Tetrahedron Letters 53 (2012) 2082–2087

Contents lists available at SciVerse ScienceDirect

Tetrahedron Letters

journal homepage: www.elsevier .com/ locate / tet let

Tris-ACE-(6,60-bipyridyl)-a-cyclodextrin: synthesis and chiralityof the tripodal metal complexes

Guillaume Poisson, Florence Dumarçay-Charbonnier, Alain Marsura ⇑SRSMC, Nancy Université & CNRS, Faculté des Sciences et Techniques, B. 70239, F-54506 Vandœuvre-lès-Nancy Cedex, France

a r t i c l e i n f o

Article history:Received 29 November 2011Revised 2 February 2012Accepted 8 February 2012Available online 15 February 2012

Keywords:Tris-ACE-a-cyclodextrinMetallocyclodextrinsCircular DichroismHelicates

0040-4039/$ - see front matter � 2012 Elsevier Ltd. Adoi:10.1016/j.tetlet.2012.02.038

⇑ Corresponding author. Tel.: +33 (0) 3 83 68 49 55E-mail address: Alain.Marsura@pharma.uhp-nancy

a b s t r a c t

This Letter describes the synthesis of a C3-symmetrical tris-ACE-(6,60-bipyridyl)-a-CyD via ‘one-pot’Staudinger-Aza-Wittig reaction (SAW). Metal complexation behavior was investigated by UV–vis and Cir-cular Dichroism. Coordination of 6,60-bipyridyl units with CuII and NiII cations, spontaneously generates astrong exciton coupling-type positive Cotton effect proving the formation of a K-helix with a left-handedscrew propeller (M).

� 2012 Elsevier Ltd. All rights reserved.

In the past decade, seminal work has described the preparationand properties of new metallocyclodextrins (metallo-CyDs) molec-ular receptors.1 Metallo-CyD architectures are potentially able toact in a wide range of applications such as: biomimeticcatalysts,2a–c photoactive devices,3a–c chemical sensors4a,b due totheir properties after the coordination to metals and formation ofhost-guest complexes with small invited molecules. A few yearsago one of our research has concerned the synthesis and complex-ation properties of novel symmetrical ureido-5,50-bipyridyl-a-CyDtripodes.5,9 As expected, selective dual metal complexation andfluorescence properties controlled via HSAB metal classificationwere detected and analyzed. Nevertheless, considering the confor-mational aspects, the ureido-5,50-bipyridyl tethers were foundunable to generate the expected chiral triple helices after coordina-tion with appropriate metal cations. It is well known that the spon-taneous folding into a helical secondary structure is based on ageneral molecular self-organization process enforced by the confor-mational information encoded within the primary structure of themolecular strand itself.6 The key features which determine the out-come of the assembly process are the ligand design, flexibility of thelinker groups joining the coordination sites, and the stereochemicalpreferences of the coordinating metal ion.7 Failure in the helicateformation with ureido-5,50-bipyridyl-a-CyD tripod was certainlydue to an improper preorganization (orientation) and hasmotivated the synthesis of C3-symmetrical tris-ACE-(6,60-bis-hetero-cylic)-a-CyDs in which the heterocyclic unit is anchored at the

ll rights reserved.

; fax: +33 (0) 3 83 68 23 45..fr (A. Marsura).

6-position. Surprisingly, this minor modification induces a dramaticchange into the conformation of the ureido-bis-heterocyclic strandsaround the upper-rim of the rigid CyD core.

The C3-symmetrical tris-ACE-(6,60-bipyridyl)-a-CyD tripod 9was synthesized from 6A,6C,6E-triazido-6A,6C,6E-trideoxy-6B,6D,6F-tri-O-acetyl-hexakis-2,3-di-O-acetyl-cyclomaltohexaose8 8(Scheme 1) by the tandem Staudinger-Aza-Wittig (SAW) key reac-tion in a 50% yield9 (Scheme 1). The 6-monoaminomethyl-60-methyl-2,20-bipyridin 4 was obtained in three steps from 6,60-di-methyl-2,20-bipyridin 1.10 Compounds 6–8 were prepared byknown methods.11 All compounds 2–9 were characterized by spec-troscopic methods. The spectroscopic data are in agreement withthe assigned structures (see References 12–14 and Supplementarydata). As expected, the IR spectrum of 9 exhibited characteristicfrequencies of the carbonyl functions: urea mCO–NH at 1640 cm�1;ester mCO–O at 1750 cm�1 (acetates) and aromatic double bonds ofbipyridyl groups at 1560 cm�1. The 13C NMR spectrum in solution(DMSO) indicates that the ligand has a C3-symmetry. The signalscorresponding to the NH–CO–NH carbonyl carbons at d158.1 ppm, the CO–O carbonyl carbons at d 171.4–162.9 ppm andthe signals corresponding to the bipyridyl carbons from d 138.1 to118.5 ppm were detected.

Metal ion complexation

The UV–vis spectrum of the free tripod 9 recorded in MeOHshows two maxima at kmax = 244 nm (14700 mol�1 dm3 cm�1)and kmax = 290 nm (32000 mol�1 dm3 cm�1). These absorptionscorrespond to the p–p⁄ transitions of bipyridyls and to the n–p⁄

urea and carboxylate carbonyl double bonds.

BrHO

OHOH

OHB, D, F

1212

6

i ii

BrAcO

OAc

B, D, F

12

A, C, E

A, C, E

N3AcO

OAc

B, D, F

12

A, C, E

iii

N N N N N N

N N

Br N3

H2N

N N

H2N

NH

AcO

OAc

B, D, F

12

A, C, EC O

HN

N N

+SAW

1 2 3

4

5 6 7

8 4

9

NBS/AIBN/CCl4

reflux, 90 min, hn

NaN3/DMSO

PPh3/NH4OH 10%

i = NBS/PPh3/DMFii = Ac2O/pyridin; 7h ,80 °Ciii = NaN3/DMF; 72h, 60 °C

SAW = PPh3/DMF/CO2 + 4

Scheme 1. Synthetic pathway of 9.

Figure 1. Schematic representation of the a-CyD tripod 9 and its potent complex-ation sites.

G. Poisson et al. / Tetrahedron Letters 53 (2012) 2082–2087 2083

The extinction value calculated from the molecular extinctioncoefficient in 9 is ca 10700 mol�1 dm3 cm�1 per bipyridine unit;this value is within the range found in the literature.5 As illustratedin Figure 1, the tripod 9 represents the first term of a new set ofheterotritopic molecular receptor having two potent metal coordi-nation sites: one formed by the bipyridyl units and the second byureas and/or carboxylates, the third site is formed by the CyD toruscavity, dedicated to the formation of host–guest complexes via theinclusion of small organic hydrophobic molecules.

The titrations of 9 with ‘hard’ to ‘borderline’ (HSAB theory)cations were monitored by UV–vis spectroscopy. As expected andbased on the results previously established5,9 in our researchgroup, ‘borderline’ cations as CuII and NiII were complexed at thebipyridyls while ‘hard’ cations as FeIII or RuIII were coordinated tothe carbonyl urea/ester functions (Fig. 2.) according to the HSABtheory. The titration curves (Fig. 3.) showed examples of mononu-clear podandate [9/CuII] and dinuclear podandate [9/FeIII/CuII]formation.

The ligand speciation curves (Fig. 4.) allowed the determinationof [1:1] and [1:1:1] stoichiometries for [9/CuII] and [9/FeIII/CuII],respectively. The complexation constants log(b11) have been also

determined20 for the four cations in MeOH as the solvent. It shouldbe assumed that the FeIII cation in 9 is coordinated and localized at

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

230 260 290n

Abs

A

0

0.05

0.1

0.15

0.2

0.25

0.3

250 270 290 310 330 350nm

Abs

C

Figure 3. Spectrophotometric titration of ligand 9 (c = 1.0 � 10�5 mol L�1) (A) in MeOH w1.0 equiv (colored lines); (C) in MeOH with RuCl3�6H2O.

Figure 2. Postulated folding pathway for the metal ion-assisted self-assembly ofCuII and FeIII/CuII left-handed (M) triple-helices of 9.

2084 G. Poisson et al. / Tetrahedron Letters 53 (2012) 2082–2087

the urea functions (‘hard’ carbonyls centers) with a log(b11)constant value of 14.9 ± 0.31 whereas for example, CuII and NiII

‘borderline’ cations are localized and coordinated in a different site,likely to the nitrogens of the three bipyridyl units giving hexa-coordinated octahedral complexes with constant values of10.1 ± 0.13 and 8.8 ± 0.06, respectively.

Formation of a triple helix by the coordination with metal cat-ions was ascertained by Circular Dichroïsm. As expected but differ-ently of 5,50 analogues,5 coordination of ureido-6,60-bipyridyl armsfor example, CuII cation spontaneously generates a strong excitoncoupling-type positive Cotton effect (De = +33 mol�1 cm�1) be-tween the bipyridine units (i.e. a negative band at 290 nm, a posi-tive band at 315 nm and an isodichroic point at approximatively302 nm) proving a high helicity induction (Fig. 5.).

The same phenomenon was observed by adding CuII cation tothe [FeIII /9] mononuclear complex, giving a FeIII/CuII dinuclearcomplex. In this case a similar positive Cotton effect of(De = +16 mol�1 cm�1) was observed. It should be noted (Fig. 5.)that coordination of FeIII at the ‘urea’ site, did not induce any supe-rior helicity into the podandate. The difference in helicity (almostthe half) between the mononuclear and the dinuclear specieswas attributed to a less degree of flexibility of ureido-bipyridylarms when urea-carbonyls are first involved in coordination withthe FeIII cation. Nevertheless, a helix is characterized by an axis, ascrew sense and a pitch. This helicity is a special case of chiralityas defined earlier by Cahn, Ingold, and Prelog.15 It may be right-handed (P) or left-handed (M). The remaining question is whetherthis induced helicity was right-handed or left-handed in the pres-ent case.

The large exciton coupling-type positive Cotton effects of forexample, [CuII/9] and [CuII/FeIII/9] CD spectra, are pointed in favor

320 350 380m

B(dinuclear complex)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

240 265 290 315 340 365nm

Abs

ith CuCl2�2H2O; (B) in MeOH with Fe2(SO4)3 1.0 equiv (black lines) then CuCl2�2H2O

00.0000010.0000020.0000030.0000040.0000050.0000060.0000070.0000080.0000090.00001

0 0.25 0.5 0.75 1 1.25 1.5Metal/Ligand equivs.

conc

.(mol

.L-1

)

00.0000010.0000020.0000030.0000040.0000050.0000060.0000070.0000080.0000090.00001

0 0.25 0.5 0.75 1 1.25 1.5 1.75 2Metal/Ligand equivs.

conc

. (m

ol.L

-1)

00.0000010.0000020.0000030.0000040.0000050.0000060.0000070.0000080.0000090.00001

0 0.25 0.5 0.75 1 1.25 1.5 1.75 2Metal/Ligand equivs.

conc

. (m

ol.L

-1)

A B

C

Figure 4. Ligand speciation curves for (A) FeIII, (B) CuII and (C) RuIII with podand 9. Free metal (dashed line), complexes (solid line), free ligand (dashed—dotted line).

-5

0

5

10

15

20

25

30

35

250 300 350λ nm λ nm

Ellip

ticité

(mde

g)

-4-202468

1012141618

280 300 320 340

Ellip

ticité

(mde

g)

FeIII

A B

Figure 5. Circular Dichroïsm titration of ligand 9 (c = 5.0 � 10�5 mol L�1) (A) in MeOH with CuCl2�2H2O; (B) in MeOH with Fe2(SO4)3 [1:1] complex, followed by CuCl2�2H2O1 equiv.

G. Poisson et al. / Tetrahedron Letters 53 (2012) 2082–2087 2085

of a single diastereoisomer as indicated earlier by Lehn andco-workers.16 Previous examinations on the chirality of C3-sym-metric FeIII tripodal-a-cyclodextrin derivatives17,18 in differentsolvents, showed ligand having short spacers (between the cyclo-dextrin torus and the metal coordination site) exhibited negativeexciton coupling and D-helicity in all solvents, while those havinglong spacers formed a K�helicate and a positive exciton cou-pling.18 It is currently, admitted that a positive coupling-typeexciton implies a complexation with a K-helix (same handednessas for the naturally ferrichrome19). Looking at the [CuII/9] and[CuII/FeIII/9] CD titration spectra, (Fig. 5) one can see the free ligand9 in MeOH displayed itself a weak exciton coupling-type positiveCotton effect. The addition of CuII aliquots to 9 strongly enhancesthis effect in the same sense, so it may be concluded reasonablyto the preferential formation of a K-helix with a left-handed screwpropeller (M) (See Fig. 6).

Coordination of NiII cation to the bipyridyl nitrogens induced anunexpectedly low kinetic to complete the [1:1] mononuclearcomplex formation of 9 (3 h) compared to the immediacy with CuII.Circular Dichroïsm titration of 9 with NiII also generates a strongexciton coupling-type positive Cotton effect (De = +91 mol�1 cm�1)between the bipyridine units (Fig. 7.). This confirms the preferen-tial formation of an analogous left-handed (M) triple K-helix.

Furthermore, it has been demonstrated that CuII and NiII com-plexation to the bipyridyl units of 9 in a protic solvent gave thesame helicity with an apparent higher rotational strength in thecase of NiII, but a large difference was observed in kinetics betweenthe two cations. The low rate of the NiII cation could be explainedin terms of equilibration reactions, by which a bipyridine ligandmoves from one NiII complex to another too slowly.21a,b

In summary, a C3-symmetrical tris-ACE-(6,60-bipyridyl)-a-CyDwas prepared. Its spectroscopic behavior was examined to investi-

00.050.1

0.150.2

0.250.3

0.350.4

0.45

230 255 280 305 330 355nm

abs

-0.045-0.04

-0.035-0.03

-0.025-0.02

-0.015-0.01

-0.0050

0 2500 5000 7500 10000time (s)

Δ-ab

s

A B

Figure 6. Spectrophotometric titration of ligand 9 (c = 1.0 � 10�5 mol L�1) (A) in MeOH with NiCl2�6H2O; (a delay of 30 min. was respected between each curve to reachequilibrium) (B) Kinetic of the [NiII/9] mononuclear complex formation (c = 1.0 � 10�5 mol L�1) with NiCl2�6H2O (1.0 equiv).

-20

0

20

40

60

80

100

245 265 285 305 325 345

nm

Ellip

ticity

(mde

g)

Figure 7. Circular Dichroïsm titration of ligand 9 (c = 5.0 � 10�5 mol L�1) in MeOHwith NiCl2�6H2O.

2086 G. Poisson et al. / Tetrahedron Letters 53 (2012) 2082–2087

gate the formation of tripodal mononuclear and dinuclear metalcomplexes. It was observed that the 2,20-bipyridine units anchoredat their 6-position, induce the formation of preferentially a-CyD-based tripod K�helicates with a high chirality. Further experi-ments are under progress to examine the complexation behaviorwith other metal cations notably with cations having a tetrahedralcoordination geometry. In extension, the synthesis and character-ization of new a-CyD-based tripods with different bis-heterocyclicunits are in progress and complexation behavior of these new com-pounds will be explored in detail.

Acknowledgments

This work was supported by the MESR and the CNRS. Authorsare grateful to Dr. J.-P. Joly for correcting the manuscript. Wewould also thank Mrs B. Fernette for recording NMR spectra, Mr.F. Dupire for recording Mass spectra.

Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.tetlet.2012.02.038.

References and notes

1. Sliwa, W.; Girek, T. Heterocycles 2003, 60, 2147–2183.2. (a) Braga, S. S. Curr. Org. Chem. 2010, 14, 1356–1361; (b) Barr, L.; Dumanski, P.;

Easton, C.; Harper, J.; Lee, K.; Lincoln, S. F.; Meyer, A.; Simpson, J. S. J. Incl.Phenom. Macrocycl. Chem. 2004, 50, 19–24; (c) Barr, L.; Easton, C.; Lee, K.;Lincoln, S. F.; Simpson, J. S. Tetrahedron Lett. 2002, 43, 7797–7800.

3. (a) Haider, J. M.; Pikramenou, Z. Chem. Soc. Rev. 2005, 34, 120–132; (b) PereiraSilva, M-J. J.; Haider, J. M.; Heck, R.; Chavarot, M.; Marsura, A.; Pikramenou, Z.

Supramol. Chem. 2003, 15, 563–571; (c) Haider, J. M.; Williams, R. M.; De Cola,L.; Pikramenou, Z. Angew. chem., Int. Ed. 2003, 42, 1830–1833.

4. (a) Ogoshi, T.; Harada, A. Sensors 2008, 8, 4961–4982; (b) Mallard-Favier, I.;Blach, P.; Cazier, F.; Delattre, F. Carbohydr. Res. 2009, 344, 161–166.

5. Heck, R.; Marsura, A. Tetrahedron Lett. 2004, 45, 281–284.6. Cuccia, L. A.; Ruiz, E.; Lehn, J.-M.; Homo, J. C.; Schmutz, M. Chem. Eur. J. 2002, 8,

3448–3457.7. (a) Bilyk, A.; Harding, M. M.; Turner, P.; Hambley, T. W. J. Chem. Soc., Dalton

Trans. 1994, 2783–2790; (b) Shu, M. H.; Sun, W. Y.; Duan, C. Y.; Fu, Y.-J.; Zhang,W.-J.; Tang, W.-X. J. Chem. Soc. Dalton Trans. 1999, 729–734.

8. Menuel, S.; Porwanski, S.; Marsura, A. N. J. Chem. 2006, 30, 603–608.9. Menuel, S.; Corvis, Y.; Rogalska, E.; Marsura, A. N. J. Chem. 2009, 33, 554–560.

10. Rodriguez-Ubis, J.-C.; Alpha, B.; Plancherel, D.; Lehn, J.-M. Helv. Chim. Acta1984, 67, 2264–2269.

11. Woods, C. R.; Benaglia, M.; Cozzi, F.; Siegel, J. F. Angew. Chem., Int. Ed. 1996, 35,1830–1833.

12. 6-monoazidomethyl-60-methyl-2,20-bipyridin (3): 6-bromomethyl-60-methyl-2,20-bipyridin (1.25.10–3 mol, 0.330 g) is added portionwise to a solution ofsodium azide (7.5.10–3 mol., 0.487 g) in DMSO (5 mL) under argon. Themixture was stirred 2 h. at 70 �C under argon, then cooled at rt. Distilled water(15 mL) is added and the solution is extracted by toluol (5 � 10 mL). Theorganic phase is dried over anh. MgSO4, filtered and evaporated to give 3 as ayellow oil (0.133 g, 47%).1H NMR (CDCl3, 400 MHz): d = 8.29 (d, 1H, H3,J = 8 Hz), 8.14 (d, 1H, H30 , J = 8 Hz), 7.71 (t, 1H, H4, J = 8 Hz), 7.59 (t, 1H, H40 ,J = 8 Hz), 7.18 (d, 1H, H5, J = 8 Hz), 7.06 (d, 1H, H50 , J = 8 Hz), 4.39 (s, 2H, CH2

bipy), 2.53 (s, 3H, CH3 bipy). 13C NMR (CDCl3, 100 MHz): d = 158.3 (q), 156.8(q), 155.6 (q), 155.5 (q), 138.2 (CH), 137.5 (CH), 123.8 (CH), 122.0 (CH), 120.6(CH), 118.7 (CH), 55.7 (CH2-N3), 25.0 (CH3).

13. 6-monoaminomethyl-60-methyl-2,20-bipyridin (4): Triphenylphosphine(1.18.10–3 mol.) was added to a solution of 3 (5.9.10–4 mol, 0.133 g) in driedDMF. The mixture was stirred 1 h. at rt under argon, then cooled to 0 �C andNH4OH (3.3 mL, 10%) was added dropwise to the solution. The reactionmixture was stirred about 2 h. until temperature is gone to rt. The reactionmixture was evaporated to dryness and the residue was dissolved into CH2Cl2/MeOH 2/1 (5 mL). The solution was then evaporated to dryness and the solidresidue was purified by column chromatography on silica gel, (CH2Cl2/MeOHgradient) to give pure 4 as a white solid (0.082 g, 70%). IRFT (KBr,m = cm�1) = 3454 (N–H), 3055 (C–H arom.), 2925 (C–H alkyls); 1H NMR(DMSO-d6, 400 MHz): d = 8.24 (t, 2H, J = 8.56 Hz, H3, H30), 7.89 (t, 1H,J = 7.8 Hz, H4), 7.82 (t, 1H, J = 7.56, H40), 7.48 (d, 1H, J = 7.28 Hz, H5), 7.30 (d,1H, J = 7.52 Hz, H50), 3.91 (s, 2H, CH2–NH2), 2.56 (s, 3H, CH3); ESIMS(C12H13N3): 200.12 [M+H]+.

14. 6A, 6C, 6E-(6-methyleneureido-60-methyl-2,20-bipyridyl)- 6A, 6C, 6E-trideoxy-6B, 6D,6+-tri-O-acetyl-hexakis-2,3-di-O-acetyl-cyclomaltohexaose (9): A solution oftriphenylphosphine (0.011 mol., 2.88 g, 40 equiv) in methylene chloride(10 mL) was added to a solution of 6A,6C,6E-triazido-6A,6C,6E-trideoxy-6B,6D,6F-tri-O-acetyl-hexakis-2,3-di-O-acetyl-cyclomaltohexaose (2.68.10–4 mol, 0.450 g, 1 equiv) 8 in methylene chloride (30 mL). The mixture wasstirred and flushed with CO2 at rt 30 min. Then 6-amino-60-methyl-2,20-bipyridine (1.07.10–3 mol., 0.214 g) in methylene chloride (10 mL) was addeddropwise into the solution. The mixture was stirred and flushed with CO2 at rt24 h. more. The solution was then evaporated to dryness and the solid residuewas purified by column chromatography on silica gel, loading using CH2Cl2 andeluting with a CH2Cl2/MeOH gradient to give pure 9 as a pale yellow powder(0.304 g, 50%). ½a�29

D +74.6 (CHCl3). IRFT (KBr, m = cm�1) = 1750 (CH3CO), 1650(NHCONH), 1580 (C@C). 1H NMR (CDCl3, 400 MHz): d = 8.15 (s, 6H, H3, H30),7.69 (s, 6H, H4, H40), 7.19 (s, 6H, H5, H50), 5.67–3.38 (m, 42H, H1, H2, H3, H4,H5, H6 cyclodextrin), 2.84 (s, 9H, CH3 bipy), 2.66 (s, 6H, CH2-bipy), 2.25–1.90(m, 45H, CO–CH3). 13C NMR (CDCl3, 100 MHz): d = 171.4, 171.1, 170.9, 170.7,169.6 [CH3CO], 158.3 [NHCONH], 158.2 [Cq], 155.5 [Cq], 123.9, 123.6 [Cq],122.5, 122.4 [CH], 120.3, 119.9 [CH], 118.7, 118.5 [CH], 97.1 [C1], 77.6 [C4],71.8, 71.5, 71.1, 69.9 [C2, C3, C5, C6], 63.6 [C06], 46.0 [CH2 bipy], 24.7 [CH3 bipy],21.4, 21.3, 21.2, 21.1 [CH3CO]. ESI-HRMS: [C105H126N12O45 2274.7900] =2298.7706 [M+Na+H]+.

G. Poisson et al. / Tetrahedron Letters 53 (2012) 2082–2087 2087

15. Cahn, R. S.; Ingold, C.; Prelog, V. Angew. Chem., Int. Ed. Engl. 1966, 5, 385–415.16. Zarges, W.; Hall, J.; Lehn, J.-M. Helv. Chim. Acta 1991, 74, 1843–1852.17. Hori, Y.; Hayashi, J.-I.; Tamagaki, S. Nippon Tagaku Kaishi 1998, 6, 417–424.18. Masuda, T.; Hayashi, J.-I.; Tamagaki, S. J. Chem. Soc., Perkin Trans. 2 2000, 161–167.19. Raymond, K. N.; Mueller, G. Matzanke Top. Curr. Chem. 1984, 123, 49–102.20. Log b values were calculated using SPECFIT� program v. 3.0, ed. R. A. Binstead,

Spectrum Software Associates 1993–2001 Gampp, H.; Maeder, M.; Meyer, C. J.;

Zuberbühler, A. D. Talanta 1985, 32, 95–101; Gampp, H.; Maeder, M.; Meyer, C.J.; Zuberbühler, A. D. Talanta 1986, 33, 943–951.

21. (a) Vander Griend, D. A.; Kwabena Bediako, D.; De Vries, M. J.; DeJong, N. A.;Heeringa, L. P. Inorg. Chem. 2008, 47, 656–662; (b) Holyer, R. H.; Hubbard, C. D.;Kettle, S. F. A.; Wilkins, R. G. Inorg. Chem. 1965, 4, 929–935.