Prof. Dr. Johann Rafelski
University of Arizona, Department of Physics

The Discovery Path of Quark-Gluon Plasma
(QGP, Quark Matter, New Phase of Matter)

What is quark-gluon plasma?

In a nutshell, QGP (also known as Quark Matter) in the contemporary use of the language is an interacting localized assembly of quarks and gluons at thermal (kinetic) and not necessarily chemical (abundance) equilibrium - heavier quark flavors such as strangeness and charm typically approach chemical equilibrium in a dynamic evolution process in present day experiments. The word plasma signals that color charged particles (quarks and/or gluons) are able to move in the volume occupied by the plasma. Since the temperature is above Hagedorn temperature Th=150 MeV and thus above the scale of light quark u,d -mass, the pressure within the hot QGP fireball created in laboratory exhibits the known Stefan-Boltzmann format, driving a rapid explosive fireball expansion.

This QGP fireball is in the so-called (color) deconfined phase of matter. This form of matter preceded in time in the early Universe the matter that we know today. For this reason QGP is often referred to as the "new" or "primordial" phase of matter and laboratory experiments "revisit" the early stage of Universe evolution, when it was 10-30 microseconds "old".

What is the Hagedorn temperature Th?

The mass spectrum of hadronic particles rises exponentially. Hagedorn temperature Th is the (inverse slope) parameter describing this rise and it is found by fitting the exponential shape to data. The experimental mass spectrum is discrete; hence a smoothing procedure is often adopted to fit the shape. The procedure is also encumbered by a low mass singularity. Our current understanding is that Th has a value
Th=140 - 155 MeV =(16,000 -18,000) Billion K .
Th is the boiling point of quarks and melting point of hadrons, the maximum temperature at which matter can exist in its usual form. Th is not a maximum temperature in the Universe (as many believed before 1980), a further heating of the quark-gluon plasma can and will continue.

The melting of hadrons into quarks occurs more easily at a finite baryon matter density. The higher the density the lower the needed temperature; it is believed that "cold" quark matter can exist for example in the center of neutron stars, a topic of current active research.

How did the term QGP come into use?

In 1987 Leon Van Hove (former scientific director general of CERN - CERN located at the Franco-Swiss frontier near Geneva, is an international particle and nuclear physics research laboratory, its acronym contains C for council, E for Europe, R for research, and N for nuclear) wrote a report entitled "Theoretical Prediction of a New State of Matter, the Quark-Gluon Plasma (also called Quark Matter)" establishing the common physics foundation of these three terms:
New State of Matter=Quark-Gluon Plasma (QGP) =Quark Matter.
However, using the phrase Quark Matter we can be misunderstood as referring to the zero-temperature limit applicable to the situation in the core of neutron stars. That is why when we speak of the new phase of matter that filled the primordial Universe for the first 20 microseconds, and which we recreate today in laboratory, the current preferred term is Quark-Gluon Plasma (QGP).

The reader seeking references will focus on these key words. S/He should remember that the language has evolved; originally QGP meant something else. In physics when naming an important new insight, older terms are recycled: for example Ed Shuryak called QGP the Feynman parton gas content of an untouched nucleon that a relativistic electron sees moving at infinite speed undergoing a deep inelastic scattering. Clearly this is not the QGP we form in laboratory disrupting and destroying and fusing the nucleons in a nucleus. The modern use of the term QGP begins in Summer 1979 when research work describing the strongly interacting quark-gluon thermal matter appears for the first time under QGP title, later published in PLB..

How did relativistic heavy ion collisions and QGP come together?

With the hindsight of 450 years I believe that Hagedorn's effort at CERN to interpret within a thermal model the particle production data in pp collisions led to the original recognition that there is an opportunity of studying quark-gluon deconfinement at high temperature. How this came together is the topic of the book: Melting Hadrons, Boiling Quarks; From Hagedorn temperature to ultra-relativistic heavy-ion collisions at CERN; with a tribute to Rolf Hagedorn published at Springer Open 2016, available for free to read on-line.

In short: before 1978/9, those working in the field of relativistic heavy ion collisions were advancing their field as an opportunity to study compression of nuclear matter to conditions prevailing in neutron stars. The idea that particle production could help achieve a high enough particle density to allow deconfinement was recognized in 1978/9. Hagedorn-Rafelski and nearly in parallel several other groups recognized that such matter made of hadrons would melt into a boiling quark-gluon plasma phase. In the following decade experimental facilities were developed, by CERN in Geneva and by Brookhaven National Laboratory (BNL) on Long Island, NY, to search for this new phase of matter.

What happens to QGP formed in laboratory?

QGP is formed in a small spatial domain as a hot fireball that explodes. In the explosion, temperature decreases and QGP undergoes the process of hadronization; that is, disintegration into individual hadrons. A method of data analysis called Statistical Hadronization Model (SHM) was invented to characterize how a blob of primordial QGP falls apart into individual hadrons. This hadronization process can occur below Th (supercooling) and particles can be emitted before QGP falls apart, so the experimentally observed T after we account for motion of exploding matter and the Doppler blue shift does not need to exactly match the same value of Th; however comparing different fireball sizes the actual analysis are not far apart. This indicates that in present day experiments a large enough fireball is formed so that the terminal speed of expansion is nearly the same.

Statistical Hadronization Model=SHM relies on the hypothesis that a hot fireball made of building blocks of future hadrons populates all available phase space cells proportional to their respective size, without regard to any additional interaction strength governing the process. Such an approach makes sense especially in the context of a strong interaction that produces only "neutral" final state particles: all color charges neutralize; free quarks and gluons are not observed.

In earlier models by Koppe, Fermi, Landau, and Hagedorn before 1978, the particles emitted were both point-like and not newly formed; they were seen as already being the constituents of the hot, dense fireball. The absence of size was particularly damaging for understanding of the physics situation as it suggested that the Hagedorn temperature Th is the maximum temperature in the Universe. Moreover, the availability of large hadron abundance without need to be created meant that these models we today call freeze-out models had no dynamical features.

Unfortunately, some authors still adhere to such precursor models 50 years late and after we found more appropriate understanding that merges the quark structure of hadrons with the statistical properties of strongly interacting matter. The reader has an easy way to recognize such grandfathered thinking: if a research publication that attempts to describe the data speaks of hadronic (produced particle) equilibrium by assumption, I recommend caution.

Today we extract from the produced hadron spectra and abundances information about the QGP hadronization condition. This helps us understand, for example, the conditions of QGP stability in the Universe. Understanding the ways by which the matter we know was produced can perhaps one day help us comprehend why matter is stable. The relativistic nuclear collision experiment provides, within a well-defined collision class, yields of many particles.

In the SHM chemical analysis we focus on integrated normalized particle spectra, the particle number-yields. The reason that such data are of particular interest is that particle yields are independent of local matter dynamics,; that is, velocity in the fireball surface, which imposes spectra deformation one often refers to as Doppler blue shift. If the experimental data is fragmentary and the spectral coverage is not full, an extrapolation of spectra needs to be made, introducing uncertainty into the data analysis. This is the primary reason it took such a long time to establish in the mind of "everybody" that QGP has been created in laboratory experiments, even though the key predicted signature - enhanced abundance of strange antibaryons - has been reported since 1990. Many relevant details how strangeness signature of QGP works were presented in the OPEN review: Discovery of Quark-Gluon Plasma: Strangness Diaries (EPJ-ST Volume 229, number 1, pp1-140) (EDP Sciences/Springer Open Access January 2020)..

When, where, and how was QGP discovered?

Both CERN and BNL held press conferences describing their experimental work. CERN advertised its position in a February 2000 event/press release. The document for scientists agreed to by those representing the seven CERN experiments said:
"The year 1994 marked the beginning of the CERN lead beam program. A beam of 33 TeV (or 160 GeV per nucleon) lead ions from the SPS now extends the CERN relativistic heavy ion program, started in the mid eighties, to the heaviest naturally occurring nuclei. A run with lead beam of 40 GeV per nucleon in fall of 1999 complemented the program towards lower energies. Seven large experiments participate in the lead beam program, measuring many different aspects of lead-lead and lead-gold collision events: NA44, NA45/CERES, NA49, NA50, NA52/NEWMASS, WA97/NA57, and WA98. .....
Physicists have long thought that a new state of matter could be reached if the short range repulsive forces between nucleons could be overcome and if squeezed nucleons would merge into one another. Present theoretical ideas provide a more precise picture for this new state of matter: it should be a quark-gluon plasma (QGP), in which quarks and gluons, the fundamental constituents of matter, are no longer confined within the dimensions of the nucleon, but free to move around over a volume in which a high enough temperature and/or density prevails. ....
A common assessment of the collected data leads us to conclude that we now have compelling evidence that a new state of matter has indeed been created, ....
The new state of matter found in heavy ion collisions at the SPS features many of the characteristics of the theoretically predicted quark-gluon plasma....".

In retrospect, one can say that CERN was scientifically right in reporting that QGP discovery. We know today that the strange antihyperon enhancement evidence of QGP observed at the time by two large experimental groups, WA97 and NA49 experiments, has withstood the test of time (see end of this report). In my opinion, many other results advanced and attached to the CERN announcement by Ulrich Heinz were either not QGP characteristic and/or were later not confirmed. In this way the QGP recognition was I believe intentionally delayed. This situation is also further discussed in above quoted Discovery of Quark-Gluon Plasma: Strangeness Diaries .

Strangeness, that is strange quark flavor production in relativistic heavy ion collisions is a signature and a diagnostic tool of QGP. The strange (anti)quark signatures of QGP was proposed in 1980, it relies on the fact that strange quarks and antiquarks are not present in matter since within a short time they undergo "weak" interaction decay, akin to naturally radioactive isotopes. However, in a hot QGP environment, strange quarks can be produced in a large abundance and subsequently due to their natural radioactivity are relatively easily observed. The ensuing production of matter particles comprising strangeness and especially MULTI STRANGE ANTIBARYONS is the gold standard signature for the formation of QGP.

A signature of anything requires a rather background free environment, and a good control of anything that is there, as no signature is truly background free. This is a reason why strange antimatter; that is, multi-strange antibaryons, stand out so strongly. In the absence of QGP their production is insignificant. Only QGP provides a ready-made reservoir of strange quark pairs from which strange antibaryons are formed. People with a significant voice at CERN, such as Hagedorn, Jacob, and van Hove, understood this concept very well. Therefore, it was no accident that the CERN-SPS research program included as a large part the exploration of the predicted strange (anti)baryon enhancement.

In CERN experimental line WA85, WA94, WA97 and NA57, the strange antibaryon enhancement effect was found, and it showed the predicted QGP pattern. Working with a different apparatus less suitable for multi-strange particles but able to survey the general pattern of strangeness production, the experimental line NA35, NA35II, and NA49 found perplexing results early on and did not report their meaning. More on this topic follows below.

The production of hadrons made entirely from newly created quarks was found at CERN to be up to 15-20 times more abundant in nuclear reactions when compared to up-scaled reference measurement. This enhancement falls with decreasing strangeness content and increasing contents of the valence quarks which are brought into collision. The pattern of enhancement follows the QGP prediction and is now at a level greater than 10 s.d. There is no known explanation of all experimental results other than QGP formation. This is also the largest medium effect observed in RHI collision experiments to date.

These discoveries made at CERN 1990-2000 are now quarter century old. They have been confirmed and extended by further results obtained in following years at CERN SPS, at the Brookhaven National Laboratory (BNL) RHI Collider (RHIC), and at the CERN Large Hadron Collider (LHC); in 2017 Nature Physics devoted its cover page to the subject publishing recent LHC results. Strange antibaryon results span a range of collision energies and participating baryon number a range that differs by a factor of nearly 200, and yet they are remarkably similar, showing that universal processes present in QGP hadronization is the mechanism of particle production.

Let us look back what happened after CERN announced that a new state of matter was discovered, note a carfully crafted low-key description of QGP as a new state of matter: In February 2000 the highly influential Director of BNL, Dr. Jeff Marburger, (a later long term Presidential Science Advisor, before President of Stony Brook campus of the NY State University System) called these remarkable results and other CERN-ion experimental results, paraphrased: pieced together indirect glimpse of QGP. Today this Marburgerism was shown to have been inaccurate: the strange antibaryon results were reconfirmed during the past 15 years of work, also at BNL, and can be said to be a direct, full panoramic sight of QGP. There is nothing more direct, spectacular, and arguably more convincing that we have seen as evidence of QGP formation in RHI collision experiments.

It is important to remember that this initial skepticism that is vibrant in remarks of Dr. Marburger became quickly infectious; in the year 2000 this point of view was soon widespread within the particle and nuclear physics communities. This pressure created a turncoat: One of the two authors of the CERN scientific QGP discovery consensus report, Ulrich Heinz, succumbed to Marburgerism and declared a few months down the road in 2000/01 that he was mistaken, no longer certain that QGP was discovered!

Dr. Marburger left to join the younger Bush administration in 2001 and was considered poorly by majority of US scientific community as we read in the opening of the Wikipedia article on the subject: "His tenure was marked by controversy" regarding his defense of the administration against allegations from over two dozen Nobel Laureates, amongst others, that scientific evidence was being suppressed or ignored in policy decisions, this is excatly what John Marburger did to QGP so that his BNL laboratory could discover it a few years later anew. BNL’s turn to announce its QGP work arrived 5 years after CERN’s, Marburger presidential office spirit permeated the April 2005 meeting of the American Physical Society, held in Tampa, Florida, where “our discovery” press conference took place on Monday, April 18, 9:00 local time.

The 10 year anniversary was relived at the 2015 BNL meeting. Berndt Mueller, the 2015 Brookhaven's Associate Laboratory Director for Nuclear and Particle Physics, is quoted as follows: "RHIC lets us look back at matter as it existed throughout our universe at the dawn of time, before QGP cooled and formed matter as we know it .... The discovery of the perfect (quark matter) liquid was a turning point in physics, and now, 10 years later, RHIC has revealed a wealth of information about this remarkable substance, which we now know to be a QGP, and is more capable than ever of measuring its most subtle and fundamental properties." Note that Berndt Mueller was my dear collaborator on pivotal 1982 and 1986 strangeness signature publications, but as director at BNL he was celebrating in 2015 only the key contribution of BNL reported in 2005. What one must remember is that ideal fluid behavior of the new phase of matter does not pinpoint any characteristic QGP property, so Berndt is right to put "quark matter" in parenthesis: unless he looks at strangeness he cannot tell what BNL discovered.

The Obstacle Race to QGP

At the beginning of this new field of research in the late 1970s, quark confinement was a profound mystery; gluons mediating the strong color force were neither discovered nor widely accepted, especially not among nuclear physics peers, and the QCD vacuum structure was just finishing kindergarten. The discussion of a new phase of de-confined quark-gluon matter was therefore in many eyes not consistent with established wisdom and considered by many too ambitious for the time.

I entered the race shortly after finishing a paper “Charged Vacuum” describing local modifications of the quantum vacuum structure in presence of strong fields. In 1974/5 at Argonne National Laboratory, in conversations with Arthur Kerman and Harry Lipkin, the connection of confinement to vacuum structure seemed natural, especially in view of the rise of what was called the MIT bag model. My publications from the period 1975-77 on dynamical models of quark structure were precursors that led me to the later insights.

However, I needed a different environment, ANL Physics Division was not into multi-particle production, a natural feature recognized already by Fermi 25 years before. As often happens, the choice was made for me. While I was considering moving from a tenure-track position at ANL to a “down-graded” fellowship (a.k.a. postdoc) at CERN, I was encouraged to take a temporary leave as my ANL colleagues saw my work on quarks and confinement as a poor fit to their nuclear science interests. I accepted the CERN offer and began to learn statistical hadron physics from Hagedorn as of September 1977.

RHI collisions aiming at QGP required the use of atomic nuclei at the highest available energy. This required cooperation between experimental nuclear and particle physicists. Their culture, background, and experience differed. A similar situation prevailed within the domain of theoretical physics, where an interdisciplinary program of research merging the three very traditional physics domains (nuclear, particle, thermal) had to form. While ideas of thermal and statistical physics needed to be applied, very few subatomic physicists, who usually deal with individual particles, were prepared to deal with many body questions.

I remember several practical issues: in which (particle, nuclear, stat-phys) journal can one publish and who could be the reviewers (other than direct competitors)? To whom to apply for funding? Which conference to contribute to? All this applied to everyone who entered this entirely new field and our work was scattered and at first hardly noticed. Key papers did not pass the referee process and if not published in conference proceedings seemed lost (that is, until the web was born at CERN).

Despite these common hardship, in the early days the small group of scientists who practiced QGP and RHI collisions were divided on many important questions. In regard to what happens in relativistic collision of nuclei the situation was most articulate and differences the greatest and well documented:
  1. One group was led by Larry McLerran, who believed (in absence of any experimental evidence) that the nuclei made of protons and neutrons pass through each other with a new phase of matter formed in a somewhat heated projectile and/or target.

  2. A few years later a more refined mechanism was recognized by JD Bjorken (which McLerran immediately embraced): a trail of energy connecting the two nuclei was to be the new phase of matter. While one generally accepts this second model as applicable to the CERN LHC collider range, it still remains a theoretical view to be confirmed by measurement. It is not correct at the lower RHIC and SPS energy range.

  3. A point of view advanced by the Frankfurt school of Walter Greiner was that in RHI collisions energy would be consumed by a shock wave compression of nuclear matter crashing into the center of momentum frame.

  4. Another group including this author argued that up to top accessible CERN-SPS collision energy, a high temperature, relatively low baryon density quark matter fireball will be formed from stopped fragments of both projectile and target, and this case turned out to be closest to reality for CERN SPS and some of lower energy precision data runs at Brookhaven National Laboratory (BNL) RHI Collider (RHIC).
Lack of unity of views of the QGP research community was compounded by aggression from outside. Those working on QGP were ridiculed as being too speculative. This added fuel to the state of uncertainty about the fate of colliding matter and the kinetic energy it carried, with disagreements that ranged across theory vs. experiment, and particle vs. nuclear physics. In this situation, QGP formation in RHI collisions was a field of research that could have easily fizzled out by 1985. When we discuss the research work at CERN below, these problems will show up as some of the mechanisms that created obstacles were hard and indeed impossible to overcome within a large experimental collaboration.

Accelerator Race

Institutional support saved the QGP research program: the US and European situation differed and evolved. Let us begin with US and follow up with Europe:

US-RHIC Early on the US research administrators realized that RHI collisions required large accelerator and large experiments, uniting much more human expertise and manpower compared to prior nuclear and even some particle physics projects. Thus work had to be centralized in pan-continental new US facilities. This meant that expertise from a few laboratories would need to be united, preferably at a site where prior investments would reduce the preparation time and minimize the cost.

The roots of the US relativistic heavy ion program go back to 1975: the Berkeley SuperHILAC, a low energy heavy ion accelerator was linked to the Bevatron, an antique particle accelerator at the time, yet capable of accelerating the injected ions to relativistic energies with the Lorentz factor above two. The system of accelerators was called the Bevalac. It offered beams of ions which were used in study of properties of compressed nuclear matter, conditions believed to be similar to those seen within collapsing neutron stars.

As interest in the study of QGP grew the Bevalac scientists formulated the future Variable Energy Nuclear Synchrotron (VENUS) heavy ion facility. When Berkley moved to define the research program for an ultra-relativistic heavy ion collider in 1983, another candidate laboratory was waiting in the wings: The Brookhaven National Laboratory (BNL) had a completed civil engineering for a collider project with 4 experimental halls. This was to be the pp collider ISABELLE, now mothballed, having been scooped by CERN's bet on the SppS collider in the race to discover the W and Z weak interaction mesons. If ISABELLE were modified to be a RHI Collider (RHIC), it was thought that it could be completed within a few years. That seemed a great idea.

This evaluation prompted a rapid major investment decision leading to creation of a new relativistic heavy ion research center at BNL: RHIC. The first data taking at RHIC began in summer 2001. When in 1984 we were told at a meeting at BNL that RHIC was to operate by 1990, a colleague working at the Bevalac asked, why not 1987 or 1988? So a big question remains today: why in the end was it 2001?

Europe-CERN In Europe the QGP formation in RHI collisions research program found its home at CERN. The CERN site benefited from being a multi-accelerator laboratory with a large pool of engineering expertise and where some of the necessary experimental equipment already existed, thanks to prior related particle physics efforts, and where a QGP theory group centered around Rolf Hagedorn was at home.

The CERN program took off by the late 80s. CERN moved on to develop the relativistic heavy ion research program under the leadership of CERN Director Herwig Schopper. Schopper, against great odds, bet on Heavy Ions to becoming one of the pillars of CERN's future. This decision was strongly supported by several national nuclear physics laboratories in Europe, where in my opinion the most important was the support offered by the German GSI laboratory in Darmstadt, and the continued development of relativistic heavy ion physics by one of GSI directors, Rudolf Bock. Many experimental components were contributed by other US and European laboratories. These include the heavy ion source and its pre-accelerator complex, required for heavy ion insertion into the CERN beam lines.

While the RHIC project took 17 years to travel from the first decision to first beam, SPS took 11 years (Pb beam capability). However, SPS was an already built, functional accelerator. Moreover, the RHIC development was hindered by the need to move heavy ion activities from the US West Coast to the East Coast, by the adaptation of ISABELLE design to fit RHIC needs, and the RHIC experiments to fit ISABELLE designed experimental halls, as well as by the typical funding constraints.

And, nobody rushed at BNL. Why rush if the money flows and there is no competition? In the late 80s and early 90s the prevailing opinion at BNL was that RHIC was invulnerable, a dream machine not to be beaten in the race to discover the new phase of matter. However, BNL executive should have known that back than no one could tell what the energy threshold for QGP formation in the very heavy ion collisions would be. The theoretical presumption made at BNL that this threshold was above the energy produced at CERN-SPS turned out to be false.

Because data taking for the RHIC beam did not happen until early 2001, about a year after CERN announced the QGP discovery, the priority in the field of heavy ions that the US pioneered in a decisive way at Berkeley in the early 1970s passed on to CERN where a large experimental program at SPS was developed, and where, as it is clear today, the energy threshold for QGP formation in Pb-Pb collisions was within SPS reach. Active QGP research at LHC and SPS at CERN and at RHIC at BNL continues today.

Science Politics

Having said all the above it is still not obvious how the US flagship RHIC at BNL arrived only at the research frontier in 2001 nearly 7 years after CERN started its experimental work that created the QGP announcement even before RHIC went on-line, and 12 years after S-W and S-S collisions in my opinion also already demonstrated QGP. The approval of the RHIC and QGP research program was made in 1983/4, essentially in parallel with the CERN program decision. The RHIC collider at BNL was to be installed in an already existent tunnel using superconductive magnet technology that was already developed for the prior BNL Isabelle project. Experimental halls were also in place. To compare, at CERN situation was similar but the needs for accelerator construction far less: CERN needed a heavy ion source and first stage heavy ion accelerator.

At the 1984 meeting when the US government DOE management offered 1990 as their forecast of the turn-key RHIC start-up, the inconvenient question asked was, why not in 1987 or 1988? Today, looking at the time-line of CERN's LEP to LHC conversion, which has been an order of magnitude more complex (again, tunnel and experimental halls are reused), one must recognize that the proposed RHIC schedule, completion around 1990, was realistic. And yet it took 16 years to ready RHIC instead of 3-6 years. One should note that John Marburger was President of Stony Brook University from 1980 until 1994, and director of Brookhaven National Laboratory from 1998 until 2001 thus in position to extract for much of this period the largest possible project pay-off for the Long Island science community.

These questions will be studied by science historians as it is today clear that this foot dragging has handed to CERN the QGP discovery opportunity. I can throw into this future analysis my 2 cents: I am familiar with independent research agencies, CERN, NASA, ESA, pursuing fundamental research programs within set budgets, choosing within constraints imposed by strategic political decisions by those sending the money their own priorities. These agencies are famous for effective use of resources and timely delivery of expected outcomes.

The operational models for the numerous US fundamental research laboratories as well as European national research laboratories is different. These labs have very limited independence. Even a large program that is approved is micromanaged by government offices staffed by changing administrators and other career individuals who often have other interests in mind than the speedy realization of a program they supervise; they are never held accountable and all committees I have ever seen evaluate the distributors of government financial largess as the "best" there is. These management weakness combines with natural desire of a laboratory that depends on its approved program for its lifeline to stretch the lifespan of programs.

Research Group Race

We now look through a magnifying glass at the two experimental teams working at CERN that are generally credited with decisive contributions to applying strangeness signature in the QGP discovery.

The first results from CERN RHI experiments appear in 1987/88. The CERN NA35 presents at the Quark Matter Conference in 1988 for 200 GeV A Sulfur beam on S-target collisions, what I would call picture-perfect QGP results, but even in the proceedings of the meeting published a year later in July 1989 all pertinent numbers, figures are overprinted with "preliminary".

These NA35 results are presented in their extended and final format two years after the meeting (submitted 17 April 1990; in revised form 2 July 1990). The title of this paper should perhaps have been "Strangeness Enhancement as the Evidence of QGP," but the 67 authors on this manuscript (a large number as measured by the "nuclear physics" norm of the time period) and the review process prevented such a claim. Thus instead the manuscript has the non-telling title, "Neutral strange particle production in sulfur-sulfur and proton-sulfur collisions at 200 GeV/nucleon" (Zeitschrift für Physik C Particles and Fields volume 48, pages191–200(1990)). The abstract states, "Significant enhancement of the multiplicities of all observed strange particles relative to negative hadrons was observed in central S-S collisions, as compared to pp and pS collisions." In the concluding section, buried in a lot of ink, one finds, "Thus our observation ... appears to be consistent with a dynamical evolution that passes through a deconfinement stage." The argument did not end with this:

This was said to set up the opposite outcome the manuscript continues rejecting this claim citing unpublished forever work recognized by others as a collection of errors and omissions, and another irrelevant publication. NA35 decided in 1990 in their initial flagship publication not to claim directly or indirectly that they were seeing QGP. This spread to other contemporary NA35 writings. At the QM1990 conference, report printed in April 1991 based on a presentation in mid-May 1990, the spokesman of NA35 Reinhard Stock says: "we have demonstrated a two-fold increase in the relative strangeness concentration in central S-S collisions, both as reflected in the K/pi ratio and in the hyperon multiplicities. A final explanation in terms of reaction dynamics has not been given as of yet." This last phrase is hard to understand since several authors of this article including the spokesman R. Stock were my former colleagues from University Frankfurt and they were well versed in strangeness signature of QGP. Thus this phrase implies that at least through 1991 the NA35 collaboration did not want, as a group, to introduce the QGP interpretation of the strangeness enhancement results. They rejected their own result and their own discovery. I now know that R. Stock himself was in the refusnik NA35 camp, responsible for this point of view: I saw that even in year 2020 he does not want to accept his own discoveries, rejecting strangeness signature of QGP.

Another CERN experimental group WA85 took, under the leadership of Emanuele Quercigh, the center stage of strangeness production and QGP search with results on: Lambda(qqs) and anti-Lambda (a CERN preprint of 18 April 1990); on Xi(qss) and anti-Xi (a CERN preprint of 8 November 1990); and a systematic exploration of QGP characteristic behavior for both (a CERN preprint of 5 July 1991). Here the WA85 collaboration takes a firm position in favor of QGP discovery with the words: "The(se) results indicate that our anti-Xi production rate, relative to anti-Lambda is enhanced with respect to pp interactions; this result is difficult to explain in terms of non-QGP models or QGP models with complete hadronization dynamics. We note, however, that sudden hadronization from QGP near equilibrium could reproduce this enhancement” referring to my work published in March 1991:

I open the abstract of my 1991 analysis with the words: "Experimental results on strange anti-baryon production in nuclear S-W collisions at 200 A GeV are described in terms of a simple model of an explosively disintegrating quark-gluon plasma (QGP)." In conclusion I close with, "We have presented here a method and provided a wealth of detailed predictions, which may be employed to study the evidence for the QGP origin of high pT strange baryons and anti-baryons." The WA85 paper I quoted above echoes this point of view, taking the WA85 collaboration in 1991 to the QGP precipice. Today, we can say that with this 1990/91 analysis method and the WA85 results and claims of the period, the QGP was discovered. This discovery had to be confirmed in Pb-Pb collisions, and to be institutionally accepted. This took 8-9 years.

Further reading

  1. The interested reader can find on my strangeness publication page the time line of relevant work.

  2. Reminiscence style review Discovery of Quark-Gluon Plasma: Strangeness Diaries
    Eur. Phys. J. Spec. Top. 229, pp1–140 (EDP Sciences/Springer Open Access 2020)

  3. Wikipedia article on Strangeness and quark-gluon plasma

v1 of September, 2015, v2.1 revised and extended July 2017, v2.2 update May 2, 2020
Return to Johann Rafelski home page