarity, and the ellipsoids are drawn at 50% probability. In (a) the complete tetranuclear structure is shown, and

in (b) only the dinuclear FeCu part of the structure is displayed.

FeꢀOꢀFe angle of 180ꢀ, indicating that the corresponding

distances in complex 7 are relatively long. Biswas et al.⁶⁵and

Dutta et al.,⁶⁶who have isolated similar μ-oxido bridged dimers,

observed a linear bridge for a Fe(III)ꢀZn(II) complex and bent

bridges for a Fe(III)ꢀCo(III) complex with FeꢀFe distances of

3.561(1), 3.564(1), and 3.549(1) Å for Zn(II)Fe(III)ꢀOꢀFe-

(III)Zn(II), Co(III)Fe(III)ꢀOꢀFe(III)Co(III) and a Fe-

(III)ꢀOꢀFe(III) dimers, respectively.

carboxylate. In 1, an octahedral rearrangement by a Bailar or

RayꢀDutt twist is in principle possible, while in the dinuclear

assemblies this would probably require that the same twist occurs

on both metals at the same time since the plane of the phenolate

aromatic ring generally prefers being signiﬁcantly tilted relative

to the plane formed by the two metals and the phenolate oxygen

O1 (2: 51.89ꢀ, 3: 52.18ꢀ, 4: 51.72ꢀ, 5: 50.83ꢀ 6: 52.53ꢀ).

Although this type of tilt is rarely discussed, it appears to be

present for most complexes with this type of 1,3-methylenea-

mine phenolate ligands.^{53,55,68ꢀ70}The tilt angle is very small

(17.40ꢀ) for the hydroxido bridged Fe(III)Zn(II) complex

reported by Neves et al. which has an unusually short distance

between the metal ions.³⁶The tilt of the phenolate relative to the

intermetal axis induces skew chirality, and there are four stereo-

genic centers in the general heterodinuclear molecule (the two

metals and the two tertiary nitrogens). A number of diastereo-

mers are thus in principle possible, and the complexes crystallize

as racemates of two enantiomers with S,S- and R,R-conﬁgura-

tions at the tertiary amines (cf. Figure 4).

IR Spectroscopy. The endogenous carboxylate donor of the

ligand IPCPMP makes infrared spectroscopy a useful tool to

study the complexation behavior of the ligand in both solid state

and solution. Especially the difference in energy for the antisym-

metric and the symmetric stretch (Δν = ν_antiꢀ ν_sym) of metal

carboxylates has been used to evaluate whether they are coordi-

nated in a bridging or monodentate mode (>200 cm^ꢀ1mono-

dentate; e180 cm^ꢀ1bridging; , 180 cm^ꢀ1chelating).⁷¹This

can be used to assign the peaks in the FTIR spectra of complexes

2 (FeZn), 3 (FeCo), 4 (FeNi), 5 (FeMn), and 7 (FeCu) in KBr

(Figure 7a) and to relate these to structural features in both solid

state and solution.

Finally, although not all structures discussed above are strictly

comparable, it should be noted that the pyridyl and carboxylate

donors of IPCPMP occupy diﬀerent positions in relation to the

other coordinating groups in diﬀerent complexes. In mono-

nuclear 1 and in 6 (FeZn-mpdp), the pyridyl group on the

carboxylate side of IPCPMP (binding to Fe(III)) is trans to the

bridging phenolate group while in 2 (FeZn), 3 (FeCo), 4 (FeNi),

and 7 (FeCu) the carboxylate on this side binds trans to the

phenolate, and the pyridyl is trans to one of the bridging acetates.

In complex 7, the pyridyl on the 2-propyl side (binding to

Cu(II)) is trans to the phenolate while for 2ꢀ4 and 6 it is trans to

one of the bridging acetates. Electronic eﬀects appear to be of

minor importance for the positions of the donor groups. In 1, the

cis position of the carboxylate in relation to the central phenolate

moiety of the ligand may be due to favorable hydrogen bonding

between the carboxylate (O1) and the nitrogen (N1) of the

second, protonated, pyridyl moiety of the ligand (cf. Figure 3). In

7, restraints caused by the square planar geometry at the Cu(II)

force the phenolate oxygen, the tertiary nitrogen, and the pyridyl

nitrogen to be in one plane, thus positioning the pyridyl trans to

the phenolate. It is rational that the bridging oxido ligand is trans

to the weak FeꢀN bond of the tertiary nitrogen. In 6, packing

eﬀects in the solid state may favor one conformer in the extended

polymeric relative to the other.

The FTIR spectra for 2ꢀ4 are very similar, as expected

because of their almost identical structures in the solid state.

The striking similarity of the IR spectrum of 5 to those of 2ꢀ4

strongly indicates that the general structure of 5 is identical to

those of the other three complexes. The IR resonances for 2ꢀ5

can be assigned by comparing their relationships in terms of

energy (wavenumber) and intensities. For 2ꢀ5 the symmetric

(ν_sym) and antisymmetric (ν_anti) vibrations of the bridging

We have not been able to determine whether diﬀerent

conformers exist in solution and, if so, how or if they interconvert

between each other. If the solid state structure of 1 (Figure 3) is

the only form found in solution, then a ligand rearrangement

is needed to form 2ꢀ5. This would probably require the

dissociation of the phenolate and subsequent formation of a

new phenolate-Fe(III) bond in trans position relative to the

3873

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 7. Fourier transform IR spectra of complexes [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2), Co (3), Ni (4), Mn (5)),

[Fe(III)Cu(II)(IPCPMP)(OAc)}₂(μ-O)][PF₆] (7), and [{Zn₂(IPCPMP)(OAc)}₂][PF₆] (Zn4) in KBr (a). Solution FTIR of 2, 3, 4 and Zn4 40 mM

in acetonitrile (b) and 2, 3, and 4 40 mM in H₂O/MeCN 1:1 v/v (c). The intensities have been adjusted using the strong absorption at 843 cm^ꢀ1for

PF₆^ꢀas internal standard. The spectra are numbered after their corresponding complexes. In the middle and bottom graphs the spectrum of 2 in KBr has

been added as a reference.

acetates are assigned to the bands ν₄and ν₂at 1419 and

1608 cm^ꢀ1, respectively, based on their intensities and similarity

to literature data.⁷¹The diﬀerence between ν_antiand ν_sym(Δν =

ν_antiꢀ ν_sym) is 189 cm^ꢀ1, which is close to the ionic value of

acetates, characteristic for bridging carboxylate ligands.⁷¹Com-

parison of these symmetric (∼1400 cm^ꢀ1, ν₄) and antisym-

metric stretches (∼1600 cm^ꢀ1, ν₂), of the bridging carboxylates

for 2 (FeZn), 3 (FeCo), 4 (FeNi), and 5 (FeMn) reveal only

minor diﬀerences despite their bridging between diﬀerent metals.

The smaller and somewhat broadened resonance at 1650 cm^ꢀ1

(ν₁) probably originates from the antisymmetric stretch of the

terminal and monodentate carboxylate of the IPCPMP ligand. The

intensity of this resonance relative to that at 1608 cm^ꢀ1(ν₂) is in

agreement with the ratio (2:1) of bridging versus terminal carbox-

ylates in the complex. This assignment is indirectly conﬁrmed by

comparison to the IR spectrum of the tetranuclear zinc complex

[{Zn₂(IPCPMP)(OAc)}₂][PF₆]⁴⁷which does not have any term-

inal monodentate carboxylate, and does not show any similar IR

resonance, except in solution where dissociation is expected to

occur.⁷²The symmetric band of this monodentate carboxylate can

also be assigned to 1342 cm^ꢀ1(ν₅) by comparing to

[{Zn₂(IPCPMP)(OAc)}₂][PF₆], which does not have any intense

peak in this region. This assignement yields Δν = 313 cm^ꢀ1, in line

with what is expected for a monodentate carboxylate.

The IR spectrum of 7 (FeCu) is quite diﬀerent from 2ꢀ4 but

can be similarly assigned by comparison to the crystal structure. It

displays somewhat broadened peaks at 1660 cm^ꢀ1(ν₁) and

1321 cm^ꢀ1(ν₆) with Δν = 339 cm^ꢀ1, corresponding well to

monodentate coordinated carboxylates and peaks at 1556 (right

of ν₂) and 1444 cm^ꢀ1(ν₄) with Δν = 112 cm^ꢀ1, which is within

limits for bridging acetates. The relative intensities of the

resonances are in agreement with the structure of 7. No stretches

corresponding to μ-oxido-diFe(III) vibrations⁶⁷have been ob-

served. They are likely concealed by the very strong absorptions

of the hexaﬂuorophosphate anion (ν₈and others) in this region.

When the complexes 2 (FeZn), 3 (FeCo), and 4 (FeNi) are

dissolved in acetonitrile, the peak at 1650 cm^ꢀ1(ν₁) for all

complexes in KBr sharpens up and shifts to higher wavenumbers

(1678, 1678, and 1682 cm^ꢀ1for 2, 3, 4, respectively, ν₁

Figure 7b). The fact that this stretch is virtually the same for

complexes 2ꢀ4 is an indication that the stretch originates from

the terminal carboxylate bound to the Fe(III) ion. This also

suggests that metal migration in the complex from one side of the

phenolate to another does not occur in acetonitrile solution

within the detection limits of this FTIR experiment. The broad

appearance of this peak in the solid state is probably due to

inhomogeneous broadening, and the various weak interactions

that cause this broadening are removed in solution.

Because of the low solubility of the complexes in water, all

reactivity studies were performed in a water/acetonitrile solution

(1:1 v/v, vide infra). Therefore, this solvent mixture was also

used in the FTIR experiments. The analysis is complicated by the

strong hydrogen-bonding capabilities of water, which can inter-

act strongly with heteroatoms such as the oxygens of the

carboxylate groups. The major change in the IR spectra when

the complexes are dissolved in the above-mentioned solvent

mixture (Figure 7c) is that the ν_antipeak at 1678 cm^ꢀ1(ν₁)

completely disappears in the case of 2 and for 3 and 4 it broadens

and shifts to lower wavenumbers, almost disappearing behind the

resonance at 1609 cm^ꢀ1. This could be explained by interaction

of the terminal carboxylate of the ligand with water through

hydrogen bonding, a form of pseudobridging. In previous

studies, absences of large Δν in solution in cases where mono-

dentate acetate coordination to metals have been determined by

crystallography have been explained by pseudobridging to hy-

drogens in the structure.⁷¹A similar explanation for complexes in

protic solvents seems viable. Another possible explanation is that

the endogenous monodentate carboxylate of IPCPMP is brid-

ging to other metals also in solution in the case of complex 2; this

could then be the reason for the peak at 1678 cm^ꢀ1disappearing

completely for 2 but not for 3 and 4. The spectra of 3 and 4 in

water/acetonitrile are, however, still similar to their spectra in the

solid state (KBr) and acetonitrile solution, except for the two

peaks at about 1678 and 1313 cm^ꢀ1(vide supra). This might

suggest that the structures of these complexes remain intact in

solution at close to neutral pH.

The peaks at 1026 cm^ꢀ1(ν₇) and 1478 cm^ꢀ1(ν₃) are assigned

to rocking vibration of the acetate methyl groups⁷³and aromatic

3874

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 8. Dependence on pH of the spectra of [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2), Co (3), Ni (4), Mn (5). The arrows

indicate the decrease and increase in intensity of speciﬁc bands when going from pH 5 to 10.5 in steps of 0.5.

CdC stretches, respectively. These bands, however, do not

change for any of the complexes or conditions discussed above.

Mass Spectrometry. To further elucidate the nature of the

heterodinuclear complexes in solution, mass spectrometric mea-

surements were performed. Electrospray ionization mass spectra

(ESI-MS) of 2 (FeZn), 3 (FeCo), 4 (FeNi), 5 (FeMn), and 7

(FeCu) in acetonitrile solutions display peaks with m/z values

and isotope patterns corresponding to the formulation

[Fe(III)M(II)(IPCPMP)(OAc)₂]^þ(M(II) = Zn, Co, Ni, and

Mn, respectively). For complex 7, a cluster of peaks at m/z 632

corresponding to the dicationic [{Fe(III)Cu(II)(IPCPMP)-

(OAc)}₂(O)]^2þ(mass = 1264 a.m.u.) is observed although the

isotope pattern is distorted due the use of unit resolution. A very

weak peak corresponding to [{Fe(III)Cu(II)(IPCPMP)-

(OAc)}₂(O)(PF₆)]^þwith m/z of 1409 and correct isotope

pattern could, however, be detected.

Upon dissolving these complexes in water/acetonitrile 1:1 (v/

v) the mass-spectra change considerably. For 2 and 4, peaks

corresponding to [Fe(III)M(II)(IPCPMP)(OAc)(OH)]^þ(m/

z = 642 and 636 for Zn(II) and Ni(II), respectively) and

[Fe(III)M(II)(IPCPMP)(O)]^þ(m/z = 582 and 576 for Zn(II)

and Ni(II), respectively) occur. Both these species are also

observed for 3 as well as a peak corresponding to

[Fe(III)Co(II)(IPCPMP)(OH)₂]^þ(m/z = 595). For all three

complexes, the spectra also contain a peak for the mononuclear

Fe(III) complex [Fe(III)(IPCPMP)]^þ(m/z = 502). However,

no peaks corresponding to either homodinuclear complexes or

mononuclear complexes of the divalent metal ion are observed,

indicating that the Fe(III) is tightly bound, probably on

the carboxylate side, while the divalent metal ion is more

labile under these conditions. Indeed, only the mononuclear

[Fe(III)(IPCPMP)]^þcomplex is observed for complex 5 in

water/acetonitrile solution (1:1 v/v). The dissociation of man-

ganese from heterodinuclear 5 is in line with the relatively weak

binding generally observed for (high spin) Mn(II) that is due to

its low charge-to-radius ratio and lack of ligand ﬁeld stabilization

energy.⁷⁴It should be noted that the pH of these water/

acetonitrile solutions was not controlled but are assumed to be

within the 5ꢀ7 pH range. Adjusting the pH of the solution of 5 to

∼pH 9 using a dilute NaOH solution yields a range of peaks but

several can be related to Fe(III)Mn(II) species of the ligand (e.g.,

[Fe(III)Mn(II)(IPCPMP)(OAc)(OH)]^þm/z = 633, [Fe(III)-

Mn(II)(IPCPMP)(OH)₂]^þm/z = 591, and [Fe(III)Mn(II)-

(IPCPMP)(O)]^þm/z = 573. Because of experimental diﬃcul-

ties, the other complexes could not be studied under the same

conditions.

In FABþ mass spectra of 2 left in H₂O/acetonitrile solution over

several hours, peaks corresponding to dinuclear Zn complexes can

be observed. The relative intensities of these peaks increase upon

treatment with base (triethyl amine). This indicates that decom-

position of the heterodinuclear complexes over extended periods of

time at high pH cannot be excluded. In contrast to complexes 2ꢀ4,

only a minute peak for the mononuclear complex is observed (<3%)

for 7 (FeCu) in water/acetonitrile solution. Three major peaks with

m/z and isotope patterns corresponding to [Fe(III)Cu-

(II)(IPCPMP)(OH)₂]^þ(m/z = 599), [Fe(III)Cu(II)(IPCPMP)-

(O)]þ (m/z = 581), and [{Fe(III)Cu(II)(IPCPMP)(OH)}₂-

(O)]^2þ(m/z = 1180/2 = 590) are shown.

For all acetate-bridged complexes at least partial substitution

of these exogenous carboxylate bridges is to be expected in

aqueous solution because of their well documented lability.⁷⁵It

should be noted, however, that several of the detected species are

degradation products due to the ionization process. For example,

the species Fe(III)M(II)(IPCPMP)(O)]^þobserved in the spec-

tra of 2, 3, and 4 in aqueous solution could be generated by loss of

a proton from a coordinated hydroxide or from Fe(III)M(II)-

(IPCPMP)(OAc)(OH)]^þby loss of acetic acid.

Electronic Spectroscopy. The strong absorption of the

phenolate-to-Fe(III) charge transfer transition is a characteristic

feature of both the purple acid phosphatases and the complexes

in this study. It is a valuable spectroscopic tool to probe changes

that occur especially in the Fe(III) coordination sphere. The

3875

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 9. UVꢀvis spectrum of (a) [Fe(III)Zn(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (2) at pH 10, 0.25 mM in acetonitrile/water 1:1, and

(b) [Fe(III)Cu(II)(IPCPMP)(OAc)}₂(μ-O)][PF₆] (7), 0.1 mM in acetonitrile. The inset displays the derivative of the absorbance vs the wavelength to

more clearly display the shoulders present in the spectra.

UVꢀvis spectra of 2 (FeZn), 3 (FeCo), 4 (FeNi), and 5 (FeMn)

were collected at several different pHs using the buffered

solutions for the reactivity studies with HPNP (vide infra) before

adding the substrate. All spectra (Figure 8) change smoothly

displaying isosbestic points, but there are differences between the

complexes. Complex 2 (FeZn) appears to be completely con-

verted to a new species at pH 9.5. This species displays absorp-

tions characteristic of oxido-to-Fe(III) dimer charge transfer

interactions⁶⁷at 336 nm (3a in Figure 9; 2990 M^ꢀ1cm^ꢀ1) and

381 nm (3b; 1650 M^ꢀ1cm^ꢀ1) as seen in Figure 9a. In principle,

three transitions are possible: p_zfFe d_z₂, which usually is

concealed by other strong absorptions below 300 nm and p_x

and p_yf Fe d_xzand d_yz. For a linear bridge between the two

Fe(III) ions, the p_xand p_ytransitions are degenerate and only

one band is observed between 300 and 500 nm. When the

FeꢀOꢀFe bridge is bent, the two transitions split, and two bands

are observed.^67,76This could indicate that the species formed

from 2 at high pH is a μ-oxido diFe(III) species with a bent oxido

bridge. The shoulder at 381 nm could in principle also originate

from the phenolate-to-Fe(III) CT transition but the very weak

shoulder at about 460 nm (1a, Figure 9a), more clearly detected

in the derivative spectrum in the inset of Figure 9a, is more likely

from this CT transition. The spectrum of 2 at pH 10 is compared

to that of 7 (FeCu) in acetonitrile in Figure 9b. For 7 the single

shoulder at 336 nm (3b) (10 000 M^ꢀ1cm^ꢀ1) fits well with a

linearly bridged μ-oxido-diFe(III) center as observed in the

crystal structure (Figure 6). For 2 a shoulder at the same

wavelength (336 nm, (3a)) is observed indicating that the

structure of 2 at high pH might be similar to that of 7 although

for the latter complex the bridge in the solid state is perfectly

linear. In a separate study, we have prepared the oxido- and

pivalate-bridged diFe(III) complex [{Fe(IPCPMP)}₂(μ-O)-

(μ-1,3-pivalate)]ClO₄.⁷⁷Its UVꢀvis spectra is not very similar

to that detected for 2 at high pH, and it is therefore not likely that

the high pH form of 2 is such a diiron species. From the fit of a

sigmoidal Boltzmann function to the change of a specific

absorption versus pH (Supporting Information, Figure S1), the

pH at 50% conversion can be extracted, and this corresponds to

the pK_aof the functional group that undergoes the assumed

protolysis. For complex 2, two different pK_a's can be estimated,

6.8 for the disappearing peak at 518 nm, and 8.3 for the two peaks

forming at 336 and 383 nm. However, for a direct transformation

of one species to another, and hence isosbestic points in general,

the estimated pH at 50% conversion should be the same for the

peaks that disappear in the starting materials and the peaks that

appear in the product at high pH. For 2 (FeZn) this is clearly not

the case. If, for two consecutive deprotonations, an intermediate

that has an absorption maximum coinciding with the spectral

traces of the starting and final species exists, then isosbestic

points may still be observed. This behavior is partially observed

for complex 4 (FeNi) and 5 (FeMn) (vide infra).

The spectra of complex 3 (FeCo, Figure 8b) show similar pH

dependence as 2. At high pH, two weak shoulders are visible in

the same region as the oxido-to-Fe(III) CT bands for 2, but 3 is

not completely converted even at pH 10.5. The band at 510 nm,

observed at low pH, has however disappeared at pH 9, and a

sigmoidal ﬁt to its change with increasing pH gives a pK_aof 7.1.

Complex 4 (FeNi) and 5 (FeMn) behave similarly to each

other (Figure 8c and d) with a shift of the phenolate to Fe(III)

CT bands from 517 and 505 nm to 446 and 440 nm, respectively,

when going from pH 5 to 10, with a slight decrease in intensities

but without the formation of oxido to Fe(III) dimer CT bands

that were observed for 2 and 3. The pK_a's for these processes,

derived from sigmoidal ﬁts to the pH dependence of the

absorbances at 517 and 505 nm, are diﬀerent, however (6.9

and 8.2, respectively). The maximum absorbance for the band

forming at high pH, for both 4 and 5, is close to the isosbestic

point, and hence the absorbance changes very little at this

wavelength and no sigmoidal ﬁt to its pH dependence is possible.

This may be related to the presence of an isosbestic point for

complex 2 (FeZn), if a similar complex with a similar absorbance

is formed as an intermediate between two consecutive deproto-

nation reactions as proposed above for 2. The nature of the

species that is formed at high pH for complexes 4 and 5, and

possibly as an intermediate in the case of 2, is at the moment not

known. Its formation may involve deprotonation of a bridging or

terminal water molecule, and the fact that the observed pK_afor

the process is diﬀerent for 4 and 5 suggests that this water

interacts with the divalent metal.

In summary, on the basis of the above-mentioned spectro-

scopic observations, we tentatively propose that complex 2

(FeZn) is converted into a dimer of heterodinuclear dimers,

linked via an FeꢀOꢀFe bond, at pH > 9. A similar dimer of

dimers may be formed at high pH in the case of complex 3

(FeCo) but appears not to form for complexes 4 (FeNi) and 5

(FeMn). The lack of clear spectroscopic signatures for the two

latter complexes prevents us from assigning any deﬁnite structure

for the species formed at high pH in their cases.

3876

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 10. M€ossbauer spectra at 80 K of [Fe(III)(H₂IPCPMP)(Cl₂)][PF₆]) (1) (a), [Fe(III)Co(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (3) (b), and

[Fe(III)Mn(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (5) (c) in the solid state.

Table 3. Magnetic Data for [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2), Co (3), Ni (4), and Mn (5)) and

Mo€ssbauer Parameters for Complexes 1ꢀ5 at 80 K

susceptibility data

M€ossbauer data

complex

J

₁₂[cm^ꢀ1

]

g_Fe(III)

g_M(II)

PI [%] (S = 5/2)

TIP [10^ꢀ6cm³/mol]

isomeric shift δ (mm/s)

quad. split ΔE_Q(mm/s)

1

2

3

4

5

0.48

0.53

0.47

0.50

0.47

0.56

1.94

2.09

1.96

1.99

0.0

0.4

0.1

0.3

558

111

152

114

ꢀ5.4

ꢀ11.2

ꢀ9.6

1.92

2.19

1.99

0.40

0.33

€

Mossbauer Spectroscopy. To probe the electronic environ-

ment of the Fe in various heterodinuclear complexes, and to

verify the oxidation and spin states of iron, the M€ossbauer spectra

of 1 (Fe-mono), 2 (FeZn), 3 (FeCo), 4 (FeNi), and 5 (FeMn)

were recorded. The spectra of 1, 3, and 5 (Figure 10 a, b and c,

respectively) show isomer shifts and quadrupole splittings typical

of a high spin Fe(III) ion (Table 3). The isomer shifts are very

similar, and the smaller quadrupole splittings for 4 and 5 as

compared to 1 are likely due to the more symmetrical environ-

ment with only nitrogen and oxygen donors at similar distances

around the Fe(III) for the former two. The Fe(III)ꢀCl bonds are

significantly (∼0.1 Å) longer than the other bonds to the metal in

1, yielding a less symmetrical environment. The broad appear-

ance and absence of quadrupole splitting for the resonances in

the spectra of 2 and 4 (Figure 11a and b, respectively) indicate

relaxation effects because the electronic spin relaxation is on the

same time scale as the Larmor frequency of the nuclei.⁷⁸

Spinꢀlattice relaxation involves a transfer of Zeeman energy of

the spin system to phonon modes of the lattice via spinꢀorbit

coupling and is temperature dependent. By this effect the spectra

get broader at lower temperature, and this is indeed observed for

2 when comparing spectra at 80 and 7 K (Supporting Informa-

tion, Figure S2 a and b, respectively) and for 4 when comparing

the spectra at 295 and 80 K (Supporting Information, Figure S3 a

and b, respectively). The presence of relaxation effects for 2

(FeZn) and 4 (FeNi) but not for 3 (FeCo) is difficult to explain

since the structures and crystallographic symmetry for all three

are very similar, assuming that crystallographic homogeneity

prevails in the bulk solid.

Magnetic Properties. Magnetic susceptibility measurements

were carried out for complexes 2 (FeZn), 3 (FeCo), 4 (FeNi),

and 5 (FeMn) at magnetic fields of 5000 Oe in the temperature

range 295ꢀ2 K. The temperature dependence of the effective

magnetic moments (μ_eff/μ_B) are shown in Figure 12aꢀd. The

experimental data for the complexes were modeled by using a

fitting procedure to the appropriate HeisenbergꢀDiracꢀvan

Vleck (HDvV) spin Hamiltonian for isotropic exchange coupling

and Zeeman splitting, eq 1.⁷⁹

^

^ ^

^

H ¼ ꢀ 2JS₁S₂þ gμ_BðS₁þ S₂ÞB

ð1Þ

3

3877

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 11. M€ossbauer spectra at 80 K of [Fe(III)Zn(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (2) (a) and [Fe(III)Ni(II)(IPCPMP)(OAc)₂-

(CH₃OH)][PF₆] (4) (b) in the solid state.

Figure 12. Eﬀective magnetic moment versus temperature for [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)] [PF₆] (M = Zn 2 (a), Co 3 (b), Ni 4 (c),

and Mn 5 (d)). The solid lines are least-squares ﬁts to the theoretical expression.

Temperature-independent paramagnetism (TIP) and a para-

magnetic impurity (PI) with spin S = 5/2 and Curie behavior

FeCo complexes of a similar phenolate ligand with J₁₂= ꢀ6 and

ꢀ10 cm

^{ꢀ1 64}The possibility of a Co(III)Fe(II) complex can be

.

were included according to χ_calc= (1 ꢀ PI) χ þ PI χ_mono

þ

ruled out since for high spin Co(III) (S = 2) and high spin Fe(II)

(S = 2) antiferromagnetic coupling would yield an S = 0 ground

state. The remaining magnetic moment of 2.71 μ_Bobserved at

low temperature is in conﬂict with this unless about 30%

paramagnetic impurities are assumed. A low spin Co(III) would

not contribute to any magnetism and hence a simple Fe(II)

paramagnetic behavior would be observed. The M€ossbauer

spectra also clearly indicate a high spin Fe(III) in the complex

(vide supra). In the case of 5, the same dilemma can not be

resolved by magnetic susceptibility measurements, but the

M€ossbauer spectra clearly indicate a high spin Fe(III) center

supporting the formulation of 5 as high spin Fe(III) and Mn(II),

that from the ﬁtting yields an antiferromagnetic exchange

3

TIP.⁸⁰The parameters from the ﬁtting are presented in Table 3.

The data for 2 (FeZn) (Figure 12a) indicate a magnetically

isolated high spin Fe(III) center, as expected, with a magnetic

moment of 5.83 μ_Bat room temperature and a calculated g value

of 1.94. The data for 3 (FeCo), 4 (FeNi), and 7 (FeMn)

(Figure 12bꢀc) all display local magnetic ordering when the

temperature is decreased, indicating magnetic exchange coupling

between the two metal centers in the complexes, and the analysis

reveals weak antiferromagnetic coupling for all three. For 3, the

data correlate well with the formulation of the complex as

containing high spin Fe(III) and Co(II) ions with an exchange

coupling J₁₂= ꢀ5.4 cm^ꢀ1, comparable to previously reported

3878

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Scheme 1

interaction of J₁₂= ꢀ9.6 cm^ꢀ1. This correlates well with

previously determined values for FeMn complexes of the phe-

nolate-based ligands BPMP (ꢀ11.5 cm^ꢀ1),⁶⁹BIMP⁸¹

(ꢀ7.7 cm^ꢀ1), and BPBPMP⁵⁹(ꢀ6.8 cm^ꢀ1) (Schematic draw-

ings are shown in Figure 2). In the case of 4 the data could

successfully be ﬁtted to a high spin Fe(III) and high spin Ni(II)

with an exchange coupling of ꢀ11.2 cm^ꢀ1in line with other

Fe(III)Ni(II) complexes of the BPBPMP ligand, both with and

without exogenous carboxylate bridges (in the range ꢀ13.2 to

ꢀ13.7 cm^ꢀ1),⁸²and the BPMP ligand (ꢀ11.5 cm

)

^{ꢀ1 55}with two

propionate bridges, suggesting that the principal superexchange

pathway is transmitted over the phenolate bridge in all cases. On

the other hand, Fe(III)ꢀCo(II) and Fe(III)ꢀNi(II) complexes

of a bis-μ-phenolato macrocyclic ligand (Figure 2c) studied by

Dutta et al.⁶⁶have a weak ferromagnetic interaction (J₁₂= 4.2 and

1.7 cm^ꢀ1). The observed ferromagnetic coupling was rationa-

lized by a more acute Fe(III)ꢀOꢀM(II) angle (92.8ꢀ). The

results here are in line with this description since the Fe(III)ꢀO-

(phenolato)ꢀM(II) angles are all signiﬁcantly larger (>110ꢀ)

than 90ꢀ, as for the previously mentioned FeNi complexes of

BPBPMP⁸²and BPMP.⁵⁵

Transesterification of 2-Hydroxypropyl-para-nitrophenol

Phosphate (HPNP). To test if complexes 2 (FeZn), 3 (FeCo), 4

(FeNi), and 5 (FeMn) also are functional models, their activities

toward cleavage of two phosphoester substrates, 2-hydroxypro-

pyl-para-nitrophenyl phosphate (HPNP) and bis(2,4-dinitro-

phenyl)phosphate (BDNPP), were tested (Scheme 1).

First the pH dependence of their reactivities was investigated,

and, as seen in the graphs in Figure 13, all complexes enhance

HPNP transesteriﬁcation although there are considerable diﬀer-

ences in their reactivities. Highest activity is shown by complex 3

(FeCo), with a rate 200 times that of the noncatalyzed reaction

(v_i,cat/v_{i, uncat}) at pH 7 and 475 times at pH 8. Complex 4 (FeNi)

also shows activity at pH > 6 (a factor of 75 at pH 7 and 120 at pH

8 compared to the noncatalyzed reaction). A sigmoidal shape for

the rate versus pH (Figure 13a) is observed for 4 until approxi-

mately pH 9.5 where the noncatalyzed reaction becomes impor-

tant. A sigmoidal Boltzmann function ﬁtted for pH 6ꢀ9.5 gives a

pH of 8.3 at v_0max/2, which might correspond to a pre-equilibrium

deprotonation of the substrate, if the alcohol group is coordinated

to one of the metals, or a coordinated water molecule. Complex 5

also shows a sigmoidal pH dependence with a rate enhancement at

pH 7 and 8 by factors of 14 and 94, respectively, compared to the

noncatalyzed reaction. The downward bend of the curve at pH 10

and above is probably due to decomposition of the complex, and at

pH > 9 precipitation of what presumably is a metal hydroxide

species is observed for complex 5.

Figure 13. Dependence on pH of HPNP transesteriﬁcation enhanced

by [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2)

(9), Co (3) (b), Ni (4) (2)and Mn (5) (1). For clarity 2 and 4 are

displayed in (a) and 3 and 5 in (b) along with the noncatalyzed reaction

((). For comparison all traces are shown in (c). Lines displayed are only

to guide the eye and are not ﬁtted curves. The data for 2 have been

corrected for changes in the background absorbance (see Experimental

Section).

activity at high pH. An absolute maximum occurs at pH 8, but the

maximum rate enhancement with a factor of 143 relative to the

uncatalyzed reaction occurs at pH 7. The decline in activity after

pH 8 may be related to the formation of the μ-oxido-bridged

species which 2 (FeZn), based on UVꢀvis spectroscopy, is

proposed to convert to at high pH.

Surprisingly, complex 2 (FeZn), which according to metal

content should be a good model for kidney bean purple acid

phosphatase, shows a diﬀerent pH dependence and much lower

3879

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 14. Scanning kinetics for the decomposition of [Fe(III)Zn(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (2) in buﬀered water-acetonitrile solution at

pH 6 (a) and 10 (b). The arrows show the direction of change over time. The dotted lines represent the spectrum after 24 h.

Fe-containing complexes in this study. Previous studies of di-

nuclear Zn complexes of the same and similar ligands show that

they are more active than 2 toward HPNP transesteriﬁcation at

pH > 7,^47,83indicating that formation of homodinuclear Zn

complexes from 2 at pH > 8 might be responsible for the increase

in rate with time. This is likely the case also for complex 3 at pH >

9 as both these complexes, according to FAB-MS data (vide

supra) generate diZn and diCo complexes upon prolonged

standing in basic solutions. The formation of the homodinuclear

complex(es) appears to be slow (half-life of 2 > 12 h); thus the

observed activity based on initial rates may safely be attributed to

the heterodinuclear species.

All complexes appear to be able to support multiple turnovers,

as calculated from the extinction coeﬃcient of the produced 4-

nitrophenol (18 500 M^ꢀ1cm^ꢀ1) and the measured absorbance

after 5500 min for complex and substrate concentrations of

Figure 15. Kinetic traces of HPNP transesteriﬁcation enhanced by

[Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2) at pH

0.025 mM and 0.8 mM, respectively. The analysis is, however,

hampered by the slow reactions and decomposition at prolonged

times. Figure 15 shows kinetic traces for HPNP transesteriﬁca-

tion by 2 (FeZn), 3 (FeCo), and 4 (FeNi) at pH 8.0, 9.0, and 9.0,

respectively. In all cases, the plots indicate multiple turnovers

although for complex 2 only two turnovers were measured

during the time of the experiment (over 4 days). Complex 3,

which is the most active one toward HPNP transesteriﬁcation,

displays 13 turnovers within the ﬁrst 1000 min (after which the

detector was saturated). Complex 4 produces six turnovers over

4 days without any discernible decomposition or other unex-

pected change in the kinetic trace.

Hydrolysis of Bis-dinitrophenol Phosphate (BDNPP). The

hydrolysis of BDNPP was studied under identical conditions as

the HPNP transesterification, and the pH dependence of the

initial rates was investigated (Figure 16). As for the reaction with

HPNP, great diversity in reactivity is observed. Complexes 2

(FeZn) and 3 (FeCo) exhibit bell-shaped behavior for pH 5ꢀ7.5,

withaninitial rate maximum ataboutpH6.5. Thistypeofbehavior

is also observed for other similar heterodinuclear complexes^56ꢀ60

as well as for the PAPs.⁸⁴Complex 5 (FeMn) also reaches a

saturation point at pH 6.5, but no decline is observed at higher pH.

Interestingly, the initial rate of BDNPP hydrolysis enhanced by

complex4(FeNi) shows zerothorderdependenceonpH between

pH 5 and 6.5. This behavior is strikingly different from that of the

previously reported heterodinuclear complexes which all show

bell-shaped pH dependence.^56ꢀ60It may be noted that the order

of reactivity of our complexes at acidic pH is similar to that

observed for substituted forms of uteroferrin. Sykes and co-

workers⁸⁵have observed that at pH 4.9, the general reactivity

8.0 (;), Co (3) at pH 9.0 (; ;), and Ni (4) at pH 9.0 (; ;). The

3

dotted lines are linear ﬁts to the initial parts (20ꢀ80 min) of the curves.

Complex and substrate concentrations are 0.025 mM and 0.80 mM,

respectively.

The kinetic traces show that for all complexes there is a short

lag time of approximately 3ꢀ5 min (Supporting Information,

Figure S4). This suggests that the complexes undergo a rearran-

gement to become active in the HPNP transesteriﬁcation (or the

BDNPP hydrolysis, vide infra), possibly involving substitution of

coordinated groups by the substrate. Complexes 3 and 4 exhibit

well-behaved kinetics (Supporting Information, Figure S4).

Complex 5 also appears to behave well at low pH, but at higher

pH loss of activity and complicated absorbance versus time

dependence indicate degradation of the complex. This decom-

position becomes signiﬁcant already at pH 8.5 (after about 300

min) and severely distorts the traces at pH > 9.5, in agreement

with the observed precipitation at high pH (vide supra).

For 2 (FeZn) and 3 (FeCo) the rate increases over time at pH >

8 and 9, respectively. This suggests that the complexes are

slowly transformed into more active species. The change over

time in the UV/vis spectra of 2 at pH 6 and 10 without added

substrate is displayed in Figure 14a and b, respectively. The

complex appears to be fairly stable at pH 6 with only a slight

decrease in absorbance after 24 h. At pH 10 and during 380 min

(∼6 h) there is a small decrease in intensity of the shoulders

around 340 and 390 nm, but the spectra after 24 h indicate

signiﬁcant decomposition and a small amount of precipitate

could indeed be observed. The latter spectra lack the character-

istic oxido or phenolate to Fe(III) CT bands characteristic of the

3880

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Scheme 2

the product (dihydrogen phosphate) or some amino acid residue

in the active site important for reactivity,^12,19while for other

heterodinuclear complexes the substrate is proposed to be

inhibited to bind to the complex because of competition with

coordinated hydroxides forming at the divalent metal at high

pH.^56ꢀ60For 2 and 3 the decrease seems to be correlated with

the formation of the proposed μ-oxido diFe(III) species. At-

tempts to ﬁt the data to an equation for a diprotic model yielded

large errors for the dissociation constants that are probably due to

competing pH dependent processes. The zeroth order depen-

dence of complex 4 is surprising and indicates that water,

presumably noncoordinated, is the active nucleophile. As men-

tioned in the introduction, it has been proposed that the active

nucleophile for uteroferrin and its Fe(III)Mn(II) derivative is a

water (hydroxide) in the secondary coordination sphere of the

metal core (vide supra and Scheme 2d). The ESI-MS of 4 in

aqueous solution indicates the presence of hydroxide complexes

of similar stoichiometry as for 2 and 3; the reason for the

diﬀerence in pH dependence of these complexes for BDNPP

hydrolysis is currently not known.

For complexes 3 (FeCo), 4 (FeNi), and 5 (FeMn) there is an

increase in activity with increasing pH once the pH rises above

pH 7ꢀ7.5; this is in contrast to what has been observed for

similar heterodinuclear complexes that hydrolyze BDNPP.^56ꢀ60

This increase in the hydrolytic activity suggests that two parallel

pathways are active and either one dominates over the other at

diﬀerent pHs. The fact that no maximum rate appears to be

approached at very high pH would suggest that there is no pre-

equilibrium deprotonation to form a coordinated and nucleo-

philic hydroxide, suggesting that it either is a noncoordinated

hydroxide making the nucleophilic attack on the metal coordi-

nated substrate, or that there is a concerted deprotonation of a

coordinated water molecule during the attack. At low pH, where

there is no dependence on OH^ꢀ(or H^þ) for 4 (vide supra),

the nucleophile is probably noncoordinated water. In compar-

ison to similar heterodinuclear complexes used in previous

studies,^56ꢀ60the divalent metal has one less donating group

from the chelating ligand in the complexes used in this study.

This nominally vacant coordination site opens the possibility for

coordination of the substrate to the divalent metal ion even

when a hydroxide is formed on this site. The electrostatic

stabilization of the bound substrate might be suﬃcient to allow

for the observed rate enhancement for a bimolecular attack of

free hydroxides, but a hydroxide that is only coordinated to the

divalent metal might be suﬃciently reactive to avoid saturation

eﬀects even at high pH.

Figure 16. Dependence on pH of the initial rates for the hydrolysis of

BDNPP enhanced by [Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)]-

[PF₆] (M = Zn (2) (9), Co (3) (b), Ni (4) (2), and Mn (5) (1) .

The curves are not ﬁtted lines but only a guide for the eye.

order is FeZn > FeMn > FeCo > FeNi (for reaction with

H₂PO₄^ꢀ), while we observe the following order for our com-

plexes: FeZn > FeCo > FeMn > FeNi (pH 6.5).

At pH > 7.5 the rate for 2 continues to decrease after the rate

maximum, while for complex 3, 4, and 5 the rates increase with

increasing pH. For 2, a sigmoidal ﬁt to the declining part of the

rate proﬁle yields a pH of 7.9 for v_0max/2, which may correspond

to the pK_avalue of a coordinated water or hydroxide. This

assumed pK_acorresponds reasonably well to that derived from

the pH dependence of the oxido to Fe(III) charge transfer band

for 2 (pK_a8.3, vide supra and Supporting Information, Figure

S1), suggesting that the presumed oxido-bridged species thus

formed is inactive in the hydrolysis of BDNPP (and HPNP). As

noted previously, complex 3 also seems to form a species

reminiscent of the oxido-species formed for 2. The conversion

is, however, incomplete even at pH 10.5 which might explain why

3 shows activity at higher pH.

Downward bends as for 2 (FeZn), 3 (FeCo), and 5 (FeMn)

around pH 6.5 (Figure 16, inset) indicate that there is a sequence

of steps in the mechanism suggesting that some pH-dependent

rearrangement of the complex (or the substrate) prior to

phosphoester cleavage is necessary for the reaction to occur.

This might involve deprotonation of a metal-bound water

molecule, a process that is frequently associated with the activa-

tion of the attacking nucleophile. In this respect it is interesting to

note that 2, 3, and 5 have initial rate maxima at about the same

pH (6.5), while some variation in pH for the maxima may be

expected if the nucleophile were bound to the divalent metal.

Complex 5 has considerably lower activity which possibly can be

explained by the lower Lewis acidity of the Mn(II) ion. It should,

however, be noted that the ESI-MSþ spectrum of 5 in water/

acetonitrile indicated dissociation of the complex into a mono-

nuclear Fe(III) complex. The lower activity might, hence, be due

to decomposition of the complex. From the observations it can

be proposed that the active nucleophile is an hydroxide termin-

ally coordinated to the Fe(III) ion.

The bell-shaped pH dependence for 2 and 4 as well as other

heterodinuclear complexes and the PAP enzymes is the result of a

second deprotonation event that inhibits the reaction at higher

pH. For the enzymes this has been related to deprotonation of

3881

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Figure 17. Catalyst (a) and substrate (b) dependence of the initial rates for BDNPP hydrolysis enhanced by

[Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn (2) pH 6.5 (9), Co pH 6.5 (3) (b), Ni pH 5.5 (4) (2), and Mn (5) pH 6.5 (1).

Table 4. Kinetic Parameters Obtained from Fitting of the MichaelisꢀMenten Equation to the Data in Figure 17b (Zn4 =

[{Zn₂(IPCPMP)(OAc)}₂][PF₆])^47,72

K_M[mM]

K_ass[M^{ꢀ1 a}

]

V_max[10^ꢀ8M s^ꢀ1

]

k_cat^b[10^ꢀ4s^ꢀ1

]

k_cat/K_M[10^ꢀ2M^ꢀ1s^ꢀ1

]

k_cat/k_uncat^c[10³]

2

27.8

9.80

9.40

14.6

17.3

36.0

102

106

68.5

57.8

3.38

1.57

0.961

2.00

3.44

6.76

3.14

1.92

4.00

6.89

2.43

3.20

2.04

2.74

3.98

21.1

9.81

6.01

12.5

21.5

3

4

5

Zn4

^aK_ass= K_M^ꢀ1. ^bk_cat= V_max/[Catalyst]. ^ck_uncat= 3.2 ꢁ 10^ꢀ8s^ꢀ1in water:acetonitrile 1:1.⁹⁸

The dependence of the initial rates on the catalyst and

substrate concentrations was also investigated, and the results

are displayed in Figure 17. At high catalyst concentrations

(Figure 17a) there is a clear saturation eﬀect for 2 (FeZn), 4

(FeNi), and 5 (FeMn) while for 3 (FeCo) only a very weak eﬀect

is visible. This saturation is probably due to a nonfavorable

dimerization, possibly via hydroxide bridge(s), or polymeriza-

tion. Dimerizations of dinuclear Zn(II) complexes of IPCPMP⁴⁷

and ICIMP⁸³and related Ni(II) complexes,^48,86where two

dinuclear entities are bridged through the endogenous carbox-

ylate of the ligand, have been observed previously. A polymer-

ization/dimerization similar to that detected in the solid state

structure of 6 (Figure 5) is also possible. In the absence of water

(pure acetonitrile), such oligomerizations of the complexes do

not appear to take place, as the IR spectra of 2, 3, and 4 show that

even at 40 mM complex concentration, monodentate and thus

nonbridging carboxylates are clearly observed.

The dependence on substrate concentration also shows

saturation behavior (Figure 17b), indicating that the substrate

binds to the catalyst in a pre-equilibrium step before the

nucleophilic attack. The data could be ﬁtted to the Michae-

lisꢀMenten equation, yielding K_M, V_max, and k_catvalues as shown

in Table 4. Comparison of the K_Mvalues indicate that 2 (FeZn)

binds the substrate weaker than 3 (FeCo), 4 (FeNi),

and 5 (FeMn). The K_Mvalues for these heterodinuclear

complexes are on the same order although slightly lower

than those for analogous complexes of the BPBPMP ligand

(2.1ꢀ11 mM),^56ꢀ60which contains a terminal phenolate donor in

the same position as the carboxylate donor of IPCPMP. The

stronger binding and better electron-feeding properties of

the terminal phenolate might favor the binding of less strong

donors such as phosphate monoanions compared to competing

carboxylates (acetates). The weakbinding of the substrate by 2 can

also be compared to the tetranuclear Zinc complex

[{Zn₂(IPCPMP)(OAc)}₂][PF₆] (Zn4),^47,72which in solution

forms isolated dinuclear complexes and also show weak binding of

BDNPP (K_M= 17.3 M). The k_catvalues for 2ꢀ5 are of similar

magnitudes and comparable or slightly lower compared to those

for the heterodinuclear complexes of BPBPMP (4.5ꢀ18 ꢁ 10^ꢀ4

s

^ꢀ1).^56ꢀ60The value for 2 is slightly higher than that of 3, 4, and 5,

indicating a transition state of lower energy for 2. The k_catvalue for

the latter complex is similar to that for the Zn(II) complex of

IPCPMP (Zn4).^47,72However, the Zn(II)-containing complexes

appear to bind the substrate less eﬃciently yielding low overall

eﬃciency (k_cat/K_M), especially for 2. Complex 2 (FeZn) still

shows highest rate enhancement in the product forming

step relative to the uncatalyzed reaction (k_cat/k_uncat) among

the complexes studied here. The overall rate enhancement

(v_i,cat/v_i,uncat) compared to the uncatalyzed reaction (which does

not include a binding step of the substrate) is in the range

100ꢀ200 except for complex 4 (FeNi) which reaches a factor of

600 at pH 10. At extended reaction times, the absorbance versus

time dependence for 2 (FeZn), 4 (FeNi), and 5 (FeMn) revealed

biphasic behavior (Supporting Information, Figure S5) with

relatively abrupt changes in rate (absorbance per unit time) after

5 h, whichmaybeindicativeofdiesterase activity, that is, hydrolysis

of the monoester product formed upon hydrolysis of the BDNPP

substrate (cf. Supporting Information).

’ CONCLUSIONS

The mononuclear synthon [Fe(III)(H₂IPCPMP)(Cl₂)]-

[PF₆] (1),⁴⁹generated in situ, has been used to systemati-

cally produce heterodinuclear complexes of the formulas

3882

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

[Fe(III)M(II)(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (M = Zn

(2), Co (3), Ni (4), Mn (5)), [Fe(III)Zn(II)(IPCPMP)-

(mpdp)][PF₆] (6), and [Fe(III)Cu(II)(IPCPMP)(OAc)}₂-

(μ-O)][PF₆] (7). All complexes except 5 have been character-

ized by crystallography, revealing virtually identical structures for

2ꢀ4. On the basis of the physical and spectroscopic properties of

5, as determined by IR, M€ossbauer, ESI-MS, and magnetic

susceptibility measurements, it can be assumed to have the same

structure as 2ꢀ4 in the solid state. Complex 6 is a heteronuclear

coordination polymer in the solid state consisting of heterodi-

nuclear monomers, while 7 forms a dimer of dimer structure of

two heterodinuclear [Fe(III)Cu(II)(IPCPMP)(OAc)] bridged

by a μ-oxido group between the two Fe(III).

nucleophile (Scheme 2f). A dimerization process for 2, 4, and

5 probably inhibits the BDNPP hydrolysis at high complex

concentrations. An analysis of the substrate dependence using

MichaelisꢀMenten kinetics shows that BDNPP binds weaker to

2 than 3ꢀ5 prior to hydrolysis, possibly indicating a diﬀerent

coordination mode of the substrate in the former case. The

overall eﬃciencies of the complexes are, however, comparable to

previously studied systems.^56ꢀ60

Further studies are directed toward elucidating the species

distribution under conditions relevant for phosphoester hydro-

lysis, as well as studying the complexes in solution by EPR and

EXAFS to get structural information on the species that are active

in the reactions.

Infrared spectroscopy conﬁrmed that the Fe(III) is selectively

bound in the tetradentate pocket for 2ꢀ4 with monodentate

coordination of the endogenous carboxylate of IPCPMP. The

structure is retained in acetonitrile solution as observed by both

IR and ESI MS, but in aqueous solution mass spectrometry

indicates that the acetate bridges may be substituted for hydro-

xides and that 2ꢀ4 partly dissociate into mononuclear Fe(III)

complexes. Complete dissociation is observed for 5 under these

conditions.

Electronic spectroscopy over a wide pH range indicates a clean

conversion of one species to another for 4 and 5; the latter

species is proposed to be a μ-hydroxido bridged heterodinuclear

complex. The same species is probably observed as an inter-

mediate for 2 and 3 before forming a μ-oxido bridged dimer of

dimer of similar structure to 7.

’ EXPERIMENTAL SECTION

General Methods. All solvents were of at least 99.5% purity and

used as received or dried either by distillation from CaH₂(methanol,

2-propanol) or by keeping over 3 Å molecular sieves in a sealed bottle

overnight (acetone). Reagents were of at least 99% purity and used

as received. The bis-hexafluorophosphate salt of the ligand 2-(N-

isopropyl-N-((2-pyridyl)methyl)aminomethyl)-6-(N-(carboxylmethyl)-

N-((2-pyridyl)-methyl)aminomethyl)-4-methylphenol (H₄IPCPMP-

(PF₆)₂H₂O or H₄L) was synthesized as described elsewhere.⁷²

3

Physical Methods. UVꢀvis spectroscopy and kinetic measure-

ments were performed on a Varian 300 Bio UV/vis spectrophotometer

equipped with a 12-position thermostatted cell changer. Infrared spectra

were collected on a Nicolet Avatar 360 FT-IR spectrometer for solid KBr

discs and a Digilab Excalibur FTIR spectrometer equipped with an

Axiom Analytical DPR-210 dipper system and a MCT detector for liquid

samples. ESI-MS were collected on a Waters Micromass ZQ 4000 probe

with capillary potential 3.5 Kvolts, source cone 20ꢀ25 V, source

temperature 70 ꢀC, and direct infusion of 20 μL/min. FAB mass spectra

were collected on a JEOL SX-102 spectrometer with 2-nitrobenzyl

alcohol (NBA) as matrix. All mass spectrometry data are reported as m/z

with probable species, and the relative intensity of the peaks in % based

on the Fe⁵⁶and Zn⁶⁴isotope given within brackets.

M€ossbauer spectra show that 1, 3, and 5 contain high spin

Fe(III) while no parameters could be extracted for 2 and 4

because of relaxation eﬀects. Magnetic susceptibility measure-

ments support the designation of 2ꢀ5 as being Fe(III)M(II) (M =

Zn, Co (high spin), Ni (high spin), Mn (high spin)).

Complexes 2ꢀ5 are active toward phosphoester cleavage

for both HPNP (an RNA mimic) and BDNPP (a DNA mimic)

but display considerable diﬀerences in reactivity. For the

HPNP reaction complex 2 shows very low activity, which is

probably due to the formation of the μ-oxido bridged dimer of

dimer at higher pH. Complex 3 is the most active for HPNP

transesteriﬁcation but shows no saturation eﬀects, indicating

that either the hydroxy functionality of the substrate is not

coordinated to a metal (and deprotonated before the attack)

or that the deprotonation of the coordinated hydroxyl group

and its attack on the substrate is a concerted process

(Scheme 2a). A similar mechanism for 5, which is less active,

can be invoked. For complex 4 a saturation behavior is

observed, and the mechanism here likely involves coordina-

tion and deprotonation of the hydroxyl group before its attack

on the phosphorus center (Scheme 2b), although pre-equi-

librium deprotonation of a water molecule coordinated to the

Ni(II) can not be excluded (Scheme 2c).

For BDNPP hydrolysis by 2ꢀ5 diﬀerent mechanisms are

observed at low and high pH, respectively. For 2, 3, and 5 at low

pH, the mechanism is proposed to involve an attack by a

hydroxide coordinated terminally to Fe(III) (Scheme 2d or e)

or, alternatively, activation by a water molecule in the second

coordination sphere by this hydroxide. For 4 a noncoordinated

water molecule is the more probable nucleophile (Scheme 2f). At

high pH, 2 shows very low activity, again probably because of

the formation of a μ-oxido dimer of dimer, while 3ꢀ5 display

an increase in activity. This increase in activity may be due

to a mechanism where a noncoordinated hydroxide acts as

Synthesis of Complexes. For the heterodinuclear complexes only

representative large scale procedures are presented, demonstrating the

two methods (A and B).

[Fe(H₂IPCPMP)Cl₂][PF₆] (1). To a solution of 68.4 mg (0.42 mmol)

FeCl₃in 4 mL of dry acetonitrile was added 295.6 mg (0.39 mmol) of

H₄IPCPMP(PF₆)₂H₂O (H₄L), and the color turned intensely blue-

3

purple. After stirring at room temperature for 2 h, butyl acetate was

added until complete precipitation of a purple solid, which was isolated

by centrifugation and washed with more butyl acetate. The solid was

recrystallized from acetone/butyl acetate 2:1 (v/v) by slow evaporation

yielding 170.2 mg (60.1%)

Elem. Anal. C₂₇H₃₅Cl₂F₆FeN₄O₃P %Calc. C, 44.10; H, 4.80; N,

7.62; Found: C, 44.48; H, 4.26; N, 8.10; FAB-MSþ m/z 560

([Fe(IPCPMP)Cl]þNa^þ, 100), 538 ([Fe(IPCPMP)Cl]þH^þ, 60);

IR (KBr, cm^ꢀ1) 3095(m), 1662(s, ꢀCO₂antisym.), 1621(s, CdC

arom.), 1582(s, CdC arom.), 1476(s, CdC arom.), 1449(m),

1407(m), 1388(m), 1260(s, sym.), 842(vs, PF₆), 769(s, PF₆), 577(s);

UV/vis acetonitrile, nm, (ε) 557 (1302 M^ꢀ1cm^ꢀ1), 358 (4061

M

^ꢀ1cm^ꢀ1), 296 (9228 M^ꢀ1cm^ꢀ1), 254 (14405 M^ꢀ1cm^ꢀ1

)

[FeZn(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (2)

Method B ꢀ Deprotonation by Excess Sodium Acetate and

Filtering through Celite. A total of 151.2 mg of H₄L (0.20 mmol)

was dissolved in 3 mL of a 1:1 (v/v) methanol/acetone mixture, and

33.8 mg of FeCl₃(0.208 mmol) was added to this solution, whereupon

the color turned intensely blue-purple (and a few dark blue crystals were

formed). No visible change occurred when 46.0 mg (0.21 mmol)

3883

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

Zn(OAc)₂2H₂O was added, but upon stepwise addition of 12.4 mL of a

[FeMn(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (5). This complex was synthe-

sized using method B and the same procedure used for 2 (vide supra) but

under inert atmosphere using Schlenk techniques and degassed solvents

starting from 161.9 mg (0.214 mmol) H₄L, 34.6 mg (0.213 mmol) of

3

0.1 M sodium acetate (1.24 mmol) solution in methanol, the color

gradually changed to pink-red. After stirring for 3 h at room temperature,

all solvent was removed to yield a pink-purple oily residue that was

dissolved in acetone and filtered through Celite to remove the sodium

salts formed. The filtrate was evaporated, and the residue was recrys-

tallized from methanol/2-propanol (3:1 v/v) by slow evaporation. This

yielded 120 mg (66.7%) of small dark pink crystals.

FeCl₃, 52.3 mg (0.213 mmol) of Mn(OAc)₂4H₂O, and 2.6 mL (1.3

3

mmol) of a 0.5 M degassed sodium acetate solution. Yielded 158.2 mg

(87%) of a dark purple solid.

Elem. Anal. [FeMn(IPCPMP)(OAc)₂(CH₃OH)][PF₆] C₃₁H₄₀F₆-

FeMnN₄O₈P %Calc. C, 43.68; H, 4.73; N, 6.57 Found C, 42.87; H,

4.23; N, 6.59; FAB-MSþ m/z 675 ([FeZn(IPCPMP)(OAc)₂]^þ, 100),

616 ([FeZn(IPCPMP)(OAc)]^þ, 60); ESI MSþ acetonitrile m/z 675

([FeZn(IPCPMP)(OAc)₂]^þ, 100); water/acetonitrile 1:1 m/z 502

Elem. Anal. [FeZn(IPCPMP)(OAc)₂(CH₃OH)]PF₆C₃₁H₄₀F₆Fe-

N₄O₈PZn Calc. C, 43.15; H, 4.67; N, 6.49; Found: C, 42.53 H, 4.5 N,

6.45; FAB-MSþ m/z 684 ([FeZn(IPCPMP)(OAc)₂]^þ, 100), 625

([FeZn(IPCPMP)(OAc)]^þ, 60); ESI-MSþ, acetonitrile, m/z 684

([FeZn(IPCPMP)(OAc)₂]^þ, 100); water/acetonitrile 1:1 (v/v) m/z

642 ([FeZn(IPCPMP)(OAc)(OH)]^þ, 80), 582 ([FeZn(IPCPMP)-

(O)]^þ, 100), 502 ([Fe(IPCPMP)]^þ, 95); IR (KBr, cm^ꢀ1) 2968(w),

2926(w), 2852(w), 2825(w), 1655(m, terminal ꢀCO₂antisym.),

1607(s, bridg.-CO₂antisym.), 1476(w, CdC arom.), 1419(m,br, bridg.

ꢀCO₂sym.), 1342(m, terminal ꢀCO₂sym.), 1264(w), 1155(w),

1023(m), 841(vs, PF₆), 556(m); UV/vis acetonitrile nm (ε) 494 (br,

([Fe(IPCPMP)]^þ, 100) IR (KBr, cm^ꢀ1

) 2969(w), 2928(w),

2854(w), 1654(m, terminal ꢀCO₂antisym.), 1606(s, bridg.-CO₂anti-

sym.), 1475(w, CdC arom.), 1419(m, bridg. ꢀCO₂sym.), 1387 (m),

1342(m, terminal ꢀCO₂sym.), 1255(w), 1153(w), 1021(m), 841(vs),

558(m); UV/vis acetonitrile nm 512 (802 M^ꢀ1cm^ꢀ1), 422 (707

M^ꢀ1cm^ꢀ1), 357 (1440 M^ꢀ1cm^ꢀ1), 284 (8380 M^ꢀ1cm^ꢀ1), 258 (10

800 M^ꢀ1cm^ꢀ1

)

1375 M^ꢀ1cm^ꢀ1), 288 (15 654 M^ꢀ1cm^ꢀ1), 257 (24 695 M^ꢀ1cm^ꢀ1

)

[FeZn(IPCPMP)(mpdp)][PF₆] (6). A total of 31.0 mg (0.041 mmol) of

[FeCo(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (3)

Method A ꢀ Deprotonation with Tributyl Amine. To 199.9 mg

H₄IPCPMP(PF₆)₂H₂O was dissolved in 0.5 mL of methanol and a few

3

drops of acetonitrile. To this solution, 6.8 mg (0.042 mmol) of FeCl₃

dissolved in 0.3 mL of methanol was added, and the color turned

intensely blue purple. No change occurred when 5.7 mg (0.042 mmol)

of ZnCl₂and 9.1 mg of H₂mpdp, both dissolved in 0.3 mL of methanol

each, were added to the solution, but when 58.7 μL (0.25 mmol) of

tributyl amine was added in portions, the color changed to deep red. One

milliliter of 2-propanol was added before filtering the solution and leaving

it to slowly evaporate which yielded crystals of X-ray diffraction quality.

FAB-MS m/z (rel. intensity, %) 787 ([FeZn(IPCPMP)(mpdp)]^þ,

100); ESI-MSþ (methanol) m/z (rel. intensity, %) 786 ([FeZn-

(IPCPMP)(mpdp)]^þ, 100).

(0.264 mmol) of H₄IPCPMP(PF₆)₂H₂O dissolved in 4 mL of

3

methanol and 1 mL of acetonitrile was added 42.8 mg (0.26 mmol) of

FeCl₃dissolved in 2 mL of methanol, and the solution turned intensely

blue-purple. Then 65.7 mg (0.26 mmol) of Co(OAc)₂4H₂O was added

3

dissolved in 2 mL of methanol, but no visible change occurred until 4

equiv (252 μL, 1.06 mmol) of tributyl amine was added dropwise,

whereupon the color changed to pink-red. About 1 mL of 2-propanol

was added, and the solution was left to slowly evaporate which over a

few days yielded 196 mg (86.7%) of dark purple crystals of X-ray

quality.

Elem. Anal. [FeCo(IPCPMP)(OAc)₂(CH₃OH)][PF₆] CH₃OH

[FeCu(IPCPMP)(OAc)}₂(μ-O)][PF₆]₂(7). This complex was synthe-

sized using method B and the same procedure used for 3 (vide supra),

starting from 200.1 mg (0.26 mmol) of H₄L, 42.7 mg (0.26 mmol) of

3

C₃₂H₄₄CoF₆FeN₄O₉P %Calc. C, 43.26; H, 4.99; N, 6.31; Found C,

43.83; H, 5.69; N 6.40; FAB-MS m/z 679 ([FeCo(IPCPMP)(OAc)₂]^þ,

100), 620 ([FeCo(IPCPMP)(OAc)]^þ, 60); ESI-MS acetonitrile m/z

679 ([FeZn(IPCPMP)(OAc)₂]^þ, 100); Water/acetonitrile 1:1 (v/v)

m/z 637 ([FeCo(IPCPMP)(OAc)(OH)]^þ, 40), 595 ([FeCo-

(IPCPMP)(OH)₂]^þ, 55), 577 ([FeCo(IPCPMP)(O)]^þ, 100), 502

([Fe(IPCPMP)]^þ, 50) IR (KBr, cm^ꢀ1) 2968(w), 2926(w), 2854(w),

2823(w), 1657(m, terminal ꢀCO₂antisym.), 1607(s, bridg. ꢀCO₂

antisym.), 1476(w, CdC arom.), 1412(m, br, bridg. ꢀCO₂sym.),

1341(m, terminal ꢀCO₂sym.), 1265(w), 1155(w), 1023(m), 841(vs),

555(m); UV/vis acetonitrile nm 491 (1440 M^ꢀ1cm^ꢀ1), 288 (16 000

FeCl₃, 52.7 mg (0.26 mmol) of Cu(OAc)₂H₂O, and 377 μL of tributyl

3

amine (1.58 mmol). The reaction yielded a brown solid that had to be

crystallized from acetone/water 4:1 (v/v) by slow evaporation which

gave 20 mg (5%) of a brown solid.

Elem. Anal. {[FeCu(IPCPMP)(OAc)]₂(O)}(PF₆)₂C₅₆H₆₆Cu₂F_12-

Fe₂N₈O₁₁P₂%Calc. C, 43.23; H, 4.28; N, 7.20; Found: 42.44, H 4.34, N

7.21; FAB-MSþ m/z 1410 ([FeCu(IPCPMP)(OAc)}₂(μ-O)]^þþ

PF₆^ꢀ, 30), 1284 ([FeCu(IPCPMP)(OAc)}₂(μ-O)]^2þþ F^ꢀ, 35),

1266 ([FeCu(IPCPMP)(OAc)}₂(μ-O)]^þ, 45), 777 ([FeCu-

(IPCPMP)(OAc)₂]

þ

NBA-NO₂-OH, 55), 684 (([FeCu-

M

^ꢀ1cm^ꢀ1), 257 (23 500 M^ꢀ1cm^ꢀ1

[FeNi(IPCPMP)(OAc)₂(CH₃OH)][PF₆] (4). This complex was synthe-

)

(IPCPMP)(OAc)₂]^þ, 30), 625 (([FeCu(IPCPMP)(OAc)]^þ, 50), 582

([FeCu(IPCPMP)(O)]^þ, 100), 566 ([FeCu(IPCPMP)]^þ, 55); ESI-

sized using method B and the same procedure used for 2 (vide supra)

starting from 147.4 mg (0.20 mmol) H₄L, 32.9 mg (0.20 mmol) of

MSþ

acetonitrile

m/z

1409

([FeCu(IPCPMP)(OAc)}₂-

(O)]^2þþPF₆^ꢀ, 2), 632 ([{FeCu(IPCPMP)(OAc)}₂(O)]^2þ, 100).

Upon increasing the cone voltage fragments corresponding to

Fe(III)Cu(II)(IPCPMP)(OAc)₂]^þ(m/z = 683), [Fe(III)Cu(II)-

(IPCPMP)(O)]^þ(m/z = 581), and [Fe(III)Cu(II)(IPCPMP)-

(OAc)(O)(F)]^þ(m/z = 1283) appear, which are reasonable fragmenta-

tion products from the μ-oxido complex (7). Water/acetonitrile m/z 599

([FeCu(IPCPMP)(OH)₂]^þ, 100), 590 ([FeCu(IPCPMP)(OH)}₂-

FeCl₃, 50.5 mg (0.203 mmol) of Ni(OAc)₂4 H₂O, and 12.4 mL of a 0.1

3

M sodium acetate (1.24 mmol) solution in methanol. Yielded 101 mg

(58.1%) of small dark pink crystals.

Elem. Anal. [FeNi(IPCPMP)(OAc)₂(CH₃OH)]PF₆C₃₁H₄₀F₆FeN_4-

NiO₈P %Calc. C, 43.49; H, 4.71; N, 6.54; Found C, 43.07; H, 4.50; N,

6.47; FAB-MSþ m/z 679 ([FeNi(IPCPMP)(OAc)₂]^þ, 100), 620

([FeNi(IPCPMP)(OAc)]^þ, 60); ESI-MSþ acetonitrile m/z 679

([FeNi(IPCPMP)(OAc)₂]^þ, 100), 1383 ([FeNi(IPCPMP)(OAc)]₂-

þPF₆}^þ, 1); Water/acetonitrile 1:1 m/z 636 ([FeNi(IPCPMP)-

(OAc)(OH)]^þ, 100), 577 ([FeNi(IPCPMP)(O)]^þ, 80), 502

([Fe(IPCPMP)]^þ, 20) IR (KBr, cm^ꢀ1) 2969(w), 2924(w), 2855(w),

2825(w), 1653(m, terminal ꢀCO₂antisym.), 1608(s, bridg.-CO₂anti-

sym.), 1477(w, CdC arom.), 1414(m, br, bridg. ꢀCO₂sym.), 1344(m,

terminal ꢀCO₂sym.), 1265(w), 1150(w), 1025(m), 842(vs), 558(m);

UV/vis acetonitrile nm 494 (1410 M^ꢀ1cm^ꢀ1), 288 (16 000

(O)]^2þ, 40), 581 ([FeCu(IPCPMP)(O)]^þ, 90). IR (KBr, cm^ꢀ1

)

2920(w), 1660(s, terminal ꢀCO₂antisym.), 1610(m), 1554(s,

bridg.-CO₂antisym.), 1479(s, CdC arom.), 1446(s, bridg. ꢀCO₂

sym.), 1321(m, terminal ꢀCO₂sym.), 1294(w), 1022(w), 847(vs),

559(s); UV/vis acetonitrile nm 565 (sh, 741 M^ꢀ1^ꢀ1), 415 (2520

s

M

^ꢀ1cm^ꢀ1), 336 (sh, 10 000 M^ꢀ1cm^ꢀ1), 280 (sh, 19 900 M^ꢀ1cm^ꢀ1),

260 (26 700 M^ꢀ1cm^ꢀ1).

Kinetic Measurements. The increase in concentration of the

products, p-nitrophenolate (PNP) and 2,4-dinitrophenolate (BNP),

M

^ꢀ1cm^ꢀ1), 257 (24 300 M^ꢀ1cm^ꢀ1

)

3884

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

were monitored at 25 ꢀC by UV/vis spectroscopy at 400 nm in quartz

suprasil cuvettes using a Cary 300 Bio spectrophotometer equipped with

a 12-position thermostatted cell changer. For the study of the pH

dependence, substrate and complex concentrations were 0.80 mM and

0.25 mM, respectively, and the ionic strength and pH were kept constant

by using total concentrations of 0.1 M NaClO₄and 0.01 M buffer (MES

pH 5ꢀ6.5, MOPS 7.0ꢀ7.5, EPPS 8.0ꢀ8.5, CHES 9.0ꢀ9.5, CAPS

10.0ꢀ11.0). The pH of the buffer was adjusted in standard solutions

using a calibrated pH meter before addition to the cuvettes. Each cuvette

was prepared by consecutive addition of 980 μL of acetonitrile, 30 μL

of a 0.0375 M standard solution of the complex in acetonitrile/H₂O

(2:1 v/v), and 970 μL of a 0.0207 M buffer solution containing

0.207 M NaClO₄. After mixing, the background absorption was mea-

sured. Then 20 μL of a 0.080 M standard solution of the substrate in

H₂O was added and after quick mixing the increase in absorption over

time was measured at 400 nm, first every minute but after 4 h, every

5 min, and after 8 h, every 15 min. The initial rates were calculated

by fitting a straight line to the curve corresponding to abs <5% of

the maximum absorption at full conversion (usually the first 30 min).

The dissociation constant of the phenol products (pK_a,PNP= 7.15,

pK_a,BNP= 4.07) were taken into account when calculating the total

concentration of phenol from the absorption of the phenolate at

400 nm (ε_PNP= 18500 M^ꢀ1cm^ꢀ1, ε_BNP= 12100 M^ꢀ1cm^ꢀ1). The

absorbance of complex 2 changes over time under the kinetic conditions,

and this process occurs over a similar time frame as the HPNP

transesterification. At pH < 6 this becomes the dominant contribution

but could partly be taken into account by subtracting the absorbance

versus time data for complex 2 under the same conditions but without

the substrate.

The Denzo-Scalepack⁸⁸program package was used for cell refine-

ments and data reductions. The structures were solved by direct

methods using the SIR97,⁸⁹SIR2004⁹⁰or SHELXS-97⁹¹programs with

the WinGX⁹²graphical user interface. A semiempirical absorption

correction (XPREP in SHELXTL,⁹³SORTAV⁹⁴or SADABS⁹⁵) was

applied to all data. Structural refinements were carried out using

SHELXL-97.⁹¹In 3 and 4, four fluorines of the PF₆anion were

disordered over two sites with occupancies 0.55/0.45 and 0.6/0.4

respectively. The disordered fluorines were restrained so that their U_ij

components approximate isotropic behavior. Furthermore, each disor-

dered pair of fluorines was refined with equal anisotropic displacement

factors. In 3, 4, 6, and 7, the OH hydrogen atoms were located from the

difference map but constrained to ride on their parent atom, with U_iso

=

1.5. Other hydrogens were positioned geometrically and constrained to

ride on their parent atoms, with CꢀH = 0.95ꢀ1.00 Å and U_iso= 1.2ꢀ1.5

U_eq(parent atom). The crystallographic details are summarized in

Table 1 and selected bond lengths and angles in Table 2.

’ ASSOCIATED CONTENT

S

Supporting Information. Crystallographic data in CIF

b

format. Further details are given in Figures S1ꢀS5 and Table S1.

This material is available free of charge via the Internet at http://

pubs.acs.org.

’ AUTHOR INFORMATION

Corresponding Author

*E-mail: ebbe.nordlander@chemphys.lu.se.

Complex and substrate concentration dependence was studied

for BDNPP hydrolysis enhanced by [Fe(III)M(II)(IPCPMP)(OAc)₂-

(CH₃OH)][PF₆] (M = Zn (2), Co (3), Ni (4), and Mn (5)) at pH 6.5

for 2, 3, and 5 and 5.5 for 4. The measurements and preparations of the

solutions in the cuvettes were made as above with total NaClO₄

concentration 0.1 M but with total buﬀer concentration increased to

0.050 M. For the dependence on complex concentration diﬀerent

volumes of stock solution of the complexes (2.0 mM in acetonitrile)

was added to yield total concentrations of 0.020, 0.050, 0.1, 0.25, 0.5, and

1.0 mM of the complex in the cuvette while keeping the substrate

concentration constant at 1.0 mM. For the dependence on substrate

concentration a similar addition of stock solution of NaBDNPP

(40.0 mM in water) yielded total concentrations of 0.25, 0.5, 1.0, 2.0,

4.0, and 6.0 mM in the cuvette while the complex concentration was kept

constant at 0.050 mM. Initial rates were calculated as above. The

substrate dependence data was analyzed using MichaelisꢀMenten

equation).

’ ACKNOWLEDGMENT

This research has been supported by a grant from the Swedish

Research Council (VR) and the Royal Physiographic Society

(www.fysiografen.org) and has been carried out within the

framework of the International Research Training Group Metal

Sites in Biomolecules: Structures, Regulation and Mechanisms

(www.biometals.eu). We thank Dr. Jean-Marc Latour, Commis-

ꢁ

sariat ꢀa I’Energie Atomique (CEA), Grenoble, France, for a

generous gift of the ligand mpdp.

’ REFERENCES

(1) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Chem. Rev. 1996,

96, 2239–2314.

(2) Burgess, B. K.; Lowe, D. J. Chem. Rev. 1996, 96, 2983–3012.

(3) Howard, J. B.; Rees, D. C. Proc. Natl. Acad. Sci. U.S.A. 2006,

103, 17088–17093.

(4) Solomon, E. I.; Sundaram, U. M.; Machonkin, T. E. Chem. Rev.

1996, 96, 2563–2606.

(5) Sazinsky, M. H.; Lippard, S. J. Acc. Chem. Res. 2006, 39, 558–566.

(6) Fontecilla-Camps, J. C.; Volbeda, A.; Cavazza, C.; Nicolet, Y.

Chem. Rev. 2007, 107, 4273–4303.

€

Mossbauer Spectroscopy. M€ossbauer spectra were recorded

with a ⁵⁷Co source in a Rh matrix using an alternating constant

acceleration Wissel M€ossbauer spectrometer operated in the transmis-

sion mode and equipped with a Janis closed-cycle helium cryostat.

Isomer shifts are given relative to iron metal at ambient temperature.

Simulation of the experimental data was performed with the Mfit

program.⁸⁷

(7) Solomon, E. I.; Xie, X. J.; Dey, A. Chem. Soc. Rev. 2008,

37, 623–638.

Magnetic Susceptibility. Magnetic susceptibility measurements

were made on a Quantum-Design MPMS-5S SQUID magnetometer

equipped with a 5 T magnet in the range from 2 to 295 K. The powdered

samples were contained in a gel bucket and fixed in a nonmagnetic

sample holder. Each raw data file for the measured magnetic moment

was corrected for the diamagnetic contribution of the sample holder and

the sample.

X-ray Structure Determinations. The crystals of 2, 3, 4, 6, and 7

were immersed in cryo-oil, mounted in a Nylon loop, and measured at a

temperature of 120 K. The X-ray diffraction data were collected on a

Nonius Kappa CCD diffractometer using Mo K_Rradiation (λ= 0.71073 Å).

(8) Stenkamp, R. E. Chem. Rev. 1994, 94, 715–726.

(9) Magnus, K. A.; Ton-That, H.; Carpenter, J. E. Chem. Rev. 1994,

94, 727–735.

(10) Dau, H.; Haumann, M. Coord. Chem. Rev. 2008, 252, 273–295.

(11) Renger, G.; Renger, T. Photosynth. Res. 2008, 98, 53–80.

(12) Wilcox, D. E. Chem. Rev. 1996, 96, 2435–2458.

(13) Karplus, P. A.; Pearson, M. A.; Hausinger, R. P. Acc. Chem. Res.

1997, 30, 330–337.

(14) Benini, S.; Rypniewski, W. R.; Wilson, K. S.; Mangani, S.; Ciurli,

S. J. Am. Chem. Soc. 2004, 126, 3714–3715.

3885

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

(15) Matsui, M.; Fowler, J. H.; Walling, L. L. Biol. Chem. 2006,

387, 1535–1544.

(48) Carlsson, H.; Haukka, M.; Bousseksou, A.; Latour, J. M.;

Nordlander, E. Inorg. Chem. 2004, 43, 8252–8262.

(16) Weston, J. Chem. Rev. 2005, 105, 2151–2174.

(17) Yang, W.; Lee, J. Y.; Nowotny, M. Mol. Cell 2006, 22, 5–13.

(18) Mitic, N.; Smith, S. J.; Neves, A.; Guddat, L. W.; Gahan, L. R.;

Schenk, G. Chem. Rev. 2006, 106, 3338–3363.

(49) Jarenmark, M.; Carlsson, H.; Trukhan, V. M.; Haukka, M.;

Canton, S. E.; Walczak, M.; Fullagar, W.; Sundstrom, V.; Nordlander, E.

Inorg. Chem. Commun. 2010, 13, 334–337.

(50) Jarenmark, M.; Carlsson, H.; Haukka, M.; Nordlander, E.,

unpublished results

(19) Twitchett, M. B.; Sykes, A. G. Eur. J. Inorg. Chem.

1999, 2105–2115.

(51) Carrell, C. J.; Carrell, H. L.; Erlebacher, J.; Glusker, J. P. J. Am.

Chem. Soc. 1988, 110, 8651–8656.

(20) Oddie, G. W.; Schenk, G.; Angel, N. Z.; Walsh, N.; Guddat,

L. W.; de Jersey, J.; Cassady, A. I.; Hamilton, S. E.; Hume, D. A. Bone

2000, 27, 575–584.

(21) Nuttleman, P.; Roberts, R. J. Biol. Chem. 1990, 265, 12192–

12199.

(22) Cashikar, A. G.; Kumaresan, R.; Rao, N. M. Plant Physiol. 1997,

114, 907–915.

(23) Str€ater, N.; Jasper, B.; Scholte, M.; Krebs, B.; Duﬀ, A. P.;

Langley, D. B.; Han, R. L.; Averill, B. A.; Freeman, H. C.; Guss, J. M.

J. Mol. Biol. 2005, 351, 233–246.

(24) Guddat, L. W.; McAlpine, A. S.; Hume, D.; Hamilton, S.; de

Jersey, J.; Martin, J. L. Structure 1999, 7, 757–767.

(25) Lindqvist, Y.; Johansson, E.; Kaija, H.; Vihko, P.; Schneider, G.

J. Mol. Biol. 1999, 291, 135–147.

(26) Uppenberg, J.; Lindqvist, F.; Svensson, C.; Ek-Rylander, B.;

Andersson, G. J. Mol. Biol. 1999, 290, 201–211.

(52) Lambert, E.; Chabut, B.; ChardonNoblat, S.; Deronzier, A.;

Chottard, G.; Bousseksou, A.; Tuchagues, J. P.; Laugier, J.; Bardet, M.;

Latour, J. M. J. Am. Chem. Soc. 1997, 119, 9424–9437.

(53) Borovik, A. S.; Papaefthymiou, V.; Taylor, L. F.; Anderson,

O. P.; Que, L. J. Am. Chem. Soc. 1989, 111, 6183–6195.

(54) Shannon, R. D. Acta Crystallogr., Sect. A 1976, 32, 751–767.

(55) Holman, T. R.; Juarez-Garcia, C.; Hendrich, M. P.; Que, L.;

Munck, E. J. Am. Chem. Soc. 1990, 112, 7611–7618.

(56) Lanznaster, M.; Neves, A.; Bortoluzzi, A. J.; Aires, V. V. E.;

Szpoganicz, B.; Terenzi, H.; Severino, P. C.; Fuller, J. M.; Drew, S. C.;

Gahan, L. R.; Hanson, G. R.; Riley, M. J.; Schenk, G. J. Biol. Inorg. Chem.

2005, 10, 319–332.

(57) Batista, S. C.; Neves, A.; Bortoluzzi, A. J.; Vencato, I.; Peralta,

R. A.; Szpoganicz, B.; Aires, V. V. E.; Terenzi, H.; Severino, P. C. Inorg.

Chem. Commun. 2003, 6, 1161–1165.

(27) Str€ater, N.; Klabunde, T.; Tucker, P.; Witzel, H.; Krebs, B.

Science 1995, 268, 1489–1492.

(58) Lanznaster, M.; Neves, A.; Bortoluzzi, A. J.; Szpoganicz, B.;

Schwingel, E. Inorg. Chem. 2002, 41, 5641–5643.

(28) Klabunde, T.; Str€ater, N.; Frohlich, R.; Witzel, H.; Krebs, B.

J. Mol. Biol. 1996, 259, 737–748.

(59) Karsten, P.; Neves, A.; Bortoluzzi, A. J.; Lanznaster, M.; Drago,

V. Inorg. Chem. 2002, 41, 4624–4626.

(29) Schenk, G.; Gahan, L. R.; Carrington, L. E.; Mitic, N.; Valiza-

deh, M.; Hamilton, S. E.; de Jersey, J.; Guddat, L. W. Proc. Natl. Acad. Sci.

U.S.A. 2005, 102, 273–278.

(30) Beck, J. L.; McConachie, L. A.; Summors, A. C.; Arnold, W. N.;

De Jersey, J.; Zerner, B. Biochim. Biophys. Acta 1986, 869, 61–68.

(31) Beck, J. L.; De Jersey, J.; Zerner, B.; Hendrich, M. P.; Debrunner,

P. G. J. Am. Chem. Soc. 1988, 110, 3317–3318.

(60) Xavier, F. R.; Neves, A.; Casellato, A.; Peralta, R. A.; Bortoluzzi,

A. J.; Szpoganicz, B.; Severino, P. C.; Terenzi, H. n.; Tomkowicz, Z.;

Ostrovsky, S.; Haase, W.; Ozarowski, A.; Krzystek, J.; Telser, J.; Schenk,

G.; Gahan, L. R. Inorg. Chem. 2009, 48, 7905–7921.

(61) Holman, T. R.; Andersen, K. A.; Anderson, O. P.; Hendrich,

M. P.; Juarez-Garcia, C.; Munck, E.; Que, L., Jr. Angew. Chem. 1990,

102, 933–5.

(32) Schenk, G.; Ge, Y.; Carrington, L. E.; Wynne, C. J.; Searle, I. R.;

Carroll, B. J.; Hamilton, S.; de Jersey, J. Arch. Biochem. Biophys. 1999,

370, 183–189.

(33) Schenk, G.; Boutchard, C. L.; Carrington, L. E.; Noble, C. J.;

Moubaraki, B.; Murray, K. S.; de Jersey, J.; Hanson, G. R.; Hamilton, S.

J. Biol. Chem. 2001, 276, 19084–19088.

(34) Smoukov, S. K.; Quaroni, L.; Wang, X.; Doan, P. E.; Hoﬀman,

B. M.; Que, L. J. Am. Chem. Soc. 2002, 124, 2595–2603.

(35) Cox, R. S.; Schenk, G.; Mitic, N.; Gahan, L. R.; Hengge, A. C.

J. Am. Chem. Soc. 2007, 129, 9550–9551.

(36) Neves, A.; Lanznaster, M.; Bortoluzzi, A. J.; Peralta, R. A.;

Casellato, A.; Castellano, E. E.; Herrald, P.; Riley, M. J.; Schenk, G. J. Am.

Chem. Soc. 2007, 129, 7486–7487.

(62) Schepers, K.; Bremer, B.; Krebs, B.; Henkel, G.; Althaus, E.;

Mosel, B.; Mueller-Warmuth, W. Angew. Chem. 1990, 102, 582–4.

(63) Jovito, R.; Neves, A.; Bortoluzzi, A. J.; Lanznaster, M.; Drago,

V.; Haase, W. Inorg. Chem. Commun. 2005, 8, 323–327.

(64) Ghiladi, M.; Jensen, K. B.; Jiang, J. Z.; McKenzie, C. J.; Morup,

S.; Sotofte, I.; Ulstrup, J. J. Chem. Soc., Dalton Trans. 1999, 2675–2681.

(65) Biswas, P.; Ghosh, M.; Dutta, S. K.; Florke, U.; Nag, K. Inorg.

Chem. 2006, 45, 4830–4844.

(66) Dutta, S. K.; Werner, R.; Florke, U.; Mohanta, S.; Nanda, K. K.;

Haase, W.; Nag, K. Inorg. Chem. 1996, 35, 2292–2300.

(67) Kurtz, D. M. Chem. Rev. 1990, 90, 585–606.

(68) Adams, H.; Bradshaw, D.; Fenton, D. E. Inorg. Chim. Acta 2002,

332, 195–200.

(37) Mitic, N.; Hadler, K. S.; Gahan, L. R.; Hengge, A. C.; Schenk, G.

J. Am. Chem. Soc. 2010, 132, 7049–7054.

(69) Holman, T. R.; Wang, Z.; Hendrich, M. P.; Que, L. Inorg. Chem.

1995, 34, 134–9.

(38) See the special issue on Biomimetic Inorganic Chemistry.

Chem. Rev., 2004, 104, 347ꢀ1200.

(70) Murch, B. P.; Bradley, F. C.; Boyle, P. D.; Papaefthymiou, V.;

Que, L. J. Am. Chem. Soc. 1987, 109, 7993–8003.

(71) Deacon, G. B.; Phillips, R. J. Coord. Chem. Rev. 1980, 33, 227–250.

(72) Jarenmark, M.; Csapo, E.; Singh, J.; W€ockel, S.; Farkas, E.;

Meyer, F.; Haukka, M.; Nordlander, E. Dalton Trans. 2010,

39, 8183–8194.

(73) Ito, K.; Bernstein, H. J. Can. J. Chem. 1956, 34, 170–178.

(74) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M.

Advanced Inorganic Chemistry; Wiley-Interscience Publication: New

York, 1999.

(75) Drueke, S.; Wieghardt, K.; Nuber, B.; Weiss, J.; Fleischhauer,

H. P.; Gehring, S.; Haase, W. J. Am. Chem. Soc. 1989, 111, 8622–8631.

(76) Brown, C. A.; Remar, G. J.; Musselman, R. L.; Solomon, E. I.

Inorg. Chem. 1995, 34, 688–717.

(39) Gavrilova, A. L.; Bosnich, B. Chem. Rev. 2004, 104, 349–384.

(40) Parkin, G. Chem. Rev. 2004, 104, 699–767.

(41) Meyer, F. Eur. J. Inorg. Chem. 2006, 3789–3800.

(42) Suzuki, M.; Kanatomi, H.; Murase, I. Chem. Lett. 1981, 1745–

1748.

(43) Ghiladi, M.; McKenzie, C. J.; Meier, A.; Powell, A. K.; Ulstrup,

J.; Wocadlo, S. J. Chem. Soc., Dalton Trans. 1997, 4011–4018.

(44) Albedyhl, S.; Averbuch-Pouchot, M. T.; Belle, C.; Krebs, B.;

Pierre, J. L.; Saint-Aman, E.; Torelli, S. Eur. J. Inorg. Chem.

2001, 1457–1464.

(45) Neves, A.; Debrito, M. A.; Drago, V.; Griesar, K.; Haase, W.

Inorg. Chim. Acta 1995, 237, 131–135.

(46) Belle, C.; Gautier-Luneau, I.; Karmazin, L.; Pierre, J. L.;

Albedyhl, S.; Krebs, B.; Bonin, M. Eur. J. Inorg. Chem. 2002, 3087–3090.

(47) Jarenmark, M.; Kappen, S.; Haukka, M.; Nordlander, E. Dalton

Trans. 2008, 993–996.

(77) Jarenmark, M.; Turitsyna, E. A.; Haukka, M.; Shteinman, A. A.;

Nordlander, E. New J. Chem. 2010, 34, 2118–2121.

(78) Greenwood, N. N.; Gibb, T. C. M€ossbauer Spectroscopy;

Chapman and Hall Ltd: London, 1971.

3886

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Inorganic Chemistry

ARTICLE

(79) Kahn, O. Molecular Magnetism; Wiley-VCH, Publishers Inc.:

New York, 1993.

(80) Simulation of the experimental magnetic data with a full-matrix

diagonalization of exchange coupling and Zeeman splitting was per-

formed with the julX program (E. Bill, Max-Planck Institute for

Bioinorganic Chemistry, M€ulheim/Ruhr, Germany).

(81) Buchanan, R. M.; Mashuta, M. S.; Richardson, J. F.; Webb, R. J.;

Oberhausen, K. J.; Nanny, M. A.; Hendrickson, D. N. Inorg. Chem. 1990,

29, 1299–301.

(82) Schenk, G.; Peralta, R. A.; Batista, S. C.; Bortoluzzi, A. J.;

Szpoganicz, B.; Dick, A. K.; Herrald, P.; Hanson, G. R.; Szilagyi, R. K.;

Riley, M. J.; Gahan, L. R.; Neves, A. J. Biol. Inorg. Chem. 2008,

13, 139–155.

(83) Carlsson, H.; Haukka, M.; Nordlander, E. Inorg. Chem. 2004,

43, 5681–5687.

(84) Dietrich, M.; M€unstermann, D.; Suerbaum, H.; Witzel, H. Eur.

J. Biochem. 1991, 199, 105–113.

(85) Twitchett, M. B.; Schenk, G.; Aquino, M. A. S.; Yiu, D. T. Y.;

Lau, T. C.; Sykes, A. G. Inorg. Chem. 2002, 41, 5787–5794.

(86) Carlsson, H.; Haukka, M.; Nordlander, E. Inorg. Chem. 2002,

41, 4981–4983.

(87) Bill, E. Mﬁt; Max-Planck Institute for Bioinorganic Chemistry:

M€ulheim/Ruhr, Germany.

(88) Otwinowski, Z.; Minor, W. In Macromolecular Crystallography,

Part A; Carter, C., Sweet, J., Eds.; Academic Press: New York, 1997;

Vol. 276, pp 307ꢀ326.

(89) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;

Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna,

R. J. Appl. Crystallogr. 1999, 32, 115–119.

(90) Burla, M. C.; Caliandro, R.; Camalli, M.; Carrozzini, B.;

Cascarano, G. L.; De Caro, L.; Giacovazzo, C.; Polidori, G.; Spagna,

R. J. Appl. Crystallogr. 2005, 38, 381–388.

(91) Sheldrick, G. Acta Crystallogr., Sect. A 2008, 64, 112–122.

(92) Farrugia, L. J. Appl. Crystallogr. 1999, 32, 837–838.

(93) Sheldrick, G. M. SHELXTL, v. 6.14-1; Bruker AXS, Inc.:

Madison, WI, 2005.

(94) Blessing, R. Acta Crystallogr., Sect. A 1995, 51, 33–38.

(95) Sheldrick, G. M. SADABS - Bruker Nonius scaling and absorption

correction, v. 2.10; Bruker AXS, Inc.: Madison, WI, 2003.

(96) Farrugia, L. J. Appl. Crystallogr. 1997, 30, 565.

(97) Macrae, C. F.; Edgington, P. R.; McCabe, P.; Pidcock, E.;

Shields, G. P.; Taylor, R.; Towler, M.; van de Streek, J. J. Appl. Crystallogr.

2006, 39, 453–457.

(98) Bunton, C. A.; Farber, S. J. J. Org. Chem. 1969, 34, 767–772.

3887

dx.doi.org/10.1021/ic1020324 |Inorg. Chem. 2011, 50, 3866–3887

Article Doi

Synthesis, characterization, and reactivity studies of heterodinuclear complexes modeling active sites in purple acid phospatases

DOI: 10.1021/ic1020324

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ic1020324

Full text of DOI:10.1021/ic1020324

Products guided by the article

R&D Labs maybe for 1285536-99-3

Relevant to this article

Hot Product