23 



over, because there then follow various thermal reactions (constituting a third 

 stage), some of known and many of largely unknown types but still definitely the 

 familiar chemical reactions considered by the chemist, with, of course, the 

 well-known complexity that the spatial distributions are not homogeneous as in 

 ordinary chemical reactions or even as in most photochemical reactions. The 

 problem of these heterogeneities has received very little serious attention; Dr. 

 Magee in particular, has pioneered in the field. I think that we can look for a 

 good deal of progress on this point in the coming years. 



This afternoon I shall discuss just the second stage, namely, the degradation 

 of kinetic energy of electrons having energy too low to excite electronic systems. 

 For simplicity I should like to consider a single electron interacting with liquid 

 water. This idealized problem should be immediately applicable to radiobiology. 

 I might mention in passing that when this problem of the loss of energy by sec- 

 ondary electrons is mentioned or treated at all in the literature, it is almost al- 

 ways treated incorrectly. For instance, the statement is commonly encountered 

 that electrons lose energy in elastic collisions, in packets of thermal size, etc. 

 This is incorrect. 



Let us start by considering the various types of interaction of a rather slow 

 electron with the medium. First I shall outline these various types and what we 

 know or do not know about them. Then I shall try to construct a picture of the 

 actual moderation of electron kinetic energy from the lowest electronic excita- 

 tion potential down to thermal energy. Finally, inquiry will be made into the 

 fate of the thermalized electron. 



The types of interaction of an electron of energy of a few electron volts with 

 liquid water are three. First, it may interact with the oscillations of the medium 

 -- the atomic oscillations. The electronic oscillations are excluded by energy 

 conservation. Second, it may interact with the "dipolar structure" of the medi- 

 um. And, third, it may suffer elastic scattering. We shall find that the first 

 two dominate the picture as far as the over-all energy loss is concerned, al- 

 though the third exercises an important influence in determining the actual path 

 of the electron. 



Consider first, then, the problem of how a moving electron interacts with a 

 system consisting of atoms which do, or can, oscillate. Here, in much of what 

 I am going to say, we have very little information and have, for the most part, 

 to be guided by general considerations. 



The theory for the "inelastic" energy transfer from a slow electron to a 

 molecule --an isolated molecule, now, and therefore one in the gaseous phase 

 --is available only for two special cases, neither of them properly applicable to 

 our problem. In the first instance there is a paper published many years ago by 

 Massey (Z) treating the impact of an electron with a homopolar diatomic mole- 

 cule, that is, a symmetrical one of the type H^ or N;?. In the second, there is a 

 paper by Wu (3) treating the interaction of a slow electron with a molecule bear- 

 ing a dipole moment which is comparatively small. This is also interesting but 

 not quite relevant because we have to concern ourselves, in considering water or 

 an aqueous medium or a biological system, with molecules, the dipole moment 

 of which is great (in atomic units, e.g. , comparable to unity). Therefore, from 

 these studies we can learn comparatively little except that the probability of 

 transfer of oscillational quanta to the molecule in impact is appreciable under 

 certain circumstances; as a matter of fact, the excitation curve in the two cases 

 that I have named looks a good deal like most other types of excitation curve. It 

 starts at a threshold, rises to a maximum, and then declines with increasing 

 energy. 



