ANTINEUTRONS PRODUCED BY CHARGE EXCHANGE OF ANTIPROTONS*)

B. Cork, G. Lambertson, O. Piccioni, W. Wenzel

The principle of invariance under charge conjugation received convincing confirmation with the discovery of antiprotons produced in the Bevatron[^1–^3]. Another prediction of this theory awaiting experimental verification was the assertion of the existence of the antineutron. Additional interest in this particle is connected with the fact that charge conjugation, when applied to a neutral particle, is somewhat less evident than in the case when it is applied to particles possessing electric charge.

The purpose of the present experiment was to detect the annihilation of antineutrons produced from antiprotons by their charge exchange. Since a small yield of antineutrons was expected, a relatively large flux of antiprotons was required. Protons with an energy of 6.2 Bev bombarded the internal beryllium target of the Bevatron (Fig. 1). By means of a system of two deflecting magnets and five magnetic lenses, a beam of negative particles with momentum \(1.4\ \text{Bev}/c\) was obtained. An arrangement of six scintillation counters connected in a coincidence circuit made it possible to distinguish antiprotons from negative mesons by time of flight.

Fig. 1. Antiproton selector: \(Q_1\)—\(Q_5\) — focusing lenses; \(A_1\)—\(A_2\) — magnetic analyzers; \(A, B, C, D, E, F\) — scintillators of size \(4 \times 4 \times 1/4\) inch.

) B. Cork, G. R. Lambertson, O. Piccioni, W. A. Wenzel, Phys. Rev. 104*, 1193 (1956). Translated by Yu. V. Orlov.

The article by the American physicists Cork, Lambertson, Piccioni, and Wenzel presented here is the first serious publication on the discovery of the antineutron. The existence of the antineutron, like the existence of the antiproton, was predicted by the modern theory of elementary particles. According to the theory, antiparticles can exist not only for particles possessing electric charge, but also for neutral particles. In this case, generally speaking, the distinction

In Figs. 1 and 2, \(F\) is the last counter of this system, which registered from 300 to 600 antiprotons per hour.

As a result of the interaction of antiprotons with matter in the thick converter \(X\) (Fig. 2), antineutrons are sometimes formed; they pass through scintillators \(S_1\) and \(S_2\) without being registered and, finally, interacting with matter in the Cherenkov counter \(C\) made of lead glass, produce in it a light pulse whose large magnitude indicates that it is due to annihilation of a nucleon and an antinucleon.

Fig. 2. Antineutron detector: \(X\)—a scintillator in which antiprotons are recharged; \(S_1\) and \(S_2\)—scintillation counters; \(C\)—Cherenkov counter made of lead glass (in another series of experiments replaced by a large scintillator).

The Cherenkov counter \(C\) is a piece of lead glass measuring \(13\times13\times14\) inches, with density 4.8 and refractive index 1.8, viewed by 16 RCA 6655 photomultipliers. This instrument is analogous to that used in the preceding experiment on the detection of antiprotons1. Between \(S_1\) and \(S_2\) a lead plate 1 inch thick is placed to absorb high-energy \(\gamma\)-rays that could be confused with antineutrons. Neutrons and neutral mesons (heavier than \(\pi\)-mesons) could be registered by the Cherenkov counter, but the magnitude of their average light pulse is much smaller than that of the pulse due to antineutron annihilation. However, even a relatively small background from these secondary neutral products would distort the spectrum of pulses from antineutrons. To separate

The content of the experiments that led to the discovery of the antineutron is clear from the article itself and requires no comment. Let us merely explain that by the term “recharging of the antiproton” in what follows are meant processes of the type
\[ \bar p+p\to \bar n+n,\qquad \bar p+n\to \bar n+n+\pi^-, \]
where \(\bar p\) and \(\bar n\) denote the antiproton and antineutron. Let us note that, regardless of the apparently low statistical accuracy of the experimental data of Cork, Lambertson, Piccioni, and Wenzel, the fact of the detection of antineutrons, thanks to convincing control experiments, appears unquestionable.

ANTINEUTRONS OBTAINED BY CHARGE EXCHANGE OF ANTIPROTONS

these accompanying neutral particles, converter \(X\) was made of a scintillating toluene–terphenyl solution, viewed by four photomultipliers connected in parallel. Thus, pulses produced by neutral particles in the Cherenkov counter (“neutral events”) could be separated according to whether they arose in the annihilation of antiprotons, to which a large pulse in \(X\) corresponds, or in the less violent process of antineutron formation as a result of charge exchange of an antiproton. A quantitative criterion for such a separation is obtained from comparison of the pulse spectra in \(X\), shown in Fig. 3. The dashed curve, obtained in a specially performed experiment, is the spectrum of pulses produced by antiprotons that passed through \(X\) but did not undergo a nuclear

Fig. 3. Spectrum of pulses in scintillator \(X\), used for charge exchange, for 74 “neutral” events registered by a counter with lead glass. The histogram includes all registered cases. The smooth solid curve is used to determine the magnitude of the pulses from antiprotons that did not cause the counters \(S_1\) or \(S_2\) to operate. The solid dashed curve refers to antiprotons that did not undergo nuclear interactions in \(X\). Both curves are normalized to the histogram (by area).

Fig. 4. Spectrum of pulses caused by neutral particles in the lead-glass Cherenkov counter. The solid histogram includes 54 cases of antineutron annihilation (pulse size in scintillator \(X\) less than 100 MeV). The dashed histogram includes 20 pulses caused by other neutral particles. The solid curve refers to antiprotons and is normalized to the solid histogram.

interaction. The sharp peak in the spectrum also provides the possibility of calibration (it is easy to calculate that the ionization losses in \(X\) for antiprotons should amount to 50 MeV). The smooth solid curve in Fig. 3, obtained under the geometrical conditions shown in Fig. 2, includes cases of all antiproton interactions in \(X\) in which no pulse is observed in \(S_1\) and \(S_2\), and a pulse either does or does not appear in \(C\). The histogram in Fig. 3 gives the distribution of pulse heights for those events in which a neutral particle produces a pulse in \(C\). The difference between the solid curve and the histogram is remarkable in that it shows that the rare interactions producing neutral particles detected by the Cherenkov counter release in \(X\) considerably less energy than other interactions not selected in this way. In fact, the peak of the histogram is located at pulses smaller than the pulses corresponding to the ionization losses of protons that have not undergone nuclear interaction (50 MeV). This is precisely what should be expected if the neutral particles are antineutrons, since in that case nucleon annihilation could not have taken place in \(X\). Conversely, the formation of other energetic neutral particles would give the characteristic large pulse of annihilation in \(X\). The histogram thus indicates that the apparatus registers a small background of events of this

of the latter type. The pulse height of 100 MeV in Fig. 3 was chosen as the boundary separating this background from pulses caused by antineutrons. Figure 4 shows the distribution of pulse heights for events producing pulses in \(X\) smaller than 100 MeV (solid histogram) and larger than 100 MeV (dashed histogram). The enormous difference between the histograms, both in mean pulse height and in shape, confirms our interpretation, according to which neutral events are divided into antineutron and background events. Calibration of the abscissa axis in Fig. 4 in terms of energy is obtained by comparing the heights of pulses produced by \(\pi\)-mesons passing through the glass with the calculated ionization energy losses of 240 MeV. Such a calibration was repeated every day. The standard for annihilation pulses is the smooth curve in Fig. 4, representing the pulse-height distribution for the annihilation of antiprotons entering the lead glass when \(S_1\), \(S_2\), and the lead plate have been removed. Comparison of the solid histogram with this antiproton curve confirms the assumption that the solid histogram was produced by the annihilation of antineutrons.

Fig. 5. Pulse spectrum in the Cherenkov counter for \(\pi\)-mesons (dashed curve) and for protons (solid curve). The curves are normalized.

For comparison with the annihilation spectra of Fig. 4, Fig. 5 shows the spectra obtained when protons (solid curve) with energy 750 MeV and negative \(\pi\)-mesons with energy 600 MeV are incident on the glass counter \(C\). These spectra indicate that large pulses are rarely produced by particles of such energies. It follows that even high-energy neutrons could not have produced a spectrum similar to the solid histogram in Fig. 4.

To determine the number of \(\gamma\)-quanta incident on \(S_1\), the lead between \(S_1\) and \(S_2\) was removed. The number of neutral events falling in this case on the incident antiproton increased by a factor of 7. With the known probability that an individual high-energy \(\gamma\)-quantum will pass through one inch of lead without absorption (3% for a \(\gamma\)-ray with energy 300 MeV), the observed increase shows that, before the selection according to pulse height in \(X\), at least 20% of the observed neutral events were due to the background of \(\gamma\)-rays. The lead-glass counter \(C\) is very sensitive to \(\gamma\)-rays and insensitive to the ionization losses of slow particles. The desirability of comparing the spectra of antineutrons and antiprotons obtained with detectors of completely different types led us to repeat the experiment with the counter \(C\) replaced by a liquid scintillator. This scintillator, 28 inches thick and 5 cubic feet in volume, was sufficiently large to

Fig. 6. Spectrum of pulses from antiprotons in the large scintillator. The dashed curve refers to all antiprotons. The solid curve refers only to antiprotons that underwent nuclear interactions, and includes a correction allowing this curve to be used for comparison with the spectrum of pulses from antineutrons.

to record a substantial part of the energy released in annihilation. For this experiment the thickness of the lead absorber between \(S_1\) and \(S_2\) was increased to 1.5 inches. As before, the antineutron detector was calibrated by antiproton annihilation. The distribution of pulse heights from antiproton annihilation in the large scintillator is given in Fig. 6. Antiprotons that did not undergo a nuclear interaction form a sharp peak.

The solid curve in Fig. 7 is identical with the solid curve of Fig. 6, containing a correction for the ionization energy losses by antiprotons before they enter the scintillator (the distribution is shifted by 70 MeV toward lower energies). After selection (with the previous criterion for the magnitude of the pulse in \(X\)), 60 events were obtained (Fig. 7). The spectra thus obtained from neutral particles and the spectrum from antiprotons are in agreement, although not as precisely as in the case of the detector with lead glass. The 60 selected events undoubtedly contain some extraneous background. This is confirmed by the form of the spectrum in \(X\) for all neutral events (Fig. 8). In this case there are many more secondary neutral particles, arising in inelastic collisions of antiprotons with nuclei, than in the experiment with lead glass, and the separation of the background pulses from the pulses caused by antineutrons is worse. The larger number of observed neutral secondary particles is probably determined by the greater sensitivity of the scintillator to neutrons.

Fig. 7. Spectrum of pulses produced in the large scintillator by neutral particles. The solid histogram refers to 60 antineutrons (pulse magnitude in scintillator \(X\) less than 100 MeV). The dashed histogram includes 65 pulses caused by other neutral particles. The solid smooth curve has been transferred from Fig. 6.

Fig. 8. Spectrum of pulses in scintillator \(X\) for 125 pulses in the large scintillator caused by neutral particles. The smooth solid and dashed curves are the same as in Fig. 3. Both curves are normalized to the histogram.

Lead glass and the scintillator have almost the same efficiency for registering antineutrons. The observed yield from a \(20\ \mathrm{g/cm^2}\) target is \(0.0030 \pm 0.0005\) antineutrons per antiproton in the case of lead glass and \(0.0028 \pm 0.0005\) in the case of the liquid scintillator.

Losses in the detector efficiency due to attenuation of the flux in \(S_1\), \(S_2\), and in the lead absorber, as well as due to the passage of particles through the detector without interaction, can be calculated on the assumption that the interaction cross section for antineutrons is the same as for antiprotons, and are found to be approximately \(50\%\). From the observed antineutron yield it follows that the mean free path of antinucleons with respect to charge exchange is about \(2300\ \mathrm{g/cm^2}\) of toluene \((\mathrm{C}_7\mathrm{H}_8)\), or, in other words, that the charge-exchange cross section is approximately \(2\%\) of the annihilation cross section for this material. This corresponds to a cross section in carbon of approximately \(8\ \mathrm{mbarn}\).

References

1. Chamberlain, Segrè, Wiegand and Ypsilantis, Phys. Rev. 100, 947 (1955).
2. Brabant, Cork, Horwitz, Moyer, Murray, Wallace and Wenzel, Phys. Rev. 101, 498 (1956).
3. Chamberlain, Chupp, Ekspong, Goldhaber, Lofgren, Segrè, Wiegand, Amaldi, Baroni, Castagnoli, Franzinetti and Manfredini, Phys. Rev. 102, 921 (1956).

1. The difference between particles and antiparticles appears in interactions of a non-electromagnetic nature. The neutron and antineutron must also differ, however, in the sign of their interaction with the electromagnetic field, since the neutron has a magnetic moment, and the signs of the magnetic moments of a particle and an antiparticle are opposite (the magnetic moment of the neutron is negative, i.e. directed opposite to the spin, whereas the magnetic moment of the antineutron is positive—directed in the same direction as the spin). The interaction of an antinucleon with an antinucleon must, to a high degree of accuracy (to the accuracy of the so-called weak interactions), be exactly the same as the interaction of nucleons with one another. However, the interaction of a nucleon–antinucleon may differ, and substantially, from the interaction of nucleons with one another. This is because, when they meet, a nucleon and an antinucleon can annihilate, i.e. transform into other particles—\(\pi\)-mesons—or, with considerably smaller probability, into electromagnetic-field quanta. The annihilation process, thanks to which antinucleons were discovered, can occur not only actually but also virtually: a nucleon and an antinucleon disappear only for a very short time and then “are born again.” In this virtual annihilation process they can “exchange” momenta, position in space, and also spin orientations. Thus, annihilation evidently leads to an additional exchange interaction characteristic of the nucleon–antinucleon system and absent in the interaction of nucleons with one another. Experimental detection of antinucleons makes it possible subsequently to study quantitatively the process of their annihilation and the specific additional interaction caused by it—this is the chief interest of experiments with antinucleons. ↩