THE UNIVERSITY OF MICHIGAN COLLEGE OF LITERATURE, SCIENCE, AND THE ARTS Department of Physics Technical Report No. 1 STATUS OF THE EXPERIMENTAL SEARCH FOR PHYSICAL QUARKS Paper presented at the International Conference on Symmetries and Quark Models, Wayne State University Lawrence W. Jones ORA Project 02507 supported by: NATIONAL SCIENCE FOUNDATION GRANT NO. GP-9332 WASHINGTON, D.C. administered through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR June 1969

Up to the present time, there is no concrete evidence for the existence of physical quarks. I would like to summarize the experimental situation behind this statement with an emphasis on the model-dependent assumptions through which these experiments are interpreted as limits on quark mass and/ or production mechanisms. I shall use the term "quark" throughout to refer to any new massive, stable hadron independent of charge. The searches for quarks fall into three discrete categories: accelerator experiments, cosmic ray searches, and studies of stable matter. If quarks are to be produced in pairs, and each quark has a mass M, the threshold energy, E, for production of a quark pair in a proton-proton collision is E/c2 2(m+M)2 / m (1) where m is the nucleon mass, and the approximation is valid 2 for E >> mc. Hence the threshold energy for making 10 GeV rest mass quarks is about 200 GeV, as an example. The cross section for producing quarks, if they exist, is highly speculative. Adair and Price note that n, K, and p production are consistent with a nu a2 a = h/mc (2) where mi is the appropriate particle mass. If this also holds for quark production, cross sections would correspond to about -1 -

a microbarn for 10 GeV rest mass quarks. The cross section would presumably rise to this value at two to four times the 2 threshold laboratory energy. On the other hand Hagedorn and 3 Feinberg have proposed various statistical or thermodynamic models of particle production at high energies which would give a production cross section falling exponentially with particle mass, a = exp(-M/T) (3) where T is a hadronic matter "temperature" in collisions, typically taken as about 0.16 GeV. If this model has any validity (and it can be made to fit data on p and d as well as r and K production up to 30 GeV), even a 2.5 GeV quark would -35 2 be produced only with a 10 cm cross section and would not be seen in even accelerator experiments. Each additional GeV -5 of quark mass would lower the cross section by 10. While Hagedorn's model was able to fit older data, recently reported d production data from Serpukhov indicate yields of antideuterons significantly in excess of the Hagedorn prediction, so that this calculation may in fact be a pessimistic lower limit. Accelerator searches for quarks are sensitive over the mass range accessible. For pair-produced quarks, this extends to 3 GeV for the Brookhaven A.G.S. Until recently, the most sensitive experiment sets a limit corresponding to d2/(dQdp) < 2x10-36 cm2r-1(GeV/c)-1 -2 -

per nucleus for p of 9 and 10 GeV/c, corresponding to a nucleon-35 2 nucleon quark production a < 1035 cm. The Fermi motion of the nucleons increases the effective threshold to 4 or 5 GeV mass for cross sections one or two orders of magnitude higher (respectively). Other earlier accelerator limits are about 5 two orders of magnitude poorer'. Two recent accelerator experiments have significantly lowered the cross section for possible quark production. At the CERN P.S. Allaby et al. have set limits corresponding to d 2/dQdp s 7.2x10-39 cm2sr -(GeV/c)1 2 -38 per nucleon for a charge of -1/3e, and d a/dQdp 5 5.2x10 8 2 -l -I cm sr (GeV/c) for a charge of -2/3 e (90% confidence level). The technique employed a beam channel tuned to momenta greater than the maximum beam momentum for integral charge and to 10.9 GeV/c for q = 1/3e. The proton beam energy of 27 GeV corresponds to a pair-produced quark threshold of 2.7 GeV without a Fermi momentum contribution. The cross section limit corresponding to the q = -1/3e limit above for isotropic quark pair production in NN reactions, a. 4.5x10 4cm, lies below the Hagedorn calculation for quark masses of about 2.4 GeV and below. The Serpukhov 70 GeV accelerator has recently been used 7 to extend the limit on mass to 5 GeV For q = -2/3e, d2a/dQdp ~ 8xlO'36 cm sr (GeV/c) per nucleon, or a < 4x10-37 cm. The corresponding limit for q = -1/3e is a S lxlO 35cm2. -3 -

Another accelerator search, particularly relevant in view of Hagedorn's gloomy production cross section prediction, was 8 carried out at SLAC. Here electromagnetic pair production was studied, and limits set on mass and charge of possible unknown particles. The limits in this case depend upon only wellestablished electrodynamics. The mass upper limits set range from 0.2 GeV for O.04e charge to 1.5 GeV for 0.7e charge (assuming the particle lifetime exceeds 10-7 sec). In order to discuss the cosmic ray searches, it is first necessary to review the flux and energy spectrum of cosmic rays entering the earth's atmosphere. To an uncertainty of about a factor of two, the integral vertical proton flux in the earth's atmosphere can be represented by -1.8 N(E,y) = 1.5E1 exp(-y/x) (4) where N is the number of protons of energy greater than E GeV per (cm sr sec), y is the depth from the top of the atmosphere in gm cm and Xa is the attenuation mean free path of nucleons -2 in air; 120 gm cm. Thus at the top of the atmosphere the integral flux of cosmic ray protons of E > 100 GeV is about 4x10 3(cm2sr sec) and this integral flux falls by about a factor of 10 for each factor of 3 in energy. At sea level the corresponding proton flux (E > 100 GeV) has fallen to 10-7(cm sr sec). Cosmic ray production of quarks consequently would be dominated by events high in the atmosphere and would probably be produced at energies close to threshold. Adair and -4 -

Price have calculated cosmic ray quark production vs. quark mass for an assumed one ub production cross section. Their result is reproduced in Fig. 1. Cosmic ray searches for quarks have taken several forms. In the largest number of experiments, a large counter "telescope" has been built to search for relativistic particles of fractional 9-15 charge (q of ~l/3e or ~2/3e) by measurement of ionization915 In order to minimize problems arising from Landau straggling of ionization loss, small pulses due to Compton electrons, and many other "dirt" effects, a large number (as many as 12) of independent ionization detectors or counters are used in each experiment. In addition, spark chambers or other track visualizing devices have been used to verify the acceptability of quark candidates. The experiments typically have a phase space admittance of 1/10 to 1 m sr and are operated for a period of months. From these many searches an upper limit to the quark flux is about 10 0(cm2sr sec). In Table I the results of some recent experiments are summarized. It has taken a very considerable effort to drive the limit from 10-9 to 101 2 -1 (cm sr sec). In view of the effort required and the negative results to date, I doubt if any group will soon push this limit another order of magnitude lower. These searches are limited in their sensitivity by several factors. 1) Obviously they would not detect integrally-charged quarks. Neither would they normally detect charges greater than unit charge (e.g., q = 4/3e) should such states be the most -5 -

stable form of quark "matter." 2) Quarks might only be produced in very high multiplicity reactions and might be accompanied (even at the earth's surface) by muons or electrons at lateral spacings of only tens of centimeters. These normal ionizing particles would then mask the quarks. 3) If quarks interact with nuclear matter_ in such a way that they are rapidly slowed down in the atmosphere through inelastic collisions, they might arrive at the earth's surface with v << c and hence an ionization measurement would not be definitive. This is unlikely if quarks interact like other hadrons. Known strong interactions are characterized by an average four-momentum transfer of about 0.5 (GeV/c)2. For a nucleon, this corresponds to an inelasticity (fraction of energy loss in a collision to incident energy) of about 50%. For a 10 GeV rest-mass quark, the same fourmomentum transfer would correspond to less than a 10% inelasticity. Hence we would generally expect quarks to reach the earth's surface with a much greater fraction of the energy they had at production than the corresponding nucleons, even assuming the same interaction cross section. A review of earlier accelerator and fractional charge cosmic ray quark searches is contained in a CERN report by 16 Massam A second class of cosmic ray search has studied the momentum and velocity of energetic cosmic ray particles with a magnetic spectrometer and time-of-flight counters. Kasha and Stefanski17 -6 -

have set a flux upper limit of 2.4x10 8(cm2sr sec)- on particles of up to 300 GeV/c, corresponding to a 2.b production cross section for particles of rest mass between 5 and 15 GeV (following the model of Adair and Pricel). This experiment is of course also sensitive to quarks of integral charge. A somewhat similar search measured range and velocity of particles, seeking evidence for particles of m > mp and/or Iql < |el, using p 18 range (in iron) and liquid threshold Cherenkov counters. Two possible deuterons were detected corresponding to a flux of -12 2 -1 1.3xJ10 (cm sr sec), or an incident sea level flux of 4.8x1 10 (cm2sr sec) -, correcting for absorption in the apparatus. These figures may be regarded as effective upper limits to quark fluxes. A third class of searches sensitive to quarks of integral or non-integral charge follows a method proposed by the Copenhagen 19 group, and is illustrated in Fig. 2. If a cosmic ray proton produces a quark of 100 GeV with a rest mass of 10 GeV along with a number of other particles (pions, etc.) which give rise to a typical cosmic ray shower, a detector located some kilometers below the production event will observe the quark arrival delayed relative to the v = c shower particles by a readily measurable time interval. This interval is given by 2 t = y/2y c where y is the distance between production and detectors and Y (assumed >>1) is the Lorentz factor of the quark. For -7 -

y = 2km and Y = 10, t = 33 ns. The experiments consist of an array of counters to detect showers and a separate detector for the quark (for example a stack of counters with absorbers between them for sampling the energy loss). Three experiments have 20521522 16 21 been published20222. The Copenhagen group and our group 21 report essentially negative results. In our experiment21 the quark detector was an ionization calorimeter: a scintillationcounter-absorber configuration capable of giving a rough (~20 or 30%) energy determination of any high energy hadron. One quark candidate was seen, although the probability is about 10% that it was a nucleon or otherwise spurious trigger. Because of the time delay technique, the limit to the quark flux depends on quark mass and on assumptions concerning the quark behavior in the atmosphere. The upper limit to the quark flux set in our experiment is given in Table II and Fig. 3. Also given is the quark production cross section upper limit for different assumed masses and for two extreme assumptions on quark interactions: 1) Xh = Xa (quarks attenuated like nucleons in the atmosphere) and 2) Xh = ( (quarks are totally unattenuated). The limit of 5x10 10(cm2sr sec)1 is a somewhat poorer limit than set by fractional-charge telescopes. However the searches are complementary in that these time delay experiments are sensitive to integrally-charged quarks, and to quarks attenuated in the earth's atmosphere (our experiment was performed at 10,600 ft. elevation). These experiments must assume that quarks are -8 -

produced with other particles and not at large transverse momentum with respect to the other secondaries (otherwise the shower particles would not arrive at the same laboratory as the quark). The experiment of Dardo et al.22 is somewhat of an enigma. They see a significant counting rate for delayed events in an underground detector. This would have shown up in our exper21 iment as a signal 100 times our upper limit. They suggest that the essential difference with our experiment is that our trigger was not sensitive to a "thin" or muon shower while theirs was, while in fact the trigger requirements were very similar. From my own personal bias and based on the experiment I know best, I am very skeptical of the results they report. Consequently I would assert that the class of time delay experiments gives no positive evidence for existence of physical quarks. Finally, quarks have been sought in stable matter of the earth's crust. In order to relate the limit set here to the cosmic ray data a very brief ourder of magnitude estimate is useful. We may assume that the cosmic ray flux has rained on the earth at the present rate for 101 years (3x1017 sec). Quarks produced by cosmic rays would mostly fall into the oceans and be mixed through sea water over this time. They would then be distributed through an average of about 2km (2xl05g/cm2) of matter. If the vertical quark flux produced by cosmic rays is cp(cm2sr sec) there will be a dinsity of quarks in matter, -9 -

p, of 3xlO17 5X112 quarks 2x 10 gm or p = 8x10 cp quark/nucleon. A cosmic ray quark flux of 10 (-cm2sr sec)- thus would -21 correspond to a quark density in stable matter of about 102 quarks per nucleon. It is unnecessary to emphasize the speculative nature of this estimate; it is only useful orientation. The searches in stable matter involve mass spectrometric methods, magnetic levitation and Millikan oil drop experiments, and optical spectroscopy. The mass spectroscopic studies have -17 -29 set limits of 10, 5x107 and 3x10 quarks per nucleon in iron meteorites, air, and sea water respectively 3. These limits are strongly model dependent,(for example they depend strongly on the assumed concentration of quarks in samples through various means) and are generally regarded as optimistic. A number of experiments were initiated during a period when some cosmologists predicted an inequality of proton and electron charge. These experiments were of course stimulated by the quark excitement. Basically, they are refinements of the Millikan oil drop experiment, wherein a non-integral charge is sought on oil droplets, graphite grains, or superconducting beads. In the latter two cases diamragnetic particles are suspended in a static magnetic field and their displacements studied in a horizontal electric field. Using the oil droplet -10 -

-20 technique, Rank has set limits of quark concentration of <102 24 - per nucleon. Morpurgo has set limits of <5xl0 19 per nucleon 18 25 of graphite (he has seen no quarks in 2x10 nucleons) Johnston and Franken, using superconducting niobium pellets, -19 26 have set limits <1019 per nucleon. 24 In a series of spectroscopic experiments, Rank2 has looked for the Lyman series hydrogen-like spectral lines of "quarkogen"; an electron bound to a +2/3 quark (or a -1/3 quark-proton nucleus). He has studied sea water, fresh water and possible biologically concentrated sources -- oysters, sea weed and plankton. The "quarkogen" is concentrated using electric fields over vaporized sources. The limits set by -18 these experiments is <10 quarks per nucleon. In conclusion, I believe that it is fair to say that the excitements of quark hunting is dying away. I know of no new cosmic ray searches being initiated. Each new, higher-energy accelerator will surely mount a serious quark search, and I am sure that someone must be planning to look into moon dirt for stable quarks. At this time, I suspect that most experimentalists feel that physical quarks are either unobservable or do not exist. -11 -

References 1. R. K. Adair and N. J. Price, Phys. Rev. 142, 844 (1966). 2. R. Hagedorn, Nuovo Cimento Supplements VI, 311 (1968). 3. D. S. Chernavsky, E. L. Feinberg, and I. N. Sissakian "Heavy-pair production with application to the problem of quark search," P. N. Lebedev Physical Institute, Moscow (preprint, 1966), and JETP 52, 545 (1967). 4. D. E. Dorfan, J. Eades, L. M. Lederman, W. Lee and C. C. Ting, Phys. Rev. Letters 14, 999 (1965). 5. P. Franzini, B. Leontic, D. Rahm, N. Samios, and M. Schwartz, Phys. Rev. Letters 17, 196 (1965). 6. J. V. Allaby, G. Bianchini, A. N. Diddens, R. W. Dobinson R. W. Hartung, E. Gygi, A. Klovning, D. H. Miller, E. J. Sacharidis, K. Schlupmann, F. Schneider, C. A. Stahlbrandt, and A. M. Wetherell, CERN preprint (submitted to Muovo Cimento, June, 1969). 7. Yu. M. Antipov, I. I. Karpov, V. P. Khromov, L. G. Landsberg, V. G. Lapshin, A. A. Lebedev, A. G. Morosov, Yu. D. Prokoshkin, Yu. V. Rodnov, V. A. Rybakov, V. A. Rykalin, V. A. Senko, B. A. Utochkin, N. K. Vishnevsky, F. A. Yetch, and A. M. Zajtzev, Phys. Letters 29B, 245 (1969). 8. E. H. Bellamy, R. Hofstadter, W. L. Lakin, M. L. Perl, and W. T. Toner, Phys. Rev. 166, 1391 (1968). 9. H. Kasha, L. B. Leipuner, and R. Adair, Phys. Rev. 150, 1140 (1966). -12 -

10. R. Gomez, H. Kobrak, A. Moline, J. Mullins, C. Orth, J. VanPutten, and G. Zweig, Phys. Rev. Letters 18, 1022 (1967). 11. A. Buhler-Broglin, P. Dalpiaz, T. Massam, and A. Zichichi Nuovo Cimento 51A, 837 (1967). 12. F. Ashton, R. B. Coates, G. N. Kelley, D. A. Simpson, N. I. Smith, and T. Takahashi, Proc. Phys. Soc. A 2, 569 (1968). 13. Y. Fukushima, T. Kifune, T. Kondo, M. Koshiba, Y. Naruse, T. Nishikawa, S. Orito, T. Suda, K. Tsumemoto, and Y. Kinura, Phys. Rev. 178, 2058 (1969). 14. E. P. Krider, T. Bowen, and R. M. Kalbadi, reported at the Midwest Cosmic Ray Conference, Louisiana State University, (March 1969). 15. R. C. Lamb, R. A. Lundy, T. B. Novey and D. D. Jovanovic, Phys. Rev. Letters 17, 1068 (1966). 16. T. Massam, "The Quark Hunter's Progress," CERN 68-24 (8 July 1968). 17. H. Kasha and R. J. Stefanski, Phys. Rev. 172, 1297 (1968). 18. F. Ashton, H. J. Edwards, and G. N. Kelly, Phys. Letters 29B, 249 (1969). 19. G. Damgaard, P. Greider, K. H. Hansen, C. Iverson, E. Lohse, B. Peters, and T. Regarajan, Phys. Letters 17, 152 (1965). 20. J. Bjornboe, G. Damgaard, K. Hansen, B. K. Chatterjee, P. Grieder, A. Klovning, E. Lillethun and B. Peters, Nuovo Cimento 53B, 241 (1968). -13 -

21. L. W. Jones, D. E. Lyon, Jr., P. V. R. Murthy, G. DeMeester R. W. Hartung, S. Mikamo, D. D. Reeder, A. Subramanian, B. Cork, B. Dayton, A. Benvenuti, E. Marquit, P. D. Kearney A. E. Bussian, F. Mills, C. Radmer, and W. R. Winter, Phys. Rev. 164, 1584 (1967). 22. M. Dardo, P. Penengo, and K. Sitte, Nuovo Cimento 58A, 59 (1968). 23. W. A. Chupka, J.P. Schiffer, and C. M. Stevens, Phys. Letters 17, 60 (1966). 24. D. M. Rank, Phys. Rev. 176, 1635 (1968). 25. G. Gallinaro and G. Morpurgo, Phys. Letters 23, 609 (1966), and private communication (1969). Also: G. Morpurgo, G. Gallinaro, and G. Palmieri,"The Magnetic Levitation Electrometer and its Use in the Search for Fractionally Charged Particles." (Preprint)(25 June 1969). 26. P. Franken (private communication). -14 -

Table I Cosmic ray limits on quark fluxes set by some experimental searches for particles of fractional charge. In almost all cases limits correspond to 90% confidence levels. 2 Group Flux limit in particles per (cm sr sec) 1 2 4 q = = 3e q = e q -3e 3 3 3 Brookhaven Yale a 2.6xlO- 9 2.0x10- 9 Argonneb 4.5xl0-10 1.6xl0-9 CERN 4.5x1010 1.7x101 1.6x0 8 Cal Techd 1.7xl0 10 2.0xl09 eurham" -10 -11 Durhame 1.15x1010 8.0x1 Tokyof 0.5xlO 10 7.5xl0-10 -11 10 Arizonag 6.8 0 ll 1.2xl010 a Reference 9 b Reference 15 c Reference 11 d Reference 10 e Reference 12 f Reference 13 g Reference 14 -15 -

Table II Quark flux and production cross section upper limits for different assumed quark masses and for two assumptions on quark mean free paths, Xh' in the atmosphere set by the experiment of Reference 17. Quark production Quark flux Quark cross section upper limit upper limit Mass (99% confidence level) (90% confidence level) (GeV) in 10-30 cm2 per nucleon in (cm dr sec) Xh = 5 0.10 3.2xl1-9 7 0.11 1.2x109 0o 0.16 5.0x1010 14 0.32 3.3xlO10 -10 20 1.57 4.5xlO = 120 gm cm 5 1.8 8.8xl010 7 3.1 5.2xl101 10 8.3 4.0xlo10 14 31.1 4.8xlo-10 20 263 1.2xlO-9 -16 -

Figure Captions Figure 1. Cosmic ray-produced flux of quarks at sea level with p/Mc > 1 for an asymptotic production cross section a = 10 c30r! as a function of quark mass. The flux is in units of particles per (cm sr sec) (after Adair and Price ). Figure 2. Time-delay method of searching for quarks as used in experiments of References 20, 21 and 22. Figure 3. Upper limits (99% confidence level) to the cross section for the productionof quarks (in pairs) in nucleon-nucleon cosmic ray collisions set by the results of the experiment of Reference 21. Curves are given for two assumed quark attenuation mean free paths; Xh = 120 g cm -and kh = -17 -

' 3x108 c8 w 10- A 0. -LJ H 3x iO-9 _ __ II iL S 109- \ - x |o"-I0 I II I I 5 10 15 20 25 PARTICLE MASS, M (GeV) Fig. 1 -18 -

SYSTEM FOR TIME DELAY DETECTION OF MASSIVE PARTICLES COSMIC RAY PROTON OF " 1000 GeV \' I// COLLISION WITH /Ii"\ AIR NUCLEUS RELATIVISTIC /1\ SECONDARIES/ ) (MOSTLY E-M / CASCADES) / / / 3/ / / t I \ POSSIBLE,/ / / ~ MASSIVE /i, / / I\P \ PARTICLE y - / / ) \ / / / 1 / / / ' I / =S: SHOWER S D Is COUNTERS 0D MASSIVE Ie v >CI I ~J --— PARTICLE -- - T.HC. — ADC DETECTOR FOR: Y= 2 km t=-33 ns Y=10 Fig. 2 -19 -

1000.0 z + -J Z 100.0 - Zt L X120gm/cm z -h z 0 5 10.05 20 ~ +-20 - 0 - 1.0-> 0 + o 0 0A 5 10 15 20 MASS IN Bev -20 -