VOLUME 39, NUMBER 17 LETTERS 24 OCTOBER 1977

I hatt. 38, 788 (1977). 6See, for instance, A. Sosin and W. Bauer, in Studies 'D. Markowitz and L. P. Kadanoff, Phys. Rev. 131, in Radiation Effects in Solids, edited by G, J. Dienes 563 (1963). (Gordon and Breach, New York, 1969), Vol. 3, p, 153. H WlesmaIlny M. Gurvltch~ Ao K. Ghosh~ Ho Lutz~ ~H. Wiesman, M. Gurvitch, A. K. Ghosh, H. Lutz, K. %'. Jones, A. N. Goland, and M. Strongin, to be O. F. Kammerer, and Myron Strongin, to be published. published. H. Lutz, H. Wiesmann, O. F. Kammerer, and Myron M. S511, R. Boning, and H. Bauer, J. Low Temp. Strongin, Phys. Rev, Lett. 36, 1576 (1976); C. L. Snead, Phys. 24, 631 (1976). Jr. , Appl. Phys. Lett. 30, 662 (1977). ~R. Viswanathan, R. Caton, and C. Pande, to be pub- S. KlaumGnzer, A. Ischenko, and P. MGller, Z. Phys. lished, 268, 189 (1974).

Distinguishing between Stars and Galaxies Composed of Matter and Using Helicity Detection

John G. Cramer DePartment of , University of Washington, Seattle, Washington 98195

Winred J. Braithwaite DePa&nzent of Physics, University of Texas, Austin, Texas 78712 (Received 14 February 1977; revised manuscript received 12 July 1977)

The produced in fusion processes in matter stars will have predominantly a "right" helicity due to the nonconversation of parity in weak interactions. This helicity is transferred to bremsstahlung and forward in-flight annilation radiation, which will be right-circularly polarized. In antimatter stars, CP symmetry wQl make the equivalent radiation left-circularly polarized. The helicity of such radiation can be used to distin- guish between astronomical objects composed of matter and antixnatter. The tantalizing possibility that macroscopic matter- antimatter annihilation or antineutrino de- quantities of antimatter may exist in our universe tection. has been the subject of calculation and speculation It would seem at first sight that the symmetry for more than two decades. ' -symmetric of the physical processes in stars composed of cosmologies have been proposed which would re- matter and of antimatter would make them obser- quired equal amounts of matter and antimatter to vationally identical so that no way could be found be present in the universe. ' It has also been pro- of distinguishing between them. However, there posed that a composed of normal mat- is one exception to this general symmetry which ter may "evaporate" equal amounts of matter and arises from the nonconservation of parity in the antimatter, ' tending to symmetrize the matter- weak interactions and from CPT symmetry. ' The antimatter balance of the universe if it were not thermonuclear processes which occur in stars, a Priori symmetric. whether involving the cycle, the carbon In a recent review article, Steigman4 pointed cycle, or nucleosynthesis of heavier elements, out that all methods of locating antimatter in the systematically convert into . universe which have so far been suggested (ex- This conversion takes place through the weak-in- cept for antineutrino detection) require the intro- teraction processes of P' decay and cap- duction of normal matter into a region of ture. When positrons (P') are emitted they will which contains antimatter, or vice versa, so that be preferentially in a "right"-helicity -antimatter annihilation takes place and strength v/c as a consequence of the nonconser- can be observed. In the present work we suggest vation of parity and the maximal parity nonconser- an alternative method which uses the helicity of vation in P decay. Thus the spins will secondary electromagnetic radiation produced by be aligned along their lines of flight. The emis- stellar fusion processes to distinguish between sion of these positrons may be accompanied by stars composed of matter and antimatter. This the emission of i~net bramsstrahlung, ' and when method, though very difficult, does not require they are slowed in matter they are likely to pro- VOLUME 39, NUMBER 17 PHYSICAL REVIEW LETTERS 24 OcroBER 1977

duce external breamsstrahlung. The helicity of escaping. It would thus appear that there is little the positron mill be transferred to both of these possibility of directly observing this radiation if types of bremsstrahlung, making them right-cir- it originates within a normal star. cularly polarized. " In other words, the brems- On the other hand, a , even one locat- strahlung helicity, like that of the positron, will ed in another galaxy, offers the possibility of di- be such that the is aligned along the photon's rect observation of the matter/antimatter helici- line of flight. ty. It is believed that this type of stellar explo- The polarized positrons may also undergo anni- sion occurs when the fusion processes in a mas- hilation in Qight in which the energetic positron sive star have terminated with the production of strikes an electron at rest, producing two pho- a large quantity of "Ni at the peak of the binding- tons, one along the line of Qight of the positron energy curve, so that the fusion fuel of the star and one in the opposite direction. The forward- is exhausted. The ~Ni decays by electron cap- going annihilation photon, which has most of the ture to ~Co, which decays by electron capture energy, will preferentially have the same helicity and positron decay (18%%uo branch and 1.46 MeV end- as the positron. ' point energy) to ~Fe." A large fraction of the A fourth source of polarized secondary radia- energy released after a supernova explosion is tion is the internal bremsstrahlung which accom- ascribed to this decay sequence, and under the panies the weak-interaction process of electron assumption that the stellar envelope blows off, capture. Here the sudden change in the charge of most of the expected energy release of energetic the nucleus by one unit produces a "shock" which occurs at the su~face. Moreover, the po- induces inner bremsstrahlung with a probability sitrons are emitted in a medium which is predom- on the order of the fine-structure constant. This inantly iron (rather than hydrogen), increasing electromagnetic radiation has preferential helici- the probability of bremsstrahlung by a factor of ty which closely corresponds to the helicity of in- (Z~, /ZH)' or 676. Therefore, the probability of ner bremsstrahlung induced by P' emission. ' directly observing the polarized secondary radia- Thus, three bremsstrahlung processes and anni- tion from this decay sequence should be greatly hilation in Qight all produce energetic radiation enhanced. which is right-circularly polarized. I et us estimate the minimum flux of energetic For antimatter stars, exactly the same process- photons from a matter/antimatter object. Follow- es mill take place except that the burning process- ing Clayton, "we consider the case of a super- es will convert antiprotons into antineutrons, pro- nova 1.0 Mpc away in which 0.14 solar masses of ducing antipositrons, i.e., . As a con- material is released in the form of "Ni, and in sequence of CPT symmetry (actually CP symme- this minimum estimate, we only consider the pho- try since time reversal is not involved), these tons produced by the inner-bremsstrahlung proc- electrons will have their spins aligned antiPaxal- ess. The later is likely to be only a small frac- lel to their lines of flight and produce brems- tion of the total photons of interest, but since it strahlung and forward in-Qight annihilation radia- occurs as part of the weak-interaction decay it tion which are left-circularly polarized. ' Thus can be estimated without making assumptions the secondary energetic photons produced by ther- about the nature of the medium in which the posi- monuclear processes in matter and antimatter trons are stopping. We reduce the estimate of stars will tend to be polarized with opposite hel- Clayton" by the factor reflecting the probability icitites. Hence, an asymmetry exists between of inner-bremsstrahlung production and find that matter and antimatter which, in principle, allows the peak flux occurs about 1 month after the on- one to distinguish between the two without the ne- set of the supernova and has a value of about 2 cessity of annihilation between them. x10 ' photons/m' sec with a mean photon energy The existence of a possible method of distin- of about 0.2 MeV and a mean helicity of about guishing between matter and antimatter stars and 25%%up. If a polarimeter 1 m in diameter, mounted galaxies brings into sharp focus the question of in space, accumulates counts for a period of whether this radiation is aetuaL. ly observable. about 1 yr starting within 2 weeks after the super- The thermonuclear processes which produce the nova is observed to initiate, the polarimeter radiation of interest normally occur deep in the would process about 3 &&10' events, which should stellar interior, End the radiation would have to permit a determination of the sign of the helicity. traverse an enormous number of mean free paths However, as previously mentioned, the photon of stellar matter before reaching the surface and flux from external bremsstraMung may be 10'-

1105 VOLUME 39, NUMBER 17 PHYSICAL REVIEW LETTERS 24 OcTQBER 1977

10' times larger than produced by inner brems- linear and circular polarization arising from in- strahlung, because of the presence of tne iron terstellar grains. " stopping medium. As a benchmark for the diffi- A quantitative estimate of the size of the mat- culty of these measurements in the present state ter/antimatter helicity in visible light is beyond of y-ray , it should be noted that even the scope of the present Letter. It will require the direct y-ray transitions from the ~Ni-decay detailed stella. r transport calculations with par- sequence in supernovae have not yet been experi- ticular attention to polarization-transfer process- mentally observed. " es. A formalism suitable for such calculations The study of continuum x-rays and y rays in has been developed by Lipps and Tolhoek, " but to search of the predicted matter/antimatter helici- our knowledge this formalism has never been ap- ty will be very difficult, but it has the distinct ad- plied to astrophysical phenomena. vantage that there are few strong competing proc- Let us finally turn to a brief discussion of the esses which can produce photons of a preferen- experimental techniques available for the mea- tial helicity in this energy region. One alterna- surement of the circular polarization of x rays tive process, however, is synchrotron radia- and y rays. Recently, a number of x-ray polar- tion. " In magnetic stars such radiation might imeters for astronomical use have been designed compete in intensity with bremsstrahlung, partic- and built. " Unfortunately, all of these instru- ularly at the lower energies, i.e., in the x-ray ments are designed to measure linem polariza- region and, moreover, this process has been tion and are completely insensitive to the circu- shown to have a nonzero helicity under some cir- lar polarization of the radiation detected. There- cumstances fore, new astronomical x-ray and/or y-ray polar- Finally, there is the possibility that a small imeters will have to be designed for the measure- remnant of the matter/antimatter helicity, aris- ments envisioned here. ing from the effects described above, might re- The only currently available technique for de- main in visible starlight because of-the infrared termining the circular polarization of photons in catastrophe in the bremsstrahlung cross section the x-ray and y-ray region uses the differential at low photon energies. Mccoy has shown that of polarized photons from for low-energy external bremsstrahlung the trans- polarized electrons. There is a sizable compo- ferred helicity decreases linearly with decreas- nent of the Compton cross section which changes ing photon energy. ' Moreover, a given positron sign, depending on the relative helicity of the col- is not limited to a single bremsstrahlung event, liding photon and electron. The polarized elec- and may produce a large number of low-energy trons employed are the valence electrons of a photons before it is stopped. magnetically saturated ferromagnetic material, It seems likely that in normal stars any resid- usually iron. ' In magnetized iron, unfortunately, ual matter/antimatter helicity in visible starli. ght only 2 out of 28 orbital electrons are spin aligned will be very smalI. . The case may be somewhat and will contribute to the polarization-dependent more favorable for light from supernovae be- cross section. Thus, even for 100%%-polarized ra- cause the escape time for visible radiation will diation the detection asymmetries would amount be much shorter and the number of scatterings to only a few-percent change in the counting rate before escape will be much smaller. when the magnetizing field reverses the spins of Any identification in visible light of the matter/ the valence electrons. antimatter helicity should be facilitated by its Recently, the prospect of more efficient ana- distinctive characteristics: (1) It has the same lyzers of circularly polarized x-rays and y rays sign of circular polarization at all wavelengths; has been raised by theoretical predictions that (2) the strength of the circular polarization should stable systems of spin-aligned hydrogen (H&), and decrease slowly with increasing wavelength; and deuterium (&), and lithium (Li&) may be feasible (3) the circular polarization should be unrelated at temperatures (0.1-1.0 K) and magnetic fields to any linear polarization which may present in (- 50 kG) which are within the reach of present the same light. Identification of this helicity, technology. " For such analyzers, either all of however, is made more difficult by a number of the electrons (H& and D4) or one out of three elec- other mechanisms which can produce circular trons (Li&) will contribute to the polarization-de- polarization in visible light from stars, particu- pendent Cornpton cross section, which would vast- larly the Zeeman effect for continuum radia- ly improve the analyzing power. tion, '~' circular polarization in sunspots, '~ and The polarization-dependent part of the Compton

1106 VOLUME 39, NUMBER 17 PHYSICAL REVIEW LETTERS 24 OcTQBER 1977 total cross section changes sign at 630 keV. This 'Q. Gamow, Phys. Rev. 74, 505 (1948), and Nature means that transmission-type polarimeters will (London) 162, 680 (1948); M. Goldhaber, Science 124, only be useful well above (or below) this energy. 218 (1956); Ya. B. Zel'dovitch, Adv. Astron. Astro- Scattering-type polarimeters, however, can be phys. 241 (1965). used at all energies. " They have characteristi- H. Alfven, Rev. Mod. Phys. 37, 652 (1965); R. Omnes, in Gamma-Ray , NASA SP-339, edited cally poorer energy resolution, but offer the ad- by Floyd W. Steeker and Jacob I. Thrombka (U.S. GPO, ditional advantage that the scattering analyzer 1973). can be made much larger than the photon detector S. W. Hawking, Commun. Math. Phys. 49, 199 (1975). itself, so that polarimeters with very large aper- 4G. Steigman, Annu. Rev. Astron. Astrophys. 14, tures and good background rejection can be con- 339 (1976). templated. Clearly then, the polarimeter design C. S. Wu and S. A. Moszkowski, (Wiley, is a very challenging problem and the designer New York, 1966), pp. 149ff. 6M. Goldhaber, L. Grodzins, and A. W. Sunyar, will have to deal effectively with the problems of Phys. Rev. 106, 826 (1957); C. Fronsdal and H. Uber- small fluxes, small asymmetry effects, and dif- all, Phys. Rev. 111, 580 (1958). ficult technologies, particularly since such a po- K. W. McVoy, Phys. Rev. 106, 828 (1957). larimeter will have to be located in space while L. A. Page, Phys. Rev. 106, 394 (1957). (This paper collecting data for the duration of the energetic- uses a polarization convention which is opposite the photon emission of a supernova, i.e. , several Basel convention used in the present work. ) ~B. months to several years. Q. Pettersson, in AlPha-, Beta-, and Gamma- Ray Spectroscopy, edited K. Siegbahn (North-Hol- In summary, we have proposed that the circu- by land, Amsterdam, 1964), Vol. II, p. 1569. lar polarization of energetic photons, and possi- 'OD. D. Clayton, in Proceedings of the Conference on bly also of visible light, may permit stars and Explosive Nucleosynthesis, edited by D. N. Schramm galaxies composed of matter and antimatter to and W. D. Arnett (Univ. of Texas Press, Austin, 1973). be distinguished. The experiments are very dif- "V. L. Ginsburg and S. I. Syrovatskii, Annu. Rev, ficult but perhaps possible, particularly for su- Astron. Astrophys. 3, 297 (1965). pernovae. ~ J. C. Kemp, Astrophys. J. 162, 169 (1970). ~3J. D. Landstreet, in Planets, Stars, and Nebulae The authors are indebeted to Dr. Peggy Dyer Studied saith Photopolarimetry, edited by T. Gehrels and Professor David Bodansky (University of (Univ. of Arizona Press, Tucson, 1974), p. 981. the Washington) for stimulating initiation of this '4D. L. Mickey and F. W. Orrall, in Planets, Stars, work and for providing very useful insights into and Nebulae Studied ~ith Photopolarimetry, edited by experimental, y-ray astronomy, to Professor T. Gehrels (Univ. of Arizona Press, Tucson, 1974), John A. Wheeler (University of Texas) for his p. 686, very helpful suggestions and encouragement, to ~5P. G. Martin, in Planets, Stars and Nebulae Studied Professor Paul E. Boynton (University of Wash- mth Photopolarimetry, edited by T. Gehrels (Univ. of Arizona ington) and Professor W. David Arnett (Universi- Press, Tucson, 1974), p. 926. F. W. Lipps and H. A. Tolhoek, Physica (Utrecht) of and su- ty Chicago) for suggesting that novae 20, 85 (1954). pernovae might be particularly favorable cases 'YR. Novick, in Planets, Stars, and Nebulae Studied for helicity observations, and to Professor E. G. aoith Photopolarimetry, edited by T. Gehrels (Univ. of Adelberger and Professor E. M. Henley (Univer- Arizona Press, Tucson, 1974), Ref. 14, p. 262. ' sity of Washington), and R. A. Matzner (Universi- H. Schopper, Nucl. Instrum. Methods 3, 158 (1958). '~W. ty of Texas) for illuminating discussions and help- C. Stwalley and L. H. Nosanow, Phys. Rev. Lett. ful comments. 36, 910 (1976); W. C. Stwalley, Phys. Rev. Lett. 37, 1628 (1976), and private communication. This work was supported in part by the U. S. Energy Research and Development Administra- tion and the Robert A. Welch Foundation.

1107