摘要:

1979 423Unstable Intermediates. Part 183.l An Electron Spin Resonance Studyof Radicals formed in Irradiated Sodium and Potassium NitriteBy Martyn C. R. Symons, Department of Chemistry, The University, Leicester LE1 7RHExposure of K[NO,] to 6oCo y-rays at 77 K gives approximately equal yields of NO,' and [N0,'I2- radicals, bysimple electron transfer. On annealing a t >77 K the e.s.r. features characteristic of these radicals are lost, largelyby recombination to give [NO,]- ions. However, relatively low yields of a two-nitrogen centre previously describedas [N,O,-] are then detected. This is stable at room temperature, and is exclusively formed on exposure at thistemperature. Similar results have been obtained from Na[NO,], except that the parallel features for [N0,I2- areextremely broad and hence identification is tentative.These results are discussed in the light of previous con-clusions. In particular, it is suggested that the two-nitrogen species is probably [N2O4I3-. The structure of thisnovel radical is discussed qualitatively. When the nitrites are doped with nitrate ions in low concentration thebehaviour at 77 K remains unchanged, but, on annealing, [NO,]2- ions are formed in addition to the two-nitrogencentre. This result confirms the primary formation of [N0,I2- in the sodium salt and shows that the electronaffinity of [NO,] - is greater than that of [NO,] -THE effect of high-energy radiation on nitrates andnitrites both in 'the crystalline state and in solution 3has been very widely studied. E.s.r.measurementshave shown that, at low temperatures, the results forsodium and potassium nitrate are remarkably simple,involving only electron ejection followed by electroncapture [equation (l)]. The radicals formed are trapped2[N03]- --.j NO,' + [NO3*l2- (1)at 77 K presumably by self-distortions. Thus, forexample, [N0,'I2- is pyramidal whereas [NO,]- isplanar.Zeldes andLivingston, irradiated Na[NO,] at 77 K, but detectedonly NO,. Since no electron-gain centre was detectedone must conclude either that its resonance was toobroad to detect, or that some mechanism operates toselectively convert the electron-gain centres into dia-magnetic products. Another, but less probable explan-ation is that these are converted by some mechanisminto NO, radicals.These workers found that on anneal-ing above 77 K the NO, radicals were lost irreversiblyand were not replaced by other detectable paramagneticspecies. In contrast, two groups studied the productsof room-temperature radiolysis of Na[NO,] and observedtwo distinct products, one being identified as NO, andthe other, which contained two equivalent 14N nuclei, as[N,04]-.5*6 These NO, radicals differed from thosedetected at 77 K in orientation: those at 77 K werenearly axially symmetric with respect to the b crystallo-graphic axis, as expected for loss of an electron from[NO,]- with no orientational change; however, thoseformed at room temperature had the a axis as the axisof near symmetry, implying a 90" r ~ t a t i o n .~ , ~ Identific-ation of NO, was supported by the observation of a broadelectronic transition in the 400 nm regionJ5 since NO,in the gas phase absorbs strongly in this region.I was prompted to re-examine these systems because ofour results for irradiated N,O4.' In this study a centrecontaining two 14N nuclei was well defined and wasidentified with some certainty as the anion, [N20,]-.The 14N hyperfine coupling (.Amax. 56.4 G 7 ) was far7 Throughout this paper: 1 G = T; 1 rad = J k g l .Results for the nitrites are less clear cut.greater than that previously assigned to [N,O,]- (8 G ) , 6 p 6and hence I felt that the previous identification waswrong. This conclusion is strongly supported bytheory, which suggests that the lowest empty orbitalfor N,O, is the 6b1, G* orbital.s This assignmentaccords with our results,' but not at all with thosepreviously assigned to [N,0,]-.5p6 Another aim of thepresent work was to endeavour to understand theapparent absence of paramagnetic electron-gain centresin the studies both at 77 K (ref.4) and room temper-a t ~ r e . ~ * ~EXPERIMENTALAnalaR grade Na[NO,] and K[NO,] were used as sup-plied, since recrystallised samples from purified water didnot give different results. Freshly supplied salts were usedto minimise the nitrate level. Aged samples gave [NO3I2-on irradiation at room temperature, showing that nitratewas formed on storage. Samples doped with nitrate gavesimilar results.Samples were irradiated as very fine powders at 77 K orroom temperature in a Vickrad s°Co y-ray source at a doserate of ca.2 Mrad h-l for between 0.1 and 2.0 h. E.s.r.spectra were recorded on a Varian E3 spectrometer at77 K. Samples were annealed in the insert Dewar afterdecanting the nitrogen, and were re-cooled to 77 K wheneversignificant changes in the e.s.r. spectra were observed.RESULTS AND DISCUSSIONIrradiation at 77 K.-Exposure of K[NO,] at 77 Kgave the e.s.r. spectrum shown in Figure 1. Thisdisplays very narrow features assigned to stationaryNO, radicals (Table) together with intense ' parallel 'features separated by 2 x 36 G, clearly assignable to[NO,],- radicals. The ' perpendicular ' features forthis radical overlap with the I = 0 features for NO,and so no reliable parameters could be obtained.How-ever, the parallel features are quite distinctive and thehyperfine coupling compares well with previous reliableestimates9 The concentrations of NO, and [N0,]2- are,in fact, comparable: that for [NO,],- appears to berelatively low because of the far greater anisotropy in thehyperfine coupling which arises because [NO,],- is a xradical whilst NO, is a cf radical.1° Thus in this case424 J.C.S. Daltonas with the nitrates, a simple electron transfer has One aspect of the spectrum for NO, in Na[NO,] isoccurred [equation (2)]. For Na[NO,] the features for noteworthy, namely that the outermost (2) features arefar broader than the x and y features (Figure 2). A 2[NO,]- -+ NO,' + [N0,'I2-second-derivative spectrum was best analysed in termsNO, were also well defined (Figure 2) and the data are of an extra triplet splitting, although the definition wasagain in good agreement with expectation (Table).In poor [Figure 2(b)]. This seems to rule out the possibilityof hyperfine coupling to 23Na. Also, such interactions 1 with neighbouring cations are usually almost isotropicbecause of the dominance of electron transfer into theouter s orbitals of the cations, whereas the presentinteraction is clearly extremely anisotropic. I tenta-tively suggest a very weak coupling to a neighbouring[NO,]- ion. Because of the large difference in bondangle between NO, (ca. 134") and [NO,]- (ca. 120°), itseems reasonable to postulate that an NO, [NO,]-pair could remain quite unsymmetrical on the e.s.r.time scale.(The relative orientations are wrong for theformation of [N,O,]-.')"0212;- J(NO~)(O) Annealing at >77 K.-On annealing, the spectra forz(No2)(-') because of the onset of significant librational or rotational(2)32506 (9-1075GH2)mo,) (0)I I (O)/ i both salts initially broadened reversibly, presumablyNW2) 6FIGURE 1 First-derivative X-band e.s.r. spectrum for KrNO,] motion. Later, there was a marked decrf3se in inten-after exposure to O0Co y-rays at 77 K for 0.1 h, showing featuresassigned to NO,' and radicalssity, the loss of NO, being greater than that of [N0,]2-.However, as the features for NO, and were lostthis case, no well defined parallel features for [NO,],- canbe seen. There is, however, distinct curvature in thea new signal was detected (radical A) (Figure 3).Theseoverlap with the M I = 0 component for NO,, but areE.s.r. data for radicals trapped in K[NO,] and Na[NO,]14N Hyperfine coupling/(;A , A" A ,50.0 47.6 68.049.0 46.4 67.536.0 ca. (-) 2 ca. (-) 237.2 (-) 4.3 (-) 4.38 6 28 b b8 2 251.9 47.1 56.464 30 3062 30 3061.4 30.2 30.2Luz, J . Magnetic Resonance, 1975,Aim.55.254.39.55.3ca. 11451.841.340.740.6gx g, gz2.005 7 1.992 2.002 02.005 7 1.992 2.002 02.002 0 ca. 2.007 3 ca. 2.007 32.002 9 2.007 1 2.007 82.005 0 2.007 5 2.006 22.005 0 b b2.005 5 2.004 7 2.014 22.003 1 1.991 5 2.003 01.997 6 2.004 1 2.004 118, 358. Unresolved. Ref. 5: identified as [N204]-spectral regions expected for the I = &l parallel lineswhich I feel fairly confidently are due to [NO,],-.Thision is a powerful electron donor and, since it is adjacentto four Na+ ions which occur in three magneticallyinequivalent sites with respect $0 [NO,]-, I postulatethat the spectrum should comprise sets of features fromhyperfine coupling to two equivalent 23Na nuclei (sevenlines) and two other inequivalent nuclei (eight lines).These 56 lines could well overlap to give the very broadfeature detected. The smaller magnetic moment for39K would account for the far narrower lines obtainedwith K[NO,]. Thus I postulate that, for both salts,reaction (2) dominates at 77 K. This result ties inexactly with our studies of aqueous and methanolicglasses containing [NO,]- ions, the primary productsagain being NO, and [N0,I2-.l1 However, these studieswere complicated by the tendency for [NO,],- ions toprotonate to give [HNO,]- ions.much narrower and hence can be located with reasonableaccuracy.The e.s.r. spectra obtained after exposure atroom temperature were similar to those obtained onannealing. Samples containing nitrate ions gave radicalB in addition to A, both sets of features appearingsimultaneously (Figure 3). Features from NO,' weresuppressed or hidden under these conditions.Radical B . Features for B were much broader forthe sodium than for the potassium salt. This againsuggests the presence of unresolved superhyperfinecoupling to a group of 23Na nuclei. For species A in thepotassium salt quite accurate parameters could bederived from the outer features (Table).These agreereasonably with those reported for Na[NO,] afterexposure at room t e r n p e r a t ~ r e , ~ * ~ if one allows for somelibratory averaging of the room-temperature spectra.However, the data are not characteristic of NO,' radicals,and I am puzzled by the previous identification of th1979 425species as N0,'.S*6 In my view, the species is [N0,'I2-, ascan be judged by comparison with data for this radical325bG (9.1 I83 GHz)i z (NO213200G t 9 4 113 GHz)FIGURE 2 (a) First-derivative X-band e.s.r. spectrum forNa[NO,] after exposure to 6oCo y-rays a t 77 K for 0.5 h,showing features assigned to NO,' and, possibly, [N0,l2-radicals.(b) A second-derivative spectrum of the ( f 1) NO,features showing possible splitting of the broad z componentobtained from nitrate ions (Table). The fact that B wasnot obtained when fresh high-grade samples were usedconfirms this identification and suggests that impuresamples were used previously; [NO3*I2- radicals werepresumably formed from [NO,'],- [equation (3)]. Thefact that [NO,'],- ions were formed on annealing doped13250 G (9 4038 GHz) nFIGURE 3 First-derivative X-band e.s.r. spectrum for KCNO,]after annealing to ca. 200 K and re-cooling to 77 K, showingouter features (B) assigned to [N0,'I2- radicals and centralfeatures (A) assigned to [N,O4I3- radicalsNa[NO,] samples strongly supports the initial formationof [NO2'I2- ions.We have previously shownll that[NO,]- + [N0,'I2- - [NO,'],- + [NO,]- (3)[NO,'],- ions can be generated from nitrite ions inaqueous alkaline solutions [equation (4)], but clearlynitrate ions are responsible in the crystalline salts.[NO,]- + 0-' ----t "Oil2- (4)If [NO,'],- is formed from nitrate ions, the postulatedmovement of the radical thought to be NO, is un-necessary (see Figure 4). As stressed above, loss of anelectron by nitrite gives NO,' radicals having theirmagnetic axes along b. However, gain of an extra0 00 NaO NFIGURE 4 Crystal structure of Na[N02] a t room temperatureelectron by [NO,]- in a nitrite site gives [N0,'I2- whosemagnetic axes are along a, as required by the single-crystal data.426 J.C.S. Daltoncrystal studies (probably associated with g, for [N,0,]3-)is unexpected, since it is apparently greater than thatfor the parent [NO,],- ions (Table).If my analysis ofthe powder spectrum is correct, the results are morereasonable as can be judged by comparing the data for[N,0,I3- and [NO,],- in the Table. In fact, Aiso. isclose to half the value for [N0,I2- and the smaller, non-axial, anisotropic component is understandable in termsof the tilted structure (2). I am therefore confident thatthe [N,O,I3- postulate is consistent with the data.If this is accepted, it would seem that the tendency for[N0,I2- to form [N,0,I3- is greater than that for NO, toform [N,O,]-. This is probably a function of therelative orientations of the parent [NO,]- ions, which areideally placed for [N2O4I3- formation (Figure 4) but forwhich a relative reorientation of 180” is required for[N,O,] - formation.The fact that is formed from [NO,],- onannealing, and is then stable at room temperature, showsthat electron transfer between [NO,],- and [NO,]- isrelatively easy, and that [NO,]- has a greater electronaffinity than [NO,]-.Had radical A been [N,O,]-,doping with [NO,]- should have increased its yield.In fact, the yield of A was not clearly changed ondoping. This shows that it was not formed from [NO,]-impurities, but sheds no light on the alternative [N,O,]-and [N,0,I3- formulations.Centre A. This centre is clearly the same as thatpreviously thought to be [N,O,]-. The quintet at 8 G(2) is reasonably well defined for both salts, but the xand y features are very difficult to extract, almostcertainly because the magnitude of the 14N hyperfinecoupling is small and significant quadrupole shifts areinvolved.Unfortunately, in the single-crystal studies,spectra near the x andy axes were so poorly defined thatno data are reported, the parameters being derived by anextrapolation. The resulting hyperfine data (Table)must be viewed as being only approximate. The moststriking difference between my results and those pre-viously reported is in g,. The crystal value was extra-polated, but the powder spectral analysis is unsatis-factory in some respects. Nevertheless, I am satisfiedthat the two species are the same.If we have correctly identified [N,O,]- in irradiatedN,O,, then centre A cannot be [N,O,]-.Also, if thision has the expected planar structure (l), the observed14N coupling constants for centre A are far too small.The only obvious alternative is the unknown radical[N20,l3-. The ion [N,0,I2- does not exist. However,the [N2O4I3- ion probably has structure (2) and, sinceX( 1 1 ( 2 1the N-N bond is weak, electron loss will probably befrom the G orbital. If so, this ion is another example ofthe relatively rare class of o1 radicals. Other examplesinclude [Ag,] +,12 [(MeO),B*B(OMe),]-,13 and [N,0,]+.8I t might be supposed that the observed 14N couplingconstants for [N,0,I3- would be approximately half thosefor [NO,I2- since the structures of [NO,],- and [NO,]-are expected to be similar, and the unpaired electron isequally divided between the two NO, groups.In fact,A,I(l4N) for [N,0,]3- (centre A) is considerably less thanAll for [NO,],- (Table). The simplest explanation forthis result is that the planes of the two NO, groups aretilted in the manner shown in (2). (This is expected byanalogy with the structure of [S,O4l2- ions.14) If the2pT orbitals on nitrogen remain almost pure 29 in thedimer, then the experimental All values will be reducedbecause the parallel field is not along either p orbital.This would result in a concomitant increase in the 14Ncoupling along the x axis. This is indeed suggested bymy interpretation of the powder spectrum. The largepositive g shift (gobs. -2.002 3) obtained from theI thank the S.K.C. for the award of a grant, and Mrs.[8/1045 Received, 5th June, 19781V. K. Thompson for experimental assistance.REFERENCESPart 182, J. Chem. Research ( S ) , 1978, 358.J. Cunningham, 5th Internat. Symp. Free Radicals, Uppsala,1961; K. Gesi and Y. Kazumata, J . Phys. SOC. Japan, 1964, 19,1981; R. Adde, Acad. Sci. Paris, Ser. C , 1967, 264, 1905; A.Reuveni and 2. Luz, J . Magnetic Resonance, 1976, 23, 265.L. Kevan, J . Phys. Chem., 1964, 68, 2590; 1965, 69, 1080;P. B. Ayscough and R. G. Collins, ibid., 1966, 70, 3128.H. Zeldes and R. Livingston, J . Chem. Phys., 1961, 35, 563.J . Tateno and K. Gesi, J . Chem. Phys., 1964, 40, 1317.N. M. Atherton, R. N. Dixon, and G. H. Kirby, Nature, 1965,D. R. Brown and M. C. R. Symons, J.C.S. Dalton, 1977,J. M. Howell and J. R. Van Wazer, J . Amer. Chem. SOC.,P. W. Atkins and M. C. K. Symons, J . Chem. SOC., 1962,10 P. W. Atkins and M. C. R. Symons, ‘The Structure ofM. C. R. Symons and D. N. Zimmerman, J.C.S. Faraduy I ,l2 R. S. Eachus and M. C. R. Symons, J . Chem. SOC. ( A ) , 1970,l3 R. L. Hudson and F. Williams, J . Amer. Chem. Soc., 1977,l4 J. P. Dunitz, J . Amer. Chem. SOC., 1956, 78, 878.206, 83.1389.1974, 96, 7902.4794.Inorganic Radicals,’ Elsevier, Amsterdam, 1967.1976, 409.3080.99, 7714

ISSN:1477-9226    

DOI:10.1039/DT9790000423

出版商:RSC    

年代:1979

数据来源: RSC