NOTES
1378
Fig. 1.-Curie-Weiss z *-'
plot of Ni203.H20.
60
e
Ni,0a.H20] g501; 50 1 11 V
I%
cu 4n 40
t
'**
gnetic Susceptibility at 25"
80
X
@ 20
70g
3
I
E 10
60:
w z 3
3 z € o0 l ' 1 ' I ' j ' I 1 ' 0 100 200 300 400 500 600 700 TEMPERATURE, OC. I
Fig. 2.-Magnetic
1 50
susceptibility and weight changes for the decomposition of Si2O8.Hz0.
decomposed on heating to Ni203.Hz0and then to N O . These authors2cQnfirmedthe presence of Ni3+by X-ray absorption edge measurements, although Hanson and Milligan, performing similar experiments, found no differences between these and ?Ti0 samples. Cairns and Ott also reported their failure to prepare anhydrous XiB03 by heating the hydrate. Recently, however, Aggarwal and Goswami4 detected lines in electron diffraction patterns attributed to hexagonal Niz03 on the oxidized surface of nickel films. The existence of nickelic oxide is denied by Vain~htein,~ who considers NizO3.xH20 to be a mixture of hydrated NiO and XO,. I n this investigation, three samples of Ki203-zH20 were prepared by adding bromine in KOH solution to solutions of nickelous chloride (a) with precipitation a t 25' and air-drying, (b) with precipitation a t 25' and drying a t loo', and (c) with precipitation a t 100' and drying in vacuo a t 2.5'. In each case chemical analyses were made for nickel (dimethylglyoxime methodj, "active" oxygen (Bunsen-Rupp test), and total oxygen (neutron activation analysis), with hydrogen obtained by difference. The results indicated the following compositions: (a) Niz03.p2-2.24H20, (b) pu'i,Os. 1.50HzO, and (c) Siz03.w~2.15H20.X-Ray diffraction patterns of all three samples are identical and agree with (2) R. W. Cairns and E. Ott, J . 4m. Chem. Soc., 56, 1094 (1934). (3) H. P.Hanson rtnd W. 0. Milligan, J . Phys. Chem., 60, 1144 (1956). (4) P. 6. Aggarwal and A. Coswami, ibid., 6S, 2105 (1961). (5) E. E. Vainshtein, Zh. Elcspeeq i Toore$. Fis., 20, 442 (1950).
Vol. 67
sample A in ref. 1, designated by Cairns and Ott as "Niz03.2H20." Magnetic susceptibility measurements were made from -196 to 100' on all samples, which demonstrated the same Curie-Weiss relationship shown in Fig. 1 for sample (b). The constants yield p = 2.488 and 6 = +21'. Sample (a) was heattreated for 24 hours at approximately 100" intervals up to 700'. Following each heat-treatment, the sample was cooled to room temperature and the magnetic susceptibility and weight measured. These data are plotted in Fig. 2. X-Ray absorption edge nieasurements revealed no differences between these samples and KiO. Conductivity and the Seebeck coefficient measurements show the material to be n-type, whereas NiO is p-type. Both the room temperature susceptibility and the weight of the oxide reach a constant value when heattreated above 400' with the susceptibility characteristic of N O . Partial loss of HzO occurs immediately upon heating to looo, but with no change in the magnetic properties. Decomposition begins at about 130°, where the weight corresponds to Ki20j.1.25H20. The maximum in susceptibility between 100 and 400' is due to the presence of small particles of NiO, which exhibit enhanced superparamagnetic susceptibilities when prepared a t lower temperatures.6 This effect decreases with higher temperatures of heat-treatment and disappears above 400'. The Curie-Weiss dependence of the susceptibility indicates the presence of one paramagnetic phase only. Chemical analysis identifies this phase as a hydrous oxide of trivalent nickel. Contrary to the findings of Cairns and Ottl but in agreement with Huttig and Peter,' this oxide may be dehydrated below Niz03-2Hz0 without changing either the structure or the magnetic properties. Decomposition to V i 0 occurs just before the monohydrate is reached. It, therefore, appears that, if this material is a true hydrate, it is NizOs-HzO with adsorbed water. Trivalent nickel in a low spin state has a spin-only magnetic moment of 1.73p. The observed moment of 2.488 suggests, therefore, a low spin state with considerable unquenched orbital contribution. The positive sign of the Weiss constant is an indication of ferromagnetic coupling between the nickel atoms. (6) J. T.Richardson and W. 0. Milligan, Phys. Rev.,102, 1289 (1956). (7) Hiittig and Peter, 2.anorg. &gem. Chem., 189, 190 (1930).
FREE RADICALS FORMED FROM 2-SUBSTITUTED NITRO COMPOUNDS1a BY A. J. TENcIrlb AND P. COPPESS'" Chpmistry Department3 Brookhaven National Laboratory, U p t o n , L . I . , New York
Receioed November 10. 1962
The photochemical rearrangement of o-iiitrobenzaldehyde to nitrobeiizoic acid was discovered in 19012 and has been investigated by various workers (see for example ref. 3). The reaction is known to occur both in the solid state and in solution. It seemed possible that a paramagnetic species could be formed during the (1) (a) Research performed under the auspices of the U. S. Atomic Energy Commission; (h) Chemistry Division, Atomic Research Establishment, Harwell, Didoot. Berks, England; (e) Crystallography Department, Weizmann Institute of Science, Rehovoth, Israel. (2) G.Ciamioian and P. Silber, Be?., S4, 2040 (1901). (3) P. Leighton and F. A. Lucy, J . Chem. Phys., 2 , 756 (1934).
June, 1963
photochemical proress and that an electron spin resonance (e.s.r.) study would give useful information. We have irradiated o-nitrobenzaldehyde within the microwave cavity of a Varian e m . spectrometer with 100 kc. field modulation. The light source was a one kilowatt G.E. mercury lamp and light shorter than 3500 A. was filtered out. Free radicals were observed in powdered o-nitrobensaldehyde at liquid nitrogen tennperature. The powder spectrum shows a distinct hyperfine structure, but anisotropic broadening of the peaks gives a large overlap and the spectrum is not easily analyzed; on warming to room temperature the signal largely disappeared. Free radical signals of comparable strength, but of a different nature, were obtained on the low temperature irradiation of o-nitrocinnamic acid and o-nitrobenzyl alcohol, which undergo reactions similar t o that of o-nitrobenzaldehyde.iL,6 Negative results were obtained, however, on irradiation of nitrobenzene and o-nitrobenzoic acid. This is expected since these compounds do not undergo the above photochemical reactions. These results indicate that the photochemical reactions of o-nit,robeiizaldehyde, onitrobenzyl alcohol, and o-nitrocinnamic acid may have a free radical intermediate. It may be noted that the ultraviolet spectra of nitrobenzene and o-nitrobenzoic acid are very similar to the ultraviolet spectrum of onitrobenzaldehyde.6--* Therefore the negative results obtained with the first two compounds a n n o t be due t o ins@Ficient absorption at wave lengths longer than 3500 A. We decided that it was preferable to study these compounds in solution since the anisotropic part of the coupling tensor is averaged The spectra obtained during the irradiation of 0.1 to 1.0 molar solutions of onitrobenzaldehyde in deoxygenated ethyl alcohol and isopropyl alcohol show a well resolved fine structure (Fig. la), which can be readily interpreted in terms of nitrogen and proton hyperfine splitting constants (Table I). The solutions were contained in a flat cell, 1 mm. thick, similar to the Varian aqueous cell. The disappearance of the radicals after the lamp has been switched off is proportional to the square of the free radical concentration. The rate constant for this reaction was found to be 1.3 1. mole-' sec.-l for an initially 1 molar solution of o-nitrobenzaldehyde in ethanol. The total number of radicals formed during the first 30 minutes was determined to be about 0.02% of the onitrobenzaldehyde concentration by comparison with a standard solution of DPPH in ethanol. In the same two samples the amount of conversion to o-nitrobenzaldehyde was determilied by reacting the iiitroso group with acidified K I and titrating the liberated iodine. About 4% of the o-nitrobensaldehyde was converted in 30 minutes. The error in the measurement of the free radical concentration may be as high as a factor of 2 but the low concentration indicates that the free radical we observe is not an jntermediate in the photochemical rearrangement to a n~itrosocompound. The main feature of the spectrum of the radical pro(4) F. Sachs and s. Hilpert, Ber., 37,3425 (1904). (5) I. Tanasescu, Bull. SOC. rhim., 41, 1074 (1927). (6) A. E. Lutskii and V. 'r. Alekseeva, J . Gen. Chem. USSR,29, 3211 (1959). (7) P. Grammaticakis, BuU. 8 o c . chim., [SI, 18,224 (1951); [SI, 20, 895 (1953). (8) H. Venner, Chem. Ber., 89, 1635 (1956). (9) s. I. Weissman, J . Chem. Phys., 22, 1378 (1954).
1379
XOTES 5 gauss
I
I
I
I
II
I II IIIII
II
I I
I1
I I I
I
5 gauss
(b) Fig. l,-(a) E.s.r. spectrum of radical produced on ultraviolet' irradiation of o-nitrobenzaldehyde; (b) e a r . spectrum of onitrobenxaldehyde negative ion in acetonitrile and 10% ethyl alcohol.
TABLE I HYPERFINE COUPLING CONSTANTS (IN GAUSS)~ Phot oradicals aNc 2-Nitrobenaaldehyde in 100% alcohol 8.1 7.9 in 10% alcoholb 5-Chloro-2-nitrobenaaldehyde in 100% alcohol 7.8 Negative ions 2-Nitrobensaldehyde in acetonitrile 7.2 in 10% alcohol' 8.6 5-Chloro-2-nitrobenaaldehyde in acetonitrile 6.55 6-Chloro-2-nitrobenaaldehyde in acetonitrile 8.55
ai
62
ai
a4
11.2 11.2
2.60d l.OOd 2.87d 0 . 9 j d
10.5
3.76
1.75
0.85
3 . 4 0 2.40 3.40 2.65
1.00 1.01
0.60 0.73
3.02
1.67
0.50
4.03
3.00
1.88
ab
0.30
