assay. Both adducts were much less mutagenic than NOF (Tables I and II). It is possible that the small degree of mutagenicity observed in the case of these adducts may arise due to the formation of small amounts of the highly mutagenic NOF via a retro-Alder-ene decomposition of the adducts in the test system. Careful studies on the stability of these free radical adducts will be neccessary before the above possibility can be discounted. The re- duced mutagenicity of the adducts compared to NOF indicates the potential of unsaturated lipids to reduce the mutagenicity of NOF. It is premature to draw con- clusions about the effect of reduced mutagenicity of these adducts on carcinogenesis because the relationship be- tween mutagenesis and carcinogenesis is not clear-cut. Although a positive correlation between mutagenicity and carcinogenicity exists for many environmental pol- lutants, 2° exceptions are also known. Thus, for instance, the myristate ester of N-hydroxy-2-acetylaminofluorene is more carcinogenic, but less rhutagenic in the Ames mutagenesis assay than N-hydroxy-2-acetylaminofluo- rene. 21 The NOF-lipid adducts have the potential to act as storage pools for the carcinogen inside the cell.

The addition of NOF to carbon-carbon double bonds in cellular membranes would lead to structural changes which may have far reaching implications for transport phenomena in cells. Formation of NOF-lipid adducts may cause membrane perturbations. Therefore, we stud- ied the effect of temperature on the ESR spectrum of the free radical generated by the reaction of NOF on phos- phatidyl choline liposomes, and the spectra obtained at different temperatures are shown (Fig. 7). The effect of temperature on the relative mobility of a nitroxyl free radical can be estimated by comparing the ratio of the high field peak to the low field peak of the spectra obtained at different temperatures. A comparison was made between the mobility of the free radical obtained

by the action of NOF on microsomes and the correspond- ing free radical generated from phosphatidyl choline liposomes (Fig. 8). The free radical generated from lipo- somes began to show enhanced mobility at 25°C whereas the microsomal radical did not show similar freedom of mobility until 45°C. The presence of proteins in micro- somes may cause the membrane to be more rigid and this may account for the above difference. Membrane alterations due to addition of NOF to unsaturated lipids in cellular membranes may be a factor in NOF-carcino- genesis. Work is in progress to establish whether NOF- lipid adducts have a protective or promotional function in the process of AAF carcinogenesis.

1. R. H. Wilson, F. DeEds, and A. J. Cox, Cancer Res. 1, 595 {1941). 2. C. C. Irving, J. Biol. Chem. 239, 1589 (1964). 3. P. H. Grantham, E. K. Weisburger, and J. H. Weisburger, Biochim. Biophys.

Acta 107, 414 {1964). 4. H. Bartsch, J. A. Miller, and E. C. Miller, Biochim. Biophys. Acta 273, 40

{1972). 5. H. Bartsch, M. Traut, and E. Hecker, Biochim. Biophys. Acta 237,556 (1971). 6. R. A. Floyd, L. M. Soong, M. A. Stuart, and D. L. Reigh, Arch. Biochem.

Biophys. 185, 450 (1978). 7. B. N. Ames, E. G. Gumey, J. A. Miller, and H. Bartsch, Proc. Natl. Acad. Sci.

U.S.A. 69, 3128 (1972). 8. B. N. Ames, J. McCann, and E. Yamasaki, Mutat. Res. 31, 347 (1975). 9. L. A. Poirier, J. A. Miller, and E. C. Miller, Cancer Res. 23, 790 (1963).

10. E. Brill, Experientia 30, 835 (1974). 11. H. E. May and P. B. McCay, J. Biol. Chem. 243, 2288 (1968). 12. J. Folch, M. Lees, and G. H. Sloane-Stanley, J. Biol. Chem. 226, 497 (1957). 13. A. N. Holden, C. Kittle, F. R. Merritt, and W. A. Yager, Phys. Rev. 77, 147

(1950). 14. R. H. Hoskins, J. Chem. Phys. 25, 788 (1956). 15. A. K. Hoffman and A. T. Henderson, J. Am. Chem. Soc. 83, 4671 (1961). 16. G. M. Coppinger and J. D. Swalen, J. Am. Chem. Soc. 83, 4900 (1961). 17. E. G. Rozantsev and M. B. Neiman, Tetrahedron 20, 131 (1962). 18. M. B. Neiman, E. G. Rozantsev, and Y. G. Mamedov, Nature 196, 472 (1962). 19. A. B. Sullivan, J. Org. Chem. 31, 2811 (1966). 20. J. McCann, E. Choi, E. Yamasaki, and B. N. Ames, Proc. Natl. Acad. Sci.

U.S.A. 73, 950 (1976). 21. H. Bartsch, C. Malaveille, H. F. Stich, E. C. Miller, and J. A. Miller, Cancer

Res. 37, 1461 (1977).

Electron Spin Resonance and Electron Nuclear Double Resonance Spectroscopy: Application to the Study of Trapped Electrons

HAROLD C. BOX and HAROLD G. FREUND Biophysics Department, Roswell Park Memorial Institute, Buffalo, New York 14263

V band electron spin resonance-ENDOR spectroscopy has been appl ied to the s tudy o f " t r a p p e d " electrons in single crystals o f po lyhydroxy and carbohydrate compounds X-irradiated at l o w temperature. Evidence is presented f r o m E N D O R measure- ments o f h y d r o x y proton coupl ings that electrons are stabi l ized

R e c e i v e d 26 D e c e m b e r 1979. * P r e s e n t e d a t t he 21st A n n u a l R o c k y M o u n t a i n Confe rence on

A n a l y t i c a l C h e m i s t r y , 30 J u l y 1979.

Volume 34, Number 3,1980

at intermolecular sites. The electron can b e s t a b i l i z e d in the dipole f i e l d s o f t w o or three hydroxy g r o u p s .

I n d e x Head ings : ESR spectroscopy; ENDOR spectroscopy; T r a p p e d electrons.

INTRODUCTION

Electron spin resonance (ESR) spectroscopy has proven to be a valuable method for studying the effects

APPLIED SPECTROSCOPY 293

of ionizing radiation at the molecular level. In the act of removing electrons from normally diamagnetic molecules ionizing radiation creates paramagnetic free radical prod- ucts which are appropriate species for analysis by the electron spin resonance technique. Primitive forms of the oxidized molecules can usually be stabilized if the tem- perature is maintained at 4.2 K {boiling point of liquid helium).

Typically the electrons ejected in the ionization proc- ess are picked up by undamaged diamagnetic molecules generating additional paramagnetic species associated with reduction processes.

The ESR spectra are invariably anisotropic in g value and in hyperfine splitting which is to say these features vary as the orientation of the free radical molecule is changed with respect to the applied magnetic field. The g value is a measure of the effective magnetic moment of the free radical. A hyperfine splitting is a measure of the magnetic interaction between the unpaired electron and a magnetic nucleus in its environment. Due to spectral anisotropy it is advantageous to examine radiation effects in single crystals wherein all molecules are of like ori- entation. Even so, electron nuclear double resonance (ENDOR) spectroscopy is often an indispensable adjunct for identifying radiation products since the ESR spectra of different products usually superimpose. Using the ENDOR method accurate measurements of hyperfine couplings can be obtained despite superposition. 1

Exceptions to the formation of radiation-induced re- duction products can occur in cases where the irradiated crystal is constituted of molecules of low electron affinity. In these instances the electron may be trapped between molecules rather than form electron adducts. This hap- pens in the case of several polyhydroxy compounds ir- radiated at 4.2 K e. The study of electrons trapped in this manner using ESR-ENDOR spectroscopy is the subject of this report.

I. EXPERIMENTAL

The data which provided the basis for this report were obtained from ESR and ENDOR measurements mainly at V band. More precisely the microwave .frequency employed to resonate the electrons was 70 GHz, although a few measurements were made at 24 GHz (K band) also. The main features of the superheterodyne ESR spec- trometer were described previously. 3 However, signifi- cant advances in millimeter microwave technology have occurred in the meantime whose incorporation into the spectrometer have resulted in improved performance. The klystron source oscillator was replaced by a Hughes IMPATT diode. Interestingly the IMPATT without ex- ternal stabilization demonstrated short-term stability comparable to the klystron locked to an external refer- ence cavity. The balanced mixer detector employing semiconductor diodes was replaced by a Baytron hybrid ring with gallium arsenide Shottky detectors. These mod- ifications improved signal/noise ratio and general relia- bility.

The transmission sample cavity resonating in the TE 011 mode has proven highly serviceable. The entire sam- ple cavity is immersed in cryogen for low temperature studies. The length of the cavity is adjustable so that the

294 Volume 34, Number 3, 1980

cavity can be resonated to the microwave frequency independent of temperature. The magnetic field required for V band spectroscopy is about 25 kG. An important modification has been the substitution of a rotatable electromagnetic so that the orientation of the magnetic field with respect to a crystal sample can be changed without disturbing the sample. The practicality of ENDOR measurements greatly improved with this ca- pability since it is necessary to trace ENDOR resonances as a function of field orientation in order to deduce hyperf'me coupling tensors.

The greater resolution of absorption components at V band compared with K band is illustrated in Fig. 1 which shows the ESR absorption spectrum from a single crystal of arabinose X-irradiated and observed at 4.2 K. The high field component is attributed to trapped electrons. The V band spectrum due to trapped electrons appears simpler due to elimination of forbidden transitions which complicate the K band spectrum. In our experience ENDOR spectroscopy often proves effective at V band when it fails at K band, whereas the inverse has not been true. However, instances are still encountered where even at V band satisfactory ENDOR signals are not obtained. The improved ENDOR results generally obtained at V band are due in part to the improved quality of the ESR signal whose perturbation by a swept radiofrequency field gives rise to the ENDOR spectrum. An improved ESR signal at V band results from the greater difference in spin populations at the higher magnetic field strength. Better separation of overlapping components at V band, even though still incomplete, as in the case of the trapped electron absorption shown in Fig. 1, also leads to im- proved ENDOR results, by discriminating against un- wanted resonances from the other components of the

(Q)

ALKOXY \ / RADICAL~ ~ /

g = 2 .0025

I00 G

TRAPPED fJ~ELECTRON

FIG. 1. E S R absorpt ion spect ra obta ined from a single crystal of arab- inose X- i r radia ted and observed at 4.2 K. The magnet ic field was paral le l to the c axis. a, K band; b, V band.

T A B L E I. H y p e r f i n e coupl ing t ensors for protons in teract ing w i t h trapped e lec trons in irradiated arabinose crysta ls .

Prinei- Iso- pal val- tropic

ues (MHz) a

Direct ion cosines a

a b c

71.26 0.749 -0.027 -0.660 A, 16.54 34.25 -0.444 0.719 -0.533

14.96 0.490 0.693 0.527

88.30 0.484 0.529 0.695 A2 31.12 49.89 0.715 0.218 -0.663

30.24 -0.503 0.819 -0.273

50.14 0.582 0.772 0.253 A.~ -2 .06 14.93 0.728 -0.356 -0.585

-3 .28 0.361 -0.525 0.770

a Direct ion cosines are wi th respect to crystal axes.

T A B L E II. Electron-proton distance, r, ca lculated f r o m Eq. (2) and t h e s p i n dens i ty at the proton, p, ca lculated f r o m Eq. (1)2

Crysta l r p

Dulci tol 1.59 0.066 Arabinose 1.60 0.035 Dulci tol 1.61 0.066 Arabinose 1.62 0.024 Arabinose 1.65 0.011 Rhamnose 1.66 0.006 Sorbitol 1.67 0.030 Sorbitol 1.68 0.023 Sucrose 1.69 0.037 Rhamnose 1.70 0.039 Sucrose 1.71 0.033 Rhamnose 1.72 0.029 Xyl i to l 1.73 0.032 Glucose phospha te 1.73 0.039 Glucose phospha te 1.74 0.014 Xyli tol 1.75 0.025

a The da ta are a compilat ion of resul ts obta ined from E N D O R meas- u rements on seven different crystals. 2

overall absorption. Of course, the quality of ENDOR signals must be optimized in each application by empir- ical adjustment of several parameters including concen- tration of paramagnetic centers, microwave power level, and radiofrequency power level.

II. T R A P P E D E L E C T R O N S

The proton hyperfine coupling tensors deduced from ENDOR measurements on the high field component of the ESR absorption obtained from arabinose (Fig. 1) are given in Table I. From studies on partially deuterated crystals, it is known that these protons are exchangeable and thus belong to hydroxy groups. Our interpretation of this absorption is that it arises from electrons stabilized at an intermolecular site. The environment of the elec- tron consists of three polar hydroxy groups each of which has its proton located relatively close to the proton.

The spin density of the electron at each proton can be calculated from the isotropic component of the hyperfine coupling, which is the mean of the principal values of the coupling tensor, using the expression

p = 3 a h / 8 ~ r g f l g n f l n . (1)

In Eq. (1) g is 2, fl is the Bohr magneton, gn is 5.5854, fin is the nuclear magneton, h is Planck's constant, and a is the isotropic coupling. Assuming a point dipole model for the electron spin distribution, the electron-proton distance can be calculated from the expression

r = [ 2 g f l g , f l n / h ( A m , x - a ) ] 1/3 (2)

In Eq. (2) Amax is the principal value of the hyperfine coupling tensor corresponding to maximum coupling. The direction of r is that indicated by the direction cosines in Table I corresponding to maximum coupling.

Table II is a compilation of results obtained to date in this laboratory from measurements on polyhydroxy and carbohydrate compounds irradiated at low temperature. Several interesting observations can be made. It appears that the electron can be stabilized in the potential well created by just two polar groups. Although the range of r in Table II is considerable, the values within a given electron trap are quite uniform. This is in accord with intuitive expectation that the electron should tend to- ward a position equidistant from the protons forming the trap. It is also interesting to note that only a single trapping site is utilized in any given crystal structure even though many of the crystal structures would appear to offer more than one such site.

The trapping of electrons in polar glasses, such as alcohols, irradiated at low temperature is a well known phenomenon. 4' ~ Unlike the few large hyperfine couplings observed in crystals, an overall weaker hyperfine inter- action from multiple protons is observed in glasses. In other respects the characteristics of trapped electron in glasses and crystals are more akin. In both media the ESR absorption can be bleached using visible light. The g values of the absorptions in both media are close to 2.001.

In retrospect perhaps it should not have been surpris- ing to find that electrons can be stabilized at intermolec- ular sites in crystals provided the constituent molecules are of low electron affinity. At 4.2 K k T is 3.6 × 10 -4 eV. Calculations based on simple models with an electron localized between two dipoles simulating OH groups yield energies approaching an electron volt.

ACKNOWLEDGMENT

Support through Department of Energy Contract EY-76SO23212 and National Cancer Institute Grant CA-25027 is gratefully acknowledged.

1. H. C. Box, Radiation Effects, ESR-ENDOR Analysis (Academic Press, New York, 1977).

2. H. C. Box, E. E. Budzinski, and H. G. Freund, J. Chem. Phys. 69, 1309 (1978); H. C. Box, E. E. Budzinski, H. G. Freund, and W. R. Potter, J. Chem. Phys. 70, 1320 (1979); E. E. Budzinski, W. R. Potter, G. Potienko, and H. C. Box, J. Chem. Phys. 70, 5040 (1979); E. E. Budzinski, W. R. Potter, and H. C. Box, J. Chem. Phys. 72, 972 (1980); S. E. Locker and H. C. Box, J. Chem. Phys. 72, 828 (1980).

3. H. C. Box, H. G. Freund, K. T. Lilga, and E. E. Budzinski, J. Phys. Chem. 74, 40 (1970).

4. L. Kevan, Int. J. Radiat. Phys. Chem. 6, 297 (1974}. 5. B. G. Ershov and A. K. Pikaer, Radiat. Res. 2, 1 (1969).

APPLIED SPECTROSCOPY 295