The universe is a complex and intricate place. We can move easily through air and yet not through a wall. The sun transmutes one element to another, bathing our planet in warmth and light. Radio waves have carried a man's voice to Earth from the surface of the moon, whereas gamma rays can inflict fatal damage on our DNA. On the face of it, these disparate phenomena have nothing to do with one another, but physicists have uncovered a handful of principles that fuse into a theory of sublime simplicity to explain all this and much more. This theory is called the Standard Model of particle physics, and it encapsulates the electromagnetic forces that make a wall feel solid, the nuclear forces that govern the sun's power plant, and the diverse family of light waves that both make modern communications possible and threaten our well-being.

The Standard Model is one of the most strikingly successful theories ever devised. In essence, it postulates that two classes of indivisible matter particles exist: quarks and leptons. Quarks of various kinds compose protons and neutrons, and the most familiar lepton is the electron. The right mix of quarks and leptons can make up any atom and, by extension, any of the different types of matter in the universe. These constituents of matter are bound together by four forces—two familiar ones, gravity and electromagnetism, and the less familiar strong and weak nuclear forces. The exchange of one or more particles known as bosons mediates the latter three forces, but all attempts to treat gravity in the microrealm have failed.

The Standard Model leaves other questions unanswered as well, such as: Why do we have four forces and not some other number? And why are there two types of fundamental particles rather than just a single one that handles everything?

These are intriguing problems. Nevertheless, for a long time now a different puzzle has captured my attention and that of many other physicists. The Standard Model views quarks and leptons as indivisible. Astoundingly, though, various clues imply that they are instead built of still smaller components. If quarks and leptons are not fundamental at all, and smaller bits do in fact exist, their presence will force extensive revisions of our theories. Just as nuclear power was inconceivable before Ernest Rutherford discovered the structure of the atom in 1911, unveiling another layer of the subatomic onion will certainly reveal phenomena we cannot yet imagine.

Generation Gaps

Resolving this issue requires scientists to smash particles together at extremely high energies. Since the observation of quarks in the 1970s, we have lacked the tools that might allow us to peer inside them. But now the Large Hadron Collider (LHC) at CERN near Geneva—the same machine that recently confirmed the existence of a Higgs boson, the last undocumented particle in the Standard Model—is gaining speed and could be up to the task.

The first hints of structure in quarks and leptons emerged from research into another—still unsolved—poser, related to the numbers of different kinds of quarks and leptons that have been discovered. Protons and neutrons consist of two types of quarks, called the up quark and the down quark. Up quarks have +2/3 the electrical charge of the proton, and down quarks have −1/3 of the proton's charge. Although only these two types of quarks, plus electrons, suffice to make up the matter of the universe, other quarks have been observed. The strange quark has the same charge as the down quark, but it is heavier. The bottom quark is an even heavier version. Similarly, the charm quark is a heavier cousin of the up quark, with the superheavy top quark rounding out the up quark family. Particle physicists have observed all these quarks, but the four heavier ones decay, in fractions of a second, into the lightest two.

The electron also has heavy, unstable cousins, the muon and the even heavier tau lepton, both of which have the same charge as the electron. And the known menagerie of particles includes three copies of neutrinos, all of which are superlightweight and electrically neutral.

This cornucopia naturally led physicists to ask: Given that the up quark, down quark and electron are the only particles necessary to build a universe, why do they have so many cousins? The question can be encapsulated in Nobel Prize–winning physicist I. I. Rabi's oft-quoted quip when he learned of the discovery of the muon: “Who ordered that?”

One way scientists went about tackling the mystery of populous particle families was to construct a chart delineating the features of all known elementary particles [see box at right], analogous to the periodic table of chemical elements. The periodic table offered physicists the first hints that the chemical elements might not be fundamental, that systematic patterns in the atom's inner structure might account for similar properties of elements in particular rows and columns.

The table of quarks and leptons has three columns called generations (which is why the mystery of particle multiplicity is now referred to as the generation problem). Generation I, at the far left, includes the up and down quark as well as the electron and electron neutrino—everything needed to explain our familiar universe. Generation II contains the somewhat more massive versions of the same particles; generation III has the most massive of all.

The Standard Model treats the quarks and leptons as pointlike particles without any internal structure. But the patterns within the table, as within chemistry's periodic table, raise the possibility that the differences in generations stem from the configuration of even smaller building blocks of matter within quarks and leptons.

Another historical precedent, near the dawn of the 20th century, that may have relevance in the search for the quark's underlying structure is the discovery of radioactive decay. Through a process not understood at the time, one element can transmute into another. We now know that by changing the number of protons and neutrons in the nucleus, it is possible to achieve the goal of medieval alchemists and convert lead into gold. The range of possible transmutations is even wider, as nuclear alchemy can even convert a neutron into a proton (or the reverse) by changing the identity of their constituent quarks. This transformation occurs via the weak nuclear force, which can also transmute leptons, although quarks cannot be changed into leptons, or vice versa. Just as the conversion of one element into another reflects the complex inner workings of the atom, so the metamorphosis of the quarks and leptons may provide yet another hint of even finer details within those particles.

Part and Particle

Many hypothetical building blocks for quarks and leptons have emerged, each with a different name, but the term “preon” has stuck as a generic descriptor for all of them. In most cases, the same name applies to the constituents of the particles that carry the forces acting on these bits of matter.

As an illustration, consider a straightforward model proposed independently in 1979 by Haim Harari, then at the Stanford Linear Accelerator Center, and Michael A. Shupe, then at the University of Illinois at Urbana-Champaign, and subsequently extended in 1981 by Harari and his student Nathan Seiberg, both then at the Weizmann Institute of Science in Rehovot, Israel [see box at left]. They proposed that two kinds of preons exist, one with an electrical charge of +1/3 and one with a charge of zero; in addition, each of these preons has an antimatter companion with opposite charge: −1/3 and zero, respectively. These preons are fermions—particles of matter—and each quark and lepton contains a unique mix of three preons. Two preons of +1/3 charge and one with zero charge, for instance, make an up quark, whereas the up quark's antimatter counterpart contains two preons of −1/3 charge and one with zero. The force-carrying bosons, meanwhile, consist of unique six-preon combinations. The positively charged W boson, for instance, which carries the weak nuclear force that acts on both quarks and leptons, has three +1/3 preons and three zeroes.

Using a series of sensible assumptions, Harari and Shupe postulated the preon content of all the particles of the first generation. The same building blocks can also account for gluons, the subatomic particles that mediate the strong nuclear force to bind quarks together inside protons and neutrons, as well as the other force-carrying bosons.

The trick in formulating any underlying structure of the well-known quarks, leptons and bosons is accounting for the myriad interactions of those particles and forces. Indeed, preons can provide a sensible language for describing subatomic processes. For instance, consider an up quark colliding with an antimatter down quark, yielding a positively charged W boson that decays into an antimatter electron, or positron, and an electron neutrino. In the preon model devised by Harari and Shupe, the incoming quarks, with their three preons apiece, combine in the collision to generate a W boson, now containing all three +1/3 charges and all three zero charges. Then the W boson comes apart, spitting out a different configuration of the same six preons: one positron (with the three +1/3 charges) and one electron neutrino (with the three zero charges).

Thus far I have discussed what might be called quark and lepton numerology. This is just a counting game, like balancing chemical or math equations, albeit a serious and feasible one. To be successful, a preon model must explain the quarks and leptons with a small number of building blocks and a few governing rules. After all, the hope is to find an underlying order that unifies superficially different particles, not a system of ad hoc definitions that accounts for their properties on a case-by-case basis. Such an explanation has been accomplished with preons, both in the Harari-Shupe model and in its successful competitors.

You may have noticed, however, that the discussion so far has included only the first generation of quarks and leptons. Things get murkier when we turn our attention to the second and third generations. Within the model proposed by Harari and Shupe, the higher generations are hypothesized to be excited states of the first-generation configurations. Just as electrons jump from one energy level to another in atoms, some unknown mechanism is thought to bind the preons together in a way that allows for multiple-particle generations from the same ingredients.

If this explanation seems a bit like hand waving, it is. Many of the details have not been worked out. The theoretical studies that first proposed the idea of quarks had a similar level of sophistication. It was only later that the strong force, which binds quarks together into protons and neutrons, was described mathematically. Still, the generation problem remains conspicuously unexplained, so several physicists have proposed competing models, including one in which one of the preons carries the generation number as well as a new charge called hypercolor, which binds the preons inside the quarks and leptons.

Although I have described a single theory of preons, do not be misled into thinking it is the only one out there. My theoretical colleagues are very smart and very creative. Literally hundreds of papers have been written proposing other preon models, although these models are often variations on a small number of basic themes. Some have preons with 1/6 charge, rather than the 1/3 in Harari and Shupe's model. Others have five preons in the quarks and leptons as opposed to three. Still others propose a mix of fermion preons and boson preons or different preon contents for the bosons than those laid out in the bottom table on the opposite page. The possibilities are actually quite rich. We physicists need more data to help weed out the alternatives.

Beyond inherent fascination with the notion that the smallest known pieces of matter might have smaller pieces still, many of us are interested in preons for another reason. If they exist, they could have something profound to say about another outstanding mystery in particle physics. The Standard Model postulates that the Higgs field is the source of mass for fundamental particles. Massive particles feel a sort of drag as they move through this ubiquitous field, whereas massless particles such as the photon glide through unmolested. If the preons that make up the second and third generations are the same as the first, presumably something about the preons enables the higher-generation particles to interact more with the Higgs field than the first generation does, thus giving the higher generations their greater mass. Whereas the Higgs mechanism can account for the masses of the particles, it cannot predict them.

Until a deeper theory is invented, the mass of subatomic particles can be determined only by measuring them one by one. Presumably by understanding the structure of quarks and leptons and how the generations differ, we will learn much more about the Higgs mechanism.

Potholes and Detours

I should note that preon theory is not without its problems. For starters, all experimental efforts to see preons have failed. That failure is disappointing but could stem simply from having inadequate equipment. Experimental questions aside, some concerns are intrinsic to the theory itself. It is a natural feature of “confining theories,” so called in this context because the preons are confined inside the quarks and leptons, that the relevant masses are inversely proportional to the confinement size. Because quarks and leptons are much smaller than protons, this rule implies that a quark made of confined preons would be much more massive than a proton, which is itself made of quarks. The proton whole would be less than the sum of its parts—less, indeed, than the individual parts themselves.

Although this problem may seem insurmountable, physicists have managed to get around a similar kink related to bosons. A quark and antiquark, for instance, can make up a boson called a pi meson, in which the confinement conundrum also seems to pose a problem. Using an idea sketched out in 1961 by Jeffrey Goldstone, then at CERN, however, theorists have long realized that symmetries in the underlying theory could overcome this difficulty. Thus, the lightness of the pi meson was not really a surprise. Unfortunately, this approach applies only to bosons, not to fermions such as quarks. Yet in 1979 Gerard 't Hooft of Utrecht University in the Netherlands worked out a related approach that does work for fermions. Whether 't Hooft's concept is borne out in actual particles remains unclear, but his ideas have at least shown that the theoretical roadblock of quark masses is not as formidable as it first appeared.

Preons are not the only avenue physicists have explored in hopes of solving the generation problem. One prominent alternative is the idea of superstrings, in which the ultimate building blocks of matter are not subatomic particles but tiny vibrating strings. Metaphorically, each of the Standard Model particles can be thought of as strings playing a different note and all of reality as an orchestra of superstrings playing a grand cosmic symphony. Happily, preons and superstrings can amiably coexist because the size scale of superstrings is much smaller than that of quarks and leptons. If superstrings exist, they could well make up not quarks and leptons but rather preons or even pre-preons or pre-pre-preons, depending on how many undiscovered onion layers of matter exist.

Another alternative to the idea of preons as ordinary, albeit undiscovered particles emerged in 2005, when Sundance Bilson-Thompson of the University of Adelaide in South Australia devised a way of describing preons as twisted braids of spacetime. This model is still in its infancy, but physicists are exploring its implications, not least because it offers one possible path to integrating a long-sought quantum theory of gravity into the Standard Model.

Proof in the Preon Pudding

Physics is ultimately an experimental science. No matter how clever the theory, if it fails to agree with measurement, it is wrong. So what can experimentalists do to prove or disprove the existence of preons? The Standard Model successfully describes the quarks, leptons and bosons of the universe without invoking preons, so physicists must look for subtle deviations from the Standard Model's predictions—tiny cracks in the edifice of modern physics. Two facets of the model in particular look like attractive areas to explore more closely.

The first is size. The Standard Model treats the quarks and leptons as pointlike—that is, particles with zero size and no inner structure. Finding a nonzero size for those particles would provide powerful evidence for preons. Measurements have shown that protons and neutrons have a radius of about 10−15 meter. Experiments at the world's leading particle colliders, past and present, have searched for evidence that quarks or leptons also have a measurable size. Thus far all the data are perfectly consistent with zero size or with a nonzero size as small as about 0.0002 to 0.001 times the size of a proton.

To distinguish between those two possibilities (zero size versus very, very tiny), we need to make more precise measurements. The LHC is a discovery machine, and the huge amount of data expected from its current collisions and a scheduled upgrade in the accelerator's energy are two ways in which we can expect to learn more about the size of quarks and leptons.

Another way to demonstrate the existence of particle substructure—for leptons, at least—is to investigate the tightly related concepts of spin and magnetic moment. With some poetic license, an electron can be thought of as a spinning ball, and physicists quantify that property with a spin quantum number. Like all fermions, the electron is said to have spin 1/2. Because the electron is electrically charged, the combination of spin and charge confers a magnetic moment, which is just a fancy way of saying it turns the electron into a familiar magnet, with a north and south pole. Assuming a lepton is a pointlike particle with spin 1/2, it should have a single and specific magnetic moment. So if a measurement of the electron or muon turns up a magnetic moment that differs from the prediction, that result would strongly suggest that those particles are not pointlike and therefore could be composed of preons.

Physicists have long known that the magnetic moments of both the electron and muon do diverge somewhat from that of a pointlike particle. This small difference has nothing to do with preons, though, and can actually be explained within the Standard Model. Each lepton is surrounded by an evanescent cloud of so-called virtual particles, which flicker in and out of existence. Because this cloud has a size, it alters the magnetic moment of the lepton ever so slightly—by about one part in 1,000. The effects of preons would be even smaller, but they could be detectable. New measurements on the horizon, at Fermilab's muon g-2 experiment, will be more than a factor of four more precise than those thus far achieved.

Physicists have also dug through collider data to look for particle decays that would be expected if preons exist and if the higher generations of particles are simply excited states of the first. One such process is a muon decaying into an electron and a photon. This decay has not yet been observed and, if it happens at all, occurs less than about one time in 100 billion.

Every direct measurement made to date is consistent with the hypothesis that quarks and leptons are, indeed, pointlike, with spin 1/2. For those of us who think that the observed generations of subatomic particles are a tantalizing hint of undiscovered physics, the past few decades have been frustrating. But now we have real opportunities to explore new territory.

In 2011 the LHC collided beams of protons at an energy of seven trillion electron volts (7 TeV), 3.5 times the previous world record (held by Fermilab's Tevatron for more than a quarter of a century). In that one year the LHC delivered half as much data from collisions as the Tevatron did over its entire 28-year operating career. In 2012 CERN raised the LHC's energy to 8 TeV before initiating a temporary shutdown at the end of the year to make repairs and improvements. The LHC is scheduled to resume operations in 2014, with a full program of colliding proton beams at 13 or 14 TeV beginning in early 2015.

The modest 2012 increase in energy meant a lot for preon searches. Improvements in accelerator operations increased the brightness and duration of the beams, quadrupling the number of collisions. In addition, the small increase in beam energy quintupled the number of highly energetic collisions—which probe the smallest sizes and which are exactly the kinds of events we need to inspect for evidence of preons. Thanks to these two improvements, the data from 2012 provided 20 times as many collisions of interest to preon hunters as in 2011 and have cut the maximum allowed size of preons in half. Ongoing analysis should enable us to again cut the maximum size in half—and perhaps even to find preons themselves! The next five years will be an exciting time for those of us looking for the ultimate building blocks of the universe.

In addition to the LHC, the Fermilab research program is undergoing a fundamental retooling, which will include a new ability to search for direct evidence of preons. Since the Tevatron was decommissioned in 2011, Fermilab's accelerators no longer tread the energy frontier of particle physics. Instead Fermilab is pushing forward into the intensity frontier, exploring rare phenomena with unprecedented precision. Two of the experiments are most relevant to the preon search. One will measure the magnetic moment of the muon, and the other will look for a muon that decays into a single electron without producing any neutrinos.

The future of hunting for structure within the quarks and leptons is brighter than it has been for a long time. As you read this article, my colleagues and I are combing through the huge amount of LHC data already taken. We are searching for evidence that quarks and leptons have a nonzero size. We are looking for a fourth generation of quarks and leptons and for some evidence that the force-carrying particles also have generations—that the W and Z bosons, which mediate the weak nuclear force, have heavier cousins.

The next few years will mark the start of a new foray into the subatomic realm, a journey the likes of which scientists last encountered more than 25 years ago, when the Tevatron began operations. Like intrepid adventurers of yesteryear, physicists are forging ahead and blazing a trail into the quantum frontier.