1690 J.C.S. DaltonStereochemistry of Bis(salicylaldiminato)metal(i!) Compounds. Part 1.Bis[N- (2,6-dial kylphenyl)salicylideneiminato]nickel(~~) Compounds andFive- and Six-co-ordinate Pyridine and Picoline ComplexesBy Brian M. Higson, Derek A. Lewton, and E. Donald McKenzie, Chemistry Department, The University,Sheffield S3 7HFThe compounds bis[N-(2,6-dimethylphenyl)salicylideneiminato] nickel ( t i ) , bis[N- (2,6-diethylphenyl)salicylidene-iminato]nickel(ii), and analogous species with substituents at the 3 and/or 5 positions of the salicyl moiety,strongly prefer the four-co-planar geometry in the solid state and in non-donor solvents. They are forced intotetrahedral structures only by bulky substitution (NO,, but not Me, OMe, Br, or CI) at the 3 position of the salicylmoiety; and they do not form six-co-ordinate oligomers as do the N-phenyl and 3- and 4-substitu:ed N-phenylcompounds.In pyridine and the picolines, however, in spite of bulk steric effects (B-strain), they form five- andsix-co-ordinate complexes with the bases, to an extent determined by ( a ) the steric effect {Me > Et (on N-phenyl) ;and y-pic >, py > a-pic) ; and ( b ) the electron-withdrawing substituents on the salicyl moiety (NO, > OMe >Br CI > Me H > 5.6-benzo: and 3- > 5-). Electronic spectra (5-30 lo3 cm-l), bulk magnetic sus-ceptibility, and I H n.m.r. data (including paramagnetic shifts) have been used to define structures.THE stereochemistry of the nickel(r1) salicylaldiminates(I) has been widely studied, and most of the presentlyavailable data are in several recent re~iews.l-~2,6-dimethylphenyl and 2,6-diethylphenyl.These sub-stituents have a combination of steric and electroniceffects, which are pertinent to two specific stereochemicalHere we report On the ('1 for is 2 S. Yamada, E. Ohno, I-, l<.;uge, A. Takeuchi, K. Yarnanouchl,K. H. Holm, G. W. Everett, and A. Chakravorty, Picogr. and K. Iawasaki, Co-ovdLnation Chrm. Rev., 1968, 3, 247.Tvatisitioiz-nl7rtal Chrw , 1968, 4, 199.Iizorg. Chem., 1966, 7, 83; R. H. Holm and 3T. J. O'Connor, ibid., L. Sacconi, Co-ordznation Chrm. lieu., 1966, 1, 192;1971, 14, 2411974 1691problems. (i) The four co-planar 4- tetrahedral equili-bria; and (ii) the formation of five- and six-co-ordinateThc nunibcring is for the identification of the n.m.r.spectra.complexes with donor molecules such as pyridine, as inequilibria (1).[NIL,; + 2B =+= [NiL,B] + B [NiL,B,] (I)During this study, as indicated in a preliminary r e p ~ r t , ~we have isolated for the first time the five-co-ordinatemono-pyridine compounds [XiL,B] (where B = pyridineor y-picoline) .,4 parallel, but less complete study, has been reportedby Yamada and his c o - w ~ r k e r s , ~ ~ ~ ~ 1 1 0 did not recognisethe five-co-ordinate species.ESPERIMESTALIn the general system of abbreviations used for the ligands,substitution on the aniline moiety is indicated by unprimednumbers and on the salicyl moiety by primed numbers.Thus, for example, (2,6-Me2-3'-NO,as) refers to the Schiffbase derived from 2,6-dimethylaniline and 3-nitrosalicylal-dehq.de.2,6-Dimethylaniline and 2,6-diethylaniline, and some ofthe salicylaldeliydes were obtained commercially ; but the3-methyl, 5-methyl, 3-chloro-, 3-bronio-, and 5-bromo-salicylaldeli ydes were prepared from the appropriate phenolby the Reimer-Tiemann procedure.6 The crude steam-distillates, containing an excess of the phenol, m-ere used forpreparing the metal compounds.Preparation of the Nzckel(I1) Covtzpounds.-The generalmethod of preparation of the bis-bidentate Schiff baseCompounds was similar to that of Yamada and co-worke~s.~We generally did not measure the quantities of reactantsaccurately, but mixed the components of the Schiff base inhot methanol, allowing a generous excess of amine, andthen added a saturated solution of nickel(I1) acetate inwater. Tlii5 was generally followed by NaHCO, (solid) toneutralise the mixture, which was then digested on a steam-bath for sel-era1 hours, and the green product was filteredoff.I t was usually recrystallised from chloroform, oftenrequiring light petroleum for reprecipitation. Where solidpyridine complexes were not formed (see Table l), a furtherpurification could conveniently be made by dissolving theproduct in pyridine and reprecipitating by the addition ofwater.The compound from 5-nitrosalicylaldehyde and 2,6-tli-methylaniline had a very low solubility in the varioussolvents employed here, and thus is largely excluded fromt sal = salicylaldehydate.D.A. Bone, E. D. McKenzie, and I<. Rowan, C h m . Coinm.,1970, 420.the physical measurements. It was obtained only in smallyield.From the preparations of both compounds of the unsub-stituted salicylaldehyde, different crystalline species wereobtained and were identified by their X-ray powder diffrac-tion patterns. A Table of X-ray diffraction data has beendeposited in Supplementary Publication hTo. Sup 20989.A suspension of [Xi(sal),(H,O),] t inethanol with an excess of 2,6-dimethylaniline at room tem-perature gave needles of a-[Ni(2,6-Me2as),] after a few days ;but they in turn were replaced by prisms of the less soluble[Ni( 2,6-Me2as) ,] ,0.5(C8HgNH,).(ii) When a mixture of the amine (2.65 g) and [Xi(sal),-(H20),] (3-72 g) was refluxed in ethanol (100 ml) for 7 h aconsiderable amount of the nickel compound remainedunreacted. An excess of amine was added, but after furtherheating (2 h) little further reaction had occurred.Themixture was halved : one half was filtered hot and this gave,on cooling, crystals of a-[Ni(B, 6-Me2as),] and the aniinesolvate [Ni(2,G-Me,as),],O-5(C8HgNH,) ; the other wastreated with pyridine (8 ml), and a clear solution formed.When this latter was cooled in the refrigerator (1 11) the a-form was obtained (0-7 g) ; but the filtrate, when set aside,soon deposited crystals of another species which we label as' @-[Ni(2,6-Me2as),] ' (0.8 g).From subsequent treatments of the filtrate we obtainedfurther quantities of the p-form, some of the amine solvateof the Schiff base compound, and two other products.Oneof the latter is identified as monoclinic [Ni(sal),(py),](Found: C, 62.4; H, 4.3; N, 5.9. Calc. for C,,H,,N20,Ni:C, 62-8; €3, 4.4; N, 6-17;. Space groupC, or C21c, a = 1447, b = 12.16, c = 14.05 A, p = 116.35".Mass spectrum peak a t m / e = 300 [Ni(sal),']}. The otherproduct is not fully characterised, but is apparently alsoan Ni(sal), species (Found: C, 63.7; H, 4 . 7 ; X, 6.3'3;.Mass spectrum peak at nz/e = 300. X-Ray powder diffrac-tion lines (20 for Cu-K, radiation): 0.3 (m), 13-15 (ms),13.9 (w), 15-1 (s), 17-25 (w), 17.9 (w), 20.7 (w), 31.1 (w),21-4 (m), 21.8 (m), 22-3 (in), - .Addition of water to pyridine solutions of a-[Si(2,6-Me,as),], the p-form, or the amine solvate gave only thea-form.The a-form could be recrystallised from CHCl,, butfrom acetone a monosolvate was obtained 1%-hich rspidlp lostacetone in the air.[Ni( 2, 6-Et2as) ,]. When a mixture of rNi(sal),( H,0),](4.0 g) and the amine (4-5 g) in ethanol (50 ml) was refluxedfor several days the product (6.5 g) was cr-[Ni(2,6-Et,asj2].A second (p) form was obtained when a pyridine solutionof the a-form was set aside to allow the solvenf to elvaporate.Such a product, however, contains significant ainoun ts of(undefined) hydrolytic material which give? a high-spinoctahedral species in pyridine. This led t o the initialerroneous conclusion that the p-form gave some octahedral[N(2,6-Et2as),(py),] in pyridii~e.~ It xas obtaincd in pcrerform by recry stallising from acetone.Recrystallisation from various solvents galre either formor mixtures as follows : CHCl,, mixture ; acetone, $-form ;pyridine, @-form by slow evaporation, and the a-form byprecipitating with water ; ethanol-pyridine (5 : l), an excessof the solid in contact with the solution a t reflux tempera-ture grew as large crystals of the a-form, but the @-formS.Yamada, -4. Takeuchi, I<. Yanianouchi, and K Ilvasaki,D. E. Armstrong and D. H. Richardson, J . Chem. Soc , 1933,[Ni(2,6-Me2as),] (i.)peff = 3.1 B.M:,30.55 (m), 32-25 (m) * a } .Bull. Chem. SOC. Japan, 1969, 42, 131.496; H. Wynberg, Chum. Rev., 1960, 60, 169n z h1.3N G > uar; x3 3II /I1.3 I-x 31974 1693crystallised out on cooling ; n-hexane, mixture (about equalamounts).Only one crystalline form each of [Ni(2,6-Me2-5’-C1as),1and [Ni(2,6-Et2-5’-Clas) ,] was obtained from the variety ofsolvents notecl above.We have not checked most of theother species for possible polymorphism, but [Ni(2,6-Me2-3’-XO,as),] crystallises from CHCI, as a monosolvate (of theplanar species, as proven by the diffuse reflectance spectrum)which soon loses the CHCI, in the air.Space groups and unit cell dimensions for a few of thecompounds, and X-ray powder diffraction patterns forothers are listed in Supplementary Publication No. SUP20989 (7 pp.).*Pyridine and Picoline Cow@lexes.-All the parentsalicylaldiminates were dissolved in pyridine and y-picolineand precipitated by the addition of water.The solidcomplexes so obtained are listed in Table 1. Where noaddition compound is listed for a particular salicylaldimin-ate, the parent species was recovered unchanged. Thefollowing salicylaldiminates also were recovered unchangedfrom or-picoline : [Ni(2,6-Me2-3’-Clas),l, [Ni(2,6-Me2-3’,5’-C1,as) J, “(2, 6-Me2-3’-Bras) J , [Ni( 2, 6-Et2-3’-Clas) ,I, [Ni-(2,6-Et,-3’,5’-Cl2as),1, [Si(2,6-Et2-3’-Bras),1, and [Ni(2,6-Et,-5’-N02as) ,I. Both the 3-nitro compounds gave intract-able oils, which could not be induced to crystallise.During the n.m.r. study in chloroform-pyridine it wasobserved that [Ni(2,6-Et2-5’-N0,as),1 went through asolubility minimum a t the 50 : 50 ratio of solvents, where thecompound is present largely as the five-co-ordinate species.This may give a useful method for isolating the latter, buthas not yet been explored.PJ~ysiccrZ Measuvements.-The following instruments wereused : Unicani SP 700 spectrophotometer for the electronicspectra, with the diffuse reflectance attachment (SP 735)for the solids; Perkin-Elmer 457 for the i.r.spectra (hexa-chlorobutadicne and paraffin oil mulls); Varian HA 100or Perkin-Elmer R12 for the lH n.m.r. spectra; A.E.I.MS 12 or RIS 9 for the mass spectra; Phillips 11-46 cmDebye-Scherrer camera with Co-K, or Cu-I<, radiation forthe X-ray powder diffraction patterns (samples in 0-2 or0.3 mni Lindernann glass capillaries).Magnetic susceptibilities in solution were measured byEvans’ lH n.m r. method.For the pyridine solutions,the solvent contained 107; rl,traniethylsilane. A sealedcapillary oi this mixture was used for the diamagneticreference signal.RESULTS AKD DISCUSSIONSteric Efccts of the Ligands.-In common with theconformation observed in all known crystal structureanalyses of N-aryl salicylaldiminates, the phenyl planeis expected to be essentially perpendicular to the salicyl-aldiminate plane. Indeed the alkyl substituents on thephenyl here make it imperative that this be so, althoughrotation from the perpendicular of up to ca. 30” should bepossible. In this conformation, the two ligands in (I)can adopt either of the two extremes: a four co-planaror a tetrahedral arrangement of the donor atoms. (Thesteric effects also will ensure that only the trans isomer of* For details of Supplementary Publications see Notice toAuthors No.7 in J.C.S. Dalton, 1973, Index issue (items less than10 pp. are sent as full-size copies).D. F. Evans, J . Chem. Soc., 1959, 2003; R. H. Holm,J . Amer. Chrvn. SOC., 1961, 83, 4683.the former exists.) Despite the comments of Yamadaand co-~orkers,~*~ there seem to be no really significantdifferences in the total steric effects in either configura-tion. The ligands, however, do provide a significantenergy barrier to the interconversion of the two con-figurations.However, for a substituent other than H in the3-position of the salicyl moiety, atom over-crowdingfavours the tetrahedral over the planar configuration,but does not preclude the latter.The bulkiness of the ligands provides a general inhibi-tion to the addition of other donor molecules such aspyridine.Thus the equilibria (1) will be expected to lieto the left.Finally, the steric effects should prevent oligomerisa-tion to octahedral paramagnetic species such as occursin wz- and 9- (but not o-) substituted N-aryl compoundsI \c I \15 10 .Wavenumber / 10 3cn:-!FIGURE 1 The electronic absorption spectrum of [Ni( 2,6-ikIe2-3’-NO,as),] in CHCl,, showing the band arising from the tetra-hedral species a t 7 lo3 cm-l(I), for which a tetrameric ‘ cubane ’ type structure isindicated, since the nitrogens cannot act as bridgingatoms. In common with the analogous 2-methyl-phenyl and 2,5-dimethylphenyl compounds,10 there isno association of [X(2,6-Me,as),] in chloroform down to240 K; and accordingly we neglect the possibility of sucholigomers in the further discussionFour-co-planar versus Tetvahedral Stweochemistry.-These compounds strongly prefer the four-co-planargeometry, with a diamagnetic groundThe electronic spectra show that [Ni(2,6-Me2-3’-NO,as),] and [Ni(2,6-Et2-3’-NO2as),] ,l but only these,give significant concentrations of the tetrahedral isomersin chloroform.[The latter are characterised by lowenergy absorptions at 7-43 lo3 cm-l (Table 2 and Figurel).] However, even these compounds crystallise out asdiamagnetic planar species (diffuse reflectance spectra,Table 2).13 P. C. Jain and E. C. Lingafelter, A d a Cryst., 1967, B23, 127;R.L. Braun and E. C. Lingafelter, ibid., p. 780; V. W. Day, M.D. Glick, and J . L. Hoard, J . Amer. Chem. SOC., 1968, 90, 4803.J. E. Andrewand A. B. Blake, J . Cheun. Soc. ( A ) , 1969,1456;J. A. Bertrand, A. P. Ginsberg, R. I. Caplan, C. E. Kirkwood,R. L. Martin, and R. C. Shenvood, Inorg. Chem., 1971, 10, 240.l o B. M. Higson and E. D. McKenzie, unpublished dataJ.C.S. DaltonBoth crystalline forms of [Ni(2,6-Et2as),] containplanar nickel(I1) species, as do the various solvates of thedifferent compounds isolated (Table 1). [Ni(2,6-Me2as),] ,-O.5(C8H,NH,) is isomorphous with the Pd analogue(X-ray powder diffraction patterns).The lH n.m.y. spectra of all compounds in CDC1, wererecorded and the details deposited in SupplementaryPublication No.20989. They give no evidence forcontact or pseudo-contact paramagnetic shifting of theresonances, even in the case of the two %NO2 compoundsfor which the electronic spectra show the presence ofquite significant amounts of the tetrahedral paramagneticisomers. This fits with the indication from molecularmodels of a significant energy barrier to the planar =+=tetrahedral interconversion, which is apparently suffi-cient to make the process too slow for the observation ofaveraged contact shifts. We shall report elsewhere themeasurement of a slow interconversion of the structuralisomers of [Cu(2,6-Me2as),].These spectra are generally unexceptionable andparallel, for example, those of our cobalt (111) salicyl-aldiminates, except for an upfield shift of 3-H (at 7 1.2-4.6) and the 3-Me resonances (at T 9.1, cf.5-Me at T cn.7.9). [The numbering system used for the aromaticprotons is shown in formula (II).J These shifts resultfrom the shielding effect of the ring current of the N-arylTABLE 3Bands appearing as shoulders are given in parentheses, and the The electronic spectra in the region 30-5 lo3 cm-1.extinction coefficients in square brackets(a) Diffuse reflectance and non-donor solventsCompound (11) QR = M e , X = H3’-Me.5’-Me3’-OMe3’-C15‘-C13’,5’-C1,3’-BrS’-Br3‘-pJ025I-NO25‘,6/-Benzo3’-OMe3’-C15’-C13’,5’-C1,3 ’ -Br5’-Br3’-NO,5’-NO,5’,6’-BenzoPhaseSolidCHC1,SolidCHCI,SolidCHCl,SolidCHCI,SolidCHCl,SolidCHC1,SolidCHC1,SolidCHCl,SolidCHC1,SolidCHCl,SolidSolidCHCl,SolidSolidCHCl,SolidCHCI,SolidCHCI,SolidCHC1,SolidCHCI,SolidSolidCHC1,CI-ICl,C6H6CE-ICI,22CHCl,SolidCHCl,SolidCHCl,SolidCHCI,27-6302829.3 [11,700]282829.0 [12,300]3030 [12,100]28.3“7.429.0 [10,500]29 [12,000]29 b> 302928.8 [13,000]28(38)25.5 b2728-4 [16,000]28.9 (28)28-329.4 [ 11,500]28.228.7 [10,300]>- 3028.029 I , [11,300]28.4> 30> 3038.529 [13,800]38.029-3 [15,000]28.7Absorption bands/103 cm-l(24.0)(24.5) [2900](23.9)(24.1) [3600](24) [3000](23.7)(24.2) [4300](24.0)(24.2) [3500](24) [3000](24.0)(24.1)(24.1) [4000](24)(23.1)(27.8)(24)(24.0)(24.0) [3400](24.1) [2900](23.8)(24.2) [3400](24) [3000](24.0)(24.2) [3200](23.8)(24.0) [4100](24.0) [3800]28-028.8 [11,500](27-0) [21,200]27.0 b(28.3) [2l,OOO]23.4 [13,000]d25.6 [23,000](24)22.522-6 [5000]22.322-7 [5800]22-322.4 14900122.322-7 [6350]22.422-8 15600122.222.5 [5100]22.322-4 [4900]22.722.7 [5700]22.122.5 [4600]22.8 c22-9 [STOO]23.622.028-6 [6500]22.432.65 [4500]22.522.7 [5300]22.322.3 [5100j9.7.322-7 [4800]22.682-7 [4900]22.122-4 [4800]22-322.3 r5300]22.4 c22.5 [5800]1-22.7 [5200]22.7 [5000]“I 699.1 c32.5 [5100]22.7 c22.922.5 [6400]21.5(21-5) [3620]21.2(21.9) [5000]21.1(21.5) [4200]21.1(21.9) [5000](21.7)21.0(21.5) [4200]21.2(21.5) [4200](22) [4800](21.5) [4000]20.5(21.5) [5000]21.3(21.7) [3700](21.7)(21.7)(21.3)(21.4) [4400]21.2(21-6) [3600]21.6(21.5) [4600](21-5) [4300]20.8( 2 1.3) [4000]21.1(21-4) [4700](21.7)20.8(21.3) 14300120.7(21.4) [5000]16.316.3 [loo]16.2(16.0) [82)15.6 [85]15.615-9 [130]15.715.9 [I50115.915.8 [90j16.216.15 [120]15.715.7 [85]15.815-65 [80]16.216-1 [I20116.2(16.4) [SOJ,7.1 p3j16.616.516.6 [I70116.115.615.6 [80]15-9(17) [lOO].15.9 [130]15.515.75 j130]16.715.5 [80]15.8 [66]16.116.1 [130]15.6155.6 [110]15.415.6 [90]15.7 [70]16.216.15 [I40116.115.8 [45],7.6 [6]16-816.7 [176]16-516-6 [160]16.25 cio0-1695 1974(b) Pyridines and picolincsCompound (IJ)I< = Ale, S = H3’-MC,5‘-MC!3’-03Ic3’-C15’-Cl3’,5’-C1,3’-Br5’-Br3‘-XQ25’. 6’-BtnZO1C = E t , S = H3‘-hlC5‘-JIc3’-031c3’-C15‘-C13‘, 5’-ClR’-Br5’-Br3‘--UC),5 ’ - 5 t.i,Base1’ yridi ney- PicolineP yrid i ney-PicolineP yri d i n ey-PicolineI’yridinc7-PicolineP yridiney-Picolinea-PicolinePyridiney-PicolinePyridiney-Pi coli nca-PicolinePyridiney-Picolinca-PicolineP yridine3) -Pic ol in ePyridincy -Picolinea-PicolinePyridiney-PicolinePyridiney-PicolinePyridiney-PicolinePyridiney-Picoli nePyridiney-PicolinePyridiney- Pi colinea-PicolincPyridiney-PicolinePyridiney-Picolixiex-PicolincPyridincy-Picolinex-Picoli xieI’yridiney- Pi colincPyridiney-Pi coli nerx-PicolincPyridine7- Picolinea-PicolinePJrridincy-PicolincTABLE 2 (Continued)Absorption bandsA -- r-(4)16.2 [lo4116.2 [70]15.8 [53]15.7 [60[16.1 [72]15-8 r27]15-9 [27]15-76 [22]15.8 [23]15.9 [60]16.0 [56]16.0 [59]15.6 [20]16.0 [20]15.7 [50]15.6 [14]15.6 [20]15.7 [35]16.8 [62]16.0 [54]16.1 ~1401gg16.4 [36]16.6 [135]16.7 [I130116.2 [IOO]16.2 [115]15.7 [GO]15-7 [50]1G.O [go]15.7 [24]15.8 [27]15.8 [23]15.8 [24]15.9 [70]15.9 [63]16-2 [70j16.0 [21]15-6 [55]15.6 [17]15.7 [19]15.5 [45]16.1 [110]16.0 [22]16-0 [a4116.1 [65](1 6.5) [20](16.1) [35]16.1 [28]16.7 [63]16.6 [I641fi16.4 [23]16-7 [ n o ](6)9.6 [3] f9.7 [5]9.6 [2] f9.2 [13]9.3 [13]9.3 [IS]9.3 [12]9.7 ~ 5 1 f9.4 [15]9.3 [15](9) PI’9.6 [6] f9.6 [7] f9.8 [18]9.9 [20]9.6 [5] f9.2 [15](22P[;2](9.2) [lo] f(9.2) [3] f(9.2) c419.0 [12]9.0 [13]9.0 [14]9.1 [13](9) [2l‘;:42’[;1;9.2 [15]9.2 [15]9.0 [15]9.1 [13](9.2) [3] f9.2 [5] f9.9 [14]9.7 [14](9) PI f(9.2) f9.8 [ l l j9.9 [15j9.7 [I41(5) ‘5.1 [lo!5-3 [12]5 [2615.2 [28]5.1 [7]5.1 [8]65-6 ClS]5.7 [16]5-8 [S]5.3 [31]5.2 [29j6 PI5.5 [7]6.0 [lo]5-9 [i5j5.7 [18]5.9 [15]5.2 [36]5.3 [28]6.3 [30]5 [415.4 [12]5.4 [20]5.3 [20]5.6 [4j6 1415.2 [8]5.1 [ 5 ]G.0 [15]6.0 [15]6.1 [8]5.3 [20]5.3 [20](5.7) [GI,6.2 [10j6.2 [23]6-2 [14]6.3 [15]6.2 [24]5.3 [18j5.6 [2oj6.6 ilSj6.0 [ l l i5.9 [17j?(ca. 5 ? ) [<3]Other bands(12.7) [4j(12.7) [5](12.9) [6](12.6) [6](12.9) [GI(12.9) PI12.7 [SP, 71(12-7) Csp, 81(12.7) P I(11.6) 1131(12.6) [7](12.7) [5](12.7) [7](12.7) [lo]12.7 [sp, 7112-7 [SP, 71(12-7) [lo](12.5) [6]12.7 [ 2 ](12.8) [5](12.8) [12j12.6 13112.6 [SF, 71(12.8) [5](12.7) [Sj11.6 [sp, 3111.6 [sp, 5111.5 [sp, 21(11.6) [5111 -5 [sp, 6111.6 [sp, 7111-6 bp, 6111.6 [SF, 41(11.5) [4j(1 1.6) [sp, 3111.6 [SP, 4111.6 [SP, 6111.6 [SP, 71I l .6 [sp, lo]11.6 [4]11.6 [SP, 31I I .A [SF, 4111.6 [sp, 61l J .6 [sp, 6111.5 [SP, 4111.5 [sp, 6j(11-4) [5](11.5) [6]11.6 [ ~ p , 71(11.5) [SP, 6111.6 [ ~ p , 6111.6 [sp, 6111-6 [sp, 4111.6 [SP, 6111.5 [sp, 10111.6 [sp, 43(11.6) [sp, 81Substitutioii on the salicyl moiety is indicated by primed numbers. Broad band. Xsyninietric to high energy. BroadHere we tabulate in separate colurrins the data which give evidence for thef The five-co-ordinate species also have a low intensity band a t ca. 9 lo3 cm-’, soThe band here isunrcsolvtd absorption between 30 and 20 10:’ cxn-l.co-ordination number \vhich heads the column.there is oi‘tcn an ambiguity as to the possible contribution of a small amount of an octahedral species here.masked b ~ . li;=and or metal-ligand absorptions.substituent (of the opposite ligand) and confirm thattlie latter is essentially perpendicular to the salicylmoiety.The spectrum of [Pd(2,G-bIe2as),] was similar to thatof the nickel analogue.Stcrcoch!~iiiistrgl in Donor Soleleizts.-In pjridine andtlie picolines these jNiN,O,1 compounds form two para-magnetic complexes as defined by the equilibria (1).Some, but not all, of these can be isolated as crystallinesolids, sometimes as mixtures (Tables 1 and 3).Various physical measurements give data on theextent of the solution equilibria.(i) The total amountof both paramagnetic species is defined by the magneticsusceptibilities (Table 4). (ii) The electronic spectragive good evidence for the species present and for theirstructures, but have not yet been refined t o give quanti-tative data on the equilibria.(iii) Paramagnetic shiftsof the lH n.m.r. spectra also prove that more than oneparamagnetic species is formed.The two high-spin compounds are seen clearly in th1696 J.C.S. DaltonTABLE 3A suiiiinary of the structural data for the complexes with pyridine and the picolinesSolution data 0:& Totalcomplex Species identified Estimated amounts of theCompound (11)” (from X,) b from spectra c various complexes Solids isolated *444444R = M e , X = H (py) 33 4 + 5 5 (35) 4 (65)(Y-PiC) 4 I 5 (+6) 5 (40) 4 (60)3’-Me (PY). 60 4 + 5 5 (60) 4 (40)5’-Me (PY) 28 4 + 5 5 (28) 4 (72) 4(Y-PiC) 4 + 5 (+6) 5 (30) 4 (70)3’- OMe (PY) 100 (4) + 6 ti (100)Y-PlC) (4) + 5 + 6 5 ((5) G (>95) 43’-C1 (py) 100 5 + 6 6 (60) 4(Y-P? 5 T 6 5 (40)(E-PlC) 4 + 5 5 (20) 4 (80)4 (20)4 (20)5/43 (py) 83 4 + 5 ( t 6 ) 5 (80)(Y -Pi c) 4 + 5 (+6) 5 (80)3’,5’-C1, (py) 100 5 -i 6 5 (10) 6 (90)(Y-PiC) 5 + 6 5( 15) 6 (85)(a-pic) 4 + 5 5 (60) 4 (40)3’-Br (py) 100 5 T 6 5 (35) 6 (65)(Y-PiC) 5 + 6 5 (40) 6 (60)(a-pic) 4 T 5 5 (60) 4 (40)5’-Br (PY) 77 4 + 5 5 (80) 4 (20)4 (20) (Y-PiC) 4 + 53’-NO, (PY) 100 6(a-Pl4 4 4- 5 (t-6) 5 (75) G (5) 4 P O + .) &f5’, 6’-Benzo (PY) g 4 (+5) 5 (<lo) 4 (>go)5 ( < l o ) 4 (>go) 45 (20) 4 (80)4 (76) (Y-PiC) 4 + 5 5 (25)3’-Me (PY) 48 4 + 5 5 (50) 4 (50)(Y-PiC) 4 + 5 5 (55) 4 (45)5’-Me (PY) 13 4 4- 5 5 (13) 4 (87)(y-pic) 4 + 53’-OMe (PY) 100 6 6 (100) 4(Y-PiC) 6 6 (100) 43’-C1 (py) 100 5 4- 6 5 (35) 6 (65)(Y-PjC) 5 + 6 5 (35) 6 (65)(X-PlC) 4 + 5 5 (20) 4 (80)5’-C1 (py) 60 4 + 5 5 (60) 4 (40)(Y-PiC) 4 + 5 5 (60) 4 (40)6 (85) 3’,5’-c12 tpy) 100 5 + 6 5 (15)(Y-PiC) 5 + 6 5 (25) 6 (75)(%-PIC) 4 + 5 5 (60) 4 (40)3’-Br (py) 100 5 + 6 5 (35) 6 (65)6 (65) (?+pic) 5 + 64 (50)(a-pic) 4 + 5 5 (60)5‘-Br (py) 63 4 + 5 5 (50)(Y-PiC) 4 + 5 5 (50) 4 (50)3’-N02 (PY) 100 6(Y-PjC)6 (55)5‘-NO, (PY).100(Y-P? c) 5 + 6 5 (20) 6 (80)4 (55) ( U-PlC) 4 1 55’, 6’-benzo (PY) g 4 (+5) 5 ((5)(Y-Plc) 4 - 5 (1-6) 5 ( G O + ) 4 (40-)5 + 64445 + 65 + 646ti44466(40) 6 (60)t! [%)(Y-PjC) 6 6 (100)444444(Y-PiC) ; ( + 5 )R = E t , X = H (py) 20 t 55 (lo+) 4 (90-)554446646 f - 6G44466f56444(35) 4 (40)6 (100)6 6 (100) : {%)(45) 4 (>95)(&-PIC) (;+I 5 + 6(?’-Pic) 4 4 (>96)(I Results refer to ‘ room temperature ’ (18 & 3 “C).b These were calculated from the data given in the Supplementary publicationParenthesesd These figures are rough estimates, basedAnThese are the co-ordination numbers of the solids isolated wheng Too low to be measured, butusing a figure of 4150 as the m’ for 100% paramagnetic.imply an ambiguity about the presence of the particular species (see Discussion section).on both the magnetic data given in the Supplementary publication and the intensities of the various bands in the spectra.of 40 seems to be about right for the five-co-ordinate species.water is added to the solutions in the various solvents.the presence of a small amount of the five-co-ordinate species is detected from the IH n.m.r.spectra.Numbers are the co-ordination number at the nickel.The products (see e) were intractable oils.* Substitution on the salicyl moiety is indicated by primed numbers1974 1697electronic spectra, such as that of [Ni(2,6-Et2-3’-Bras),]in pyridine (Figure 3). The bands at ca. 9 and 16 lo3cm-l are quite characteristic of a six-co-ordinate [NiL,B,]species; whilst that a t ca. 6 lo3 cm-l is a prominentcharacteristic band of the new five-co-ordinate speciesi 5 10 5’The diffuse reflectance spectra of: a, [Ni(2,6-Me,-5’-Bras) ,] (four-co-ordinate) ; b, [Ni(2,6-Me,-3’-Clas),(py)] (five-co-ordinate) ; and c, ~Ni(2,6-Me2-3’-Bras),(y-pic),] (six-co-ordinate)Wavenumber / 10 3crn-’FIGURE 2TABLE 4The magnetic susceptibility data for pyridine solutions(ca.20 g 1-1)XM XM’1272 14532249 2458746 9553867 40843942 4155322 1 34343676 391 73897 40883858 40902949 318161.3 8461754 2010304 5603932 41963874 41342231 24923813 41013863 41013951 41893879 41592333 2613fzaPeff1.882-441 *523-143.172.883.083.153-152-78I -432-211-163.193.1 c,2-463-153-153.1 83.182.52Thc lxiraiiiagnetic signal was not resolved from the dia-niagnctic 011c.which MY define here. The full spectrum of the latteris best seen in the diffuse reflectance spectrum of, forexample, [Si(2,6-Et2-3’-Clas),(py)] (Figure 2), showingthat this species also has a band at ca.16 lo3 cm-l (asdoes any residual diamagnetic four-co-ordinate speciesthat may be present). Thus the bands at ca. 10 andca. 6 lo3 cm I , and only these, can be used to detect thepresence of significant amounts of the six- and/or thefive-co-ordinate complexes, respectively. They cannot1~ used to dcfine exactly the equilibria (1) for we do nothave any information on absolute extinction coefficientsin these solutions; and there is some ambiguity in thedetection of small amounts of the six-co-ordinate speciesin solutions containing mainly the five-co-ordinatespecies, since the latter appears to have a weak band atcn. 10 lo3 cm-1 (Figure 2).Still some rough but usefulestimates of the relative amounts of each can be given(Table 3).These assignments of the spectra to five- and six-co-ordinate species are further confirmed by the spectra ofFigure 3 (obtained at different temperatures). Whenthe solutions are heated, the equilibria (1) are displacedto the left, so that the five-co-ordinate species becomesmore prominent.Our study of the paramagnetic shifts of the lH ? ‘ I . ~ Z . Y .spectra was similar to La Mar’s study12 of the com-pound (I), for which R = Et. By contrast, in oursystems, the five-co-ordinate species persist over a muchwider range of pyridine concentrations. Figure 4 showswhat happens when drops of pyridine (deuteriated) areadded to 10 ml of a 0 . 0 5 ~ solution of [INi(2,6-Et2-3’-Bras),] in CDCl,.The various signals move up or clown-field in the order expected for a contact-shift mechan-ism .l2In this range of low pyridine concentrations theparamagnetic shifts change monotonically with pyridine,i I15 10 5Wave n urn be r / 1 0 crn-’FIGURE 3 The electronic absorption spectra of [Ki(2,0-1CtZ-3’-Bras),] pyridine a t 20, 30, 40, and 50 “C, as labelled, showingthe equilibria between the five- and six-co-ordinate speciesbut, as the latter is further increased, the shifts are nolonger monotonic and some of them change direction.Figure 5 gives a plot of signal us. pyridine concentrationfor [Ni (2,6-Me2-3’-Bras) ,] and [Ni (2,6-E t ,-5’-h’O,as)up to the present effective limit of observation.[Wehave stayed within the range +10 to -20 p.p.m.(w.r.t. tetramethylsilane), and some signals broadenmarkedly and are lost in the background noise at higherpyridine concent rations. jThe paramagnetic shifts appear qualitatively to beof the normal contact shift type. An exception, howT-ever, is apparent in the case of the methylene protons ofthe ethyl substituted compounds. Figure 4 showsD. Cummins, B. Sf. Higson, and E. D. McKfienzie, J.C S.Dalton, 1973, 414 and 1359.l2 G. N. La Mar, I>zot,g. Chiitz. ,4cla, 1969, 3, 1831698 J.C.S. Daltonclearly how the initial quartet collapses into two inde-pendent signals, which move down-field at differentIChemical s h i f t /p.p.rn. from tetrcmethylsilaneFIGURE 4 The lH n.m.r. spectra of [Ni(2,6-Et2-3’-Bras),1 inCDCl, (10 ml) showing the paramagnetic shifts which ensuewhen drops of C,D,N are added as follows: a, no pyridine;b, 1 drop; c, 3 drops; and d, 5 drops100f 60.-H 1 l0IIIIIIIIIII0b\1\I2ok 0rates (Figure 5 ) as pyridine is added.This contrastswith the behaviour of the methyl protons of the 2,6-dimethylaniline compounds, and we currently believethat it results from restricted rotation of the -CH,Rgroups, making the two methylene protons non-equiva-lent in the contact shift mechanisms. Further experi-ments, including the use of differently substitutedanilines and an X-ray analysis of a five-co-ordinate solid,should help to distinguish between such a mechanismand an alternative pseudo-contact one.Such lH n.m.r. data also are the most powerful in de-tecting small concentrations of the paramagnetic species.At the concentrations available, we cannot use the elec-tronic spectra or the magnetic susceptibilities to detectunequivocally high-spin complexes of the compoundsderived from I-hydroxy-2-naphthaldehyde ; but theshifts of some of the lH n.m.r.resonances of these com-pounds in pyridine make it clear that small concentra-tions of [NiL,B] are indeed present.Polymov+hism.-Dimorphism and the formation ofcrystalline solvates are observed, but appear to beuncommon here. We shall discuss elsewhere a greatervariety among the cobalt and copper analogues. Thevarious species characterised are given in the Experi-mental section and in Table 1.Special comments arenecessary only for ‘ p-[Ni(2,6-Me2as),] ’. Mass spectraprove the presence of [Ni(2,6-Me2as),] and the productcan be converted to the a-form by dissolving in pyridineand adding water. However, the X-ray molecularweight of 460 is low (507.3 for [Ni(2,6-Me2as),]}; and theelectronic spectra (diffuse reflectance) and magneticsusceptibility (XM = 2093, peff = 2.2) suggest a mixedcrystal of diamagnetic [Ni(2,6-Me2as),] and some as yet0IIIIIIIIII0Paramagnetic s h i f t / p.p.m. from tetramefhy1s;LaneFIGURE 6 Plots of Paramagnetic shift vs. pyridine concentration (py/CI)Cl, mixtures) for the various resonances in [Ni(2, 6-Me2-3’-In each case, H7 moved to lower field too Bras),] (full lines -0.04 g/ml) and [Ni(2,6-Me2-5’-NO,as),1 (broken lines -0.06 g/ml).soon to be representable on this Figure, as did H, of the forme1974undefined high-spin octahedral nickel@) compound-perhaps a salicylaldehydato-species.Substituegzt Efects and Stereochemistry .-The variousdata pertaining to the five- and six-co-ordinate speciesare summarised in Table 3. Higher co-ordinationnumbers are favoured in the following orders: (i) Me >Et for the aniline substituent; (ii) NO, > OMe > Br 3C1 > Me 2 H > 5,6-benzo for the salicyl substituents(in positions 3 or 5 ) ; (iii) 3-substitution > 5-substitu-tion; and (iu) y-pic 2 py > ct-pic.Bulk steric effects (R-strain 13) undoubtedly accountfor both (i) and (271); and (ii) can be rationalised, atleast partly, in terms of the electron-withdrawing powerof the ring substituent. The greater the latter, thelower the electron density at the metal atom, whichshould favour the higher co-ordination numbers (Paul-ing’s electroneutrality principle). Yamada and co-workers 2 9 5 prefer to argue in terms of the ligand-fieldstrength of the salicylaldiminates, but the two rational-isations are not mutually exclusive. There is no im-mediately obvious explanation for the observations (iii),but they may reflect the entropy part of the total freeenergy.We are currently looking further at this problem of thesu bstituent effects, using several series of compoundswith the same substituent at positions 3-, 4-, 5-, and 6- ofthe salicyl ring.We are indebted to Messrs. D. A. Bone, A. Cox, and K.Rowan for some of the preparations, and to Dr. B. E. Mannfor helpful discussions.[3/2386 Received, 21st November, 19731l3 J. G. Gibson and E. D. McKenzie, J . Chem. SOC. ( A ) , 1971,1666