The periodic table · and where it stops

Beyond the Table

Everyone meets the periodic table on a classroom wall: 118 boxes in tidy rows, one for each element. It is a very good chart. It is also a chart of how these elements behave in a warm room at ordinary air pressure, which is the one place any of us have ever stood.

Take them somewhere else and they stop obeying it. Squeeze sodium, a metal soft enough to cut with a butter knife, and it turns into a see-through solid you could look through like glass. Squeeze helium, which refuses to react with anything, and it will join to sodium and stay there. Neither of those is a guess: both have been done in a laboratory and the pressures are on this page.

What never changes is the count of protons in the middle of the atom. Eighty protons is mercury wherever you take it. Nearly everything else you could say about mercury, that it is a liquid, that it is silver, that it is a metal at all, is a fact about the conditions as much as about the element.

80protons in the middle: this never changes
Hgso it is mercury, anywhere in the universe
−38.8 °Cbut it only melts here at this temperature, and only under the conditions in this room
80 to 85%of the 2.11 volts a lead-acid cell delivers comes from relativity, by the calculation Ahuja and colleagues published in 2011
200 GPathe squeeze at which sodium, a metal soft enough to cut with a knife, turns into a see-through insulator
113 GPawhere helium, the element that bonds to nothing at your desk, makes a stable compound with sodium
52 °Cthe melting point calculated for oganesson, 325 kelvin, which sits under the noble gases and is expected to be a solid

The table already bends, in the room you are in

Relativity has a reputation for living somewhere else: near a black hole, aboard a spacecraft, inside a particle accelerator. It is also in your car, your wedding ring and the thermometer in the shed. Heavy atoms hold a lot of positive charge in the middle, and the electrons closest in have to move fast to stay in orbit around it. On the simplest estimate, an innermost electron's speed is the atomic number divided by 137.036, so in gold, with 79 protons, that comes to about 58 per cent of the speed of light. At that speed the electron behaves as though it carries about 22 per cent more mass than at rest, its orbit pulls in tight, and the outer shells that do the chemistry are dragged in with it.

Three of the results are sitting in your house.

Mercury pours, and its neighbours do not

Cadmium sits directly above mercury and melts at 321 degrees Celsius (594 kelvin). Gold sits beside it and melts at 1,064 (1,337 kelvin). Mercury melts at minus 38.8 degrees Celsius (234.32 kelvin), which is 38.8 degrees below freezing, and that is why it filled thermometers. Florent Calvo, Elke Pahl, Michael Wormit and Peter Schwerdtfeger ran the calculation both ways and published it in Angewandte Chemie in 2013. With relativity in the equations the melting point comes out at minus 23 degrees Celsius (250 kelvin), close to the measured minus 38.8. With the relativistic terms switched off it comes out at 82 degrees Celsius (355 kelvin): mercury would be a solid you could hold in your hand, and hot tea would not melt it.

The mechanism is the pulled-in outer shell. Mercury's outermost pair of electrons is held so tightly by the contracted shell that it barely reaches out to the next atom, so mercury atoms grip each other about as weakly as argon atoms do. A metal whose atoms hardly hold hands is a metal that runs.

Gold is yellow because its electrons are heavy

Silver and gold sit one above the other and are built the same way. Silver reflects the whole visible band evenly, which is what a mirror finish is. Gold absorbs the blue end, near 2.4 electronvolts, so what bounces back to your eye is short of blue and reads as warm yellow. The absorption is electrons jumping from a full inner shell up to the outer one. Niels Egede Christensen and Bernard Seraphin ran that band structure both with and without relativity at the Technical University of Denmark, published in Physical Review B in 1971, and found the relativistic shifts running to about an electronvolt, the same size as the gaps themselves. Take the shift away and the first absorption moves up into the ultraviolet, out of sight, and the metal reflects the visible band evenly. Gold would look like silver.

Four fifths of your car battery

A lead-acid cell delivers 2.11 volts, and six of them in a row give the 12.7 volts that turns your starter motor. Rajeev Ahuja, Andreas Blomqvist, Peter Larsson, Pekka Pyykkö and Patryk Zaleski-Ejgierd calculated the energies of the solid materials in that cell from first principles at three levels, non-relativistic, scalar relativistic and fully relativistic, and published the result in Physical Review Letters in 2011. Their average calculated voltage was 2.13 against the measured 2.11, and every version of the calculation agreed that 1.7 to 1.8 volts of it come from relativity, mostly from the lead dioxide plate. Take those volts away and each cell drops to somewhere between 0.33 and 0.43 volts, and the six-cell battery under your bonnet would hold about 2 to 2.6 volts instead of 12.7.

Lead is element 82. It is in the sinker on a fishing line and in the flashing on a roof. The relativity is not exotic; it is load bearing.

How to read a rule. The group logic of the periodic table is not handed down from anywhere. It follows from a premise: that only the outermost electrons take part in bonding, so elements with the same outer arrangement behave alike. Conservation laws work the same way. Emmy Noether proved at Göttingen in 1918 that each one follows from a symmetry: energy conservation from the fact that the rules do not change with the time of day, momentum from the fact that they do not change with the place. That makes a conservation law a theorem with premises, not a decree, and a premise is something you can go and check. Squeeze an atom until its core electrons start taking part in bonding and the premise behind the group logic fails. Everything below is what happens when it does.


Switch relativity off and on

Six properties, each with a published pair of numbers: what the calculation returns with the relativistic terms in, and what the same calculation returns with them out. Nothing about the atoms changes. Only the physics used to describe them.

Mercury melts atelement 80
−23 °C (250 K), measured at −38.8 °C (234.32 K)

A liquid metal at room temperature. The outermost pair of electrons is held in tight, so mercury atoms grip each other about as weakly as argon atoms do.

Gold lookselement 79
absorbs near 2.4 eV, in the blue

Yellow. The light that comes back to your eye is short of blue, and short-of-blue reads as warm gold.

Car battery holdselement 82, six cells
12.7 V

Six cells at 2.11 volts each, the measured figure. The calculation returns 2.13 volts a cell, and 1.7 to 1.8 of those volts are relativistic.

Copernicium iselement 112, under mercury
band gap 6.4 eV, melts at 10 ± 11 °C (283 K)

A volatile liquid held together by the weak attraction between closed shells, with a density close to mercury's and an electronic structure like a noble gas. Kenneth Pitzer suggested exactly this in 1975.

Oganesson melts atelement 118, a noble gas by column
52 °C (325 K)

A solid on the bench at room temperature, with a liquid range of 125 degrees above that. Scalar relativity supplies 48 degrees of the shift and spin-orbit coupling another 56.

Francium grips its electronelement 87, below caesium
392.8 kJ/mol, harder than caesium's 375.7

Going down group 1 the grip loosens all the way to caesium, and then tightens again at francium. The outer shell has been pulled in, so the last electron sits closer to the charge and is harder to take.

One more, without a switch to flip. Gold holds a spare electron nearly as well as iodine does: 222.8 kilojoules per mole against iodine's 295.2. Mix caesium and gold and you do not get an alloy, you get caesium auride, a salt with a gold anion and a 2.6 electronvolt band gap, which is to say a semiconductor made of two metals.


Squeeze it and the chemistry changes

Pressure is measured in pascals, and the useful unit here is the gigapascal. One gigapascal is about 9,900 atmospheres, which is roughly what you get by standing eight adult elephants on a single postage stamp. Two diamonds with tips a few tenths of a millimetre across, pressed together with a screw, reach hundreds of gigapascals: that instrument is the diamond anvil cell, and it fits on a laboratory bench.

Once the squeeze is hard enough that the inner electron shells of neighbouring atoms start to overlap, those inner electrons stop being spectators and join the bonding. The premise behind the group logic fails, and elements start behaving like elements from other parts of the table. Here is what that looks like, with the depth inside our own planet where each pressure is reached.

21 GPaabout 600 km down

Xenon starts restacking

Solid xenon changes its packing from cubic to hexagonal. John Caldwell and Jeffrey Nguyen put the equilibrium boundary at 21 plus or minus 3 gigapascals in Science in 1997, and noted the change is sluggish enough to be caught anywhere between 3 and 70 gigapascals depending on how you get there.

78 GPaabout 1,820 km down

Xenon takes on oxygen

Agnès Dewaele, Nicholas Worth, Chris Pickard and colleagues made two xenon oxides above 78 gigapascals and reported them in Nature Chemistry in 2016, with the xenon carrying mixed charges and sitting in extended networks of shared oxygen. A noble gas building a framework solid.

113 GPaabout 2,490 km down

Helium bonds to sodium

Helium holds its electrons harder than any other element, 2,372 kilojoules per mole, and at your desk it bonds to nothing. Xiao Dong, Artem Oganov, Alexander Goncharov and colleagues loaded sodium into helium in a laser-heated diamond anvil cell, took it to 155 gigapascals, and above 113 saw new single-crystal reflections appear after heating past about 1,230 degrees Celsius (1,500 kelvin). The compound is Na2He, built like fluorite, and it is an insulator: the electron pairs sit in the empty spaces between the atoms rather than on any atom. Nature Chemistry, 2017.

118 GPaabout 2,580 km down

Sodium melts at room temperature

Sodium's melting point climbs with pressure to a peak near 730 degrees Celsius (1,000 kelvin) at 31 gigapascals, then falls off a cliff. Eugene Gregoryanz and colleagues followed it with synchrotron diffraction and reported in Physical Review Letters in 2005 that it bottoms out at 27 degrees Celsius (300 kelvin) at 118 gigapascals. Squeeze sodium hard enough and it turns into a liquid at room temperature. The measured line stops at 130 gigapascals.

132 GPaabout 2,820 km down

Xenon becomes a metal

Kathleen Goettel and colleagues reported optical evidence for the metallisation of xenon at 132 plus or minus 5 gigapascals in Physical Review Letters in 1989; Rachel Reichlin and colleagues put the onset nearer 150 in the same issue. The element with a famously full outer shell conducts electricity.

200 GPaabout 3,510 km down

Sodium turns transparent

Yanming Ma, Mikhail Eremets, Artem Oganov and colleagues compressed sodium fivefold and photographed it: at 200 gigapascals it is optically transparent, with a wide band gap and no metallic sheen at all. Six-coordinated, distorted, close packed. They attribute it not to atoms pairing up but to valence electrons being pushed by the atomic cores into the gaps in the lattice. Nature, 2009.

216 GPaabout 3,680 km down

Aluminium restacks, then restacks again

The metal in your window frames changes packing from cubic to hexagonal, then to body-centred cubic at 321 plus or minus 12 gigapascals. Danae Polsin and colleagues watched both changes happen in nanoseconds under laser-driven ramp compression with x-ray diffraction, and followed the last phase out to 475 gigapascals. Physical Review Letters, 2017.

425 GPapast Earth's centre

Hydrogen's gap closes

Eugene Wigner and Hillard Bell Huntington predicted in 1935 that hydrogen would turn into a metal, and put the pressure at 25 gigapascals using the compressibility measured at zero pressure for every pressure, which is the part that was wrong rather than the idea. Paul Loubeyre, Florent Occelli and Paul Dumas took hydrogen to 425 gigapascals at minus 193 degrees Celsius (80 kelvin) in a toroidal diamond anvil cell and watched the direct band gap drop from 0.6 electronvolts to under 0.1, reported in Nature in 2020.

495 GPapast Earth's centre

Hydrogen, reflecting

Ranga Dias and Isaac Silvera reported in Science in 2017 that hydrogen at 495 gigapascals and 5.5 kelvin, five and a half degrees above absolute zero, reflected light in the way a free-electron metal does. Comments published in the same journal dispute what the reflectance data support, and the authors issued a correction to the fit. The reading, the objections and the correction are all on the record together, which is the ordinary condition of a frontier measurement.

The pattern under all of that: the periodic table is a map of behaviour at one atmosphere, and there are other atmospheres. A column tells you what an element does in the conditions the column was drawn in. Change the conditions and elements swap jobs. A metal insulates. A noble gas bonds. A soft solid pours at room temperature.


The conditions bench

Pick an element and drive the two dials. The panel reports what that element is understood to be at those conditions, who found that out and on what instrument, or, where nobody has been there yet, what the calculation returns and what would have to be built to check it.

Element
1 atmosphere
1 atmosphereeach step is ten times the last500 GPa
20.3 °C (293.4 K)
0.001 K above absolute zeroeach step is ten times the last1,000,000 °C
1 atmthe air around you
the surfacedepth inside the Earth reaching this pressure
nothingsix-tonne elephants standing on one postage stamp
20 °Cabout room temperature
Mercury · 80 protons

A liquid metal

Loading.

Temperatures read in degrees Celsius, with the kelvin figure alongside where the physics uses it. Very close to absolute zero the kelvin figure leads, because Celsius has stopped telling you much down there; past about ten thousand degrees the two scales read the same to the eye, so only one figure is given. Both dials are squashed so that many steps of ten fit across them, which is the only way a warm room and the inside of a star share one track. The elephant is a six-tonne African bush elephant and the stamp is 24 by 20 millimetres, worked out from its weight over that area. Depths come from integrating the density profile of the Preliminary Reference Earth Model, which puts the core and mantle boundary at 2,891 kilometres and 135.8 gigapascals and the centre at 6,371 kilometres and 364 gigapascals.


Where the table must end, and why that is interesting

Elements are counted in protons, so the natural question is how high the count can go. There is a real answer in the mathematics, and it is not a wall so much as a door.

Paul Dirac's equation for an electron near a nucleus, with the nucleus treated as a single point of charge, stops returning sensible answers once the charge passes about 137. That number is not a coincidence and it is where the fine structure constant got its nickname: its reciprocal is 137.035999177, one of the most precisely known numbers in all of measurement, and an innermost electron's speed on the simplest estimate is the atomic number divided by it. At 137 that estimate reaches the speed of light itself, and the equation has nothing left to say.

Real nuclei are not points. They are balls a few femtometres across, and giving the charge a realistic size changes the answer: the trouble moves out to roughly 172 to 173 protons. What the calculation says happens there is the interesting part. The energy of the innermost bound level falls all the way down to meet the sea of negative-energy states below, and empty space around the nucleus stops being stable: it starts producing electron and positron pairs, the electron dropping into the level and the positron flying off free.

The atom's own definition changes at that point. Below it, an atom is a nucleus plus however many electrons you give it. Above it, the vacuum itself supplies electrons whether you offer them or not, and a bare nucleus is not a thing that can be left bare. The question is no longer what element it is; it is whether "element" is still the right word for the object.

What would be measured. Nobody has to make an atom with 173 protons to test this: two heavy nuclei passing close enough together share a charge for a fleeting moment, and 92 plus 92 is 184. Heavy-ion collisions between uranium nuclei put the combined charge over the threshold for around 10-21 seconds, and the signature to look for is a sharp line in the positrons coming out, at an energy fixed by the depth the level reached rather than by the collision. That experiment has been run before, at GSI in Darmstadt in the 1980s, and it produced lines that later work did not reproduce. It is being set up again with better detectors and better control of the collision geometry, and the quantity it returns is a positron energy spectrum: a number, from an instrument, that either shows the line or does not.

Between the heaviest element made so far and that threshold there are 55 more proton counts. Teams are firing beams at 119 and 120 right now. In 2024 a group led by Jacklyn Gates at the Lawrence Berkeley National Laboratory made two atoms of livermorium over 22 days by firing titanium-50 at plutonium-244, which was the first time an element in that range had been built with a beam other than calcium-48, and it is the step the search for element 120 was waiting on.


The heavy end already disagrees with the table

You do not have to go to 173 to find the pattern breaking down. It is breaking down at the bottom of the table we already have. In every one of these cases the column makes a prediction, and the calculation or the experiment returns something else.

🟣

Oganesson, the noble gas that is not a gas

Element 118 sits under radon at the foot of group 18. Odile Smits, Jan-Michael Mewes, Paul Jerabek and Peter Schwerdtfeger ran two independent methods against a high-accuracy reference and published in Angewandte Chemie in 2020: melting point 52 degrees Celsius plus or minus 15 (325 kelvin), boiling point 177 plus or minus 10 (450 kelvin), so a solid at ambient conditions with an unusually wide liquid range of 125 degrees.

A year earlier the same group with Georg Kresse had put its band gap at 1.5 electronvolts against radon's 7.1, which makes solid oganesson a semiconductor rather than the transparent insulator every other noble gas freezes into. And in 2018 Paul Jerabek, Bastian Schuetrumpf, Schwerdtfeger and Witold Nazarewicz showed in Physical Review Letters that its electrons lose their shell structure altogether and smear into something closer to a uniform gas.

Five atoms have ever been made. Oganesson-294 lasts about 0.69 milliseconds.
🟢

Flerovium, argued both ways from the data

Element 114 sits under lead, so the column says metal. Two gas chromatography experiments sent single atoms down a gold surface and measured how strongly they stuck. The first, reported by Robert Eichler and colleagues in 2010 from the decay chains of three atoms, gave a stickiness that read as noble-gas-like. The second pointed at a volatile metal instead.

A 2022 study led by Alexander Yakushev, accounting for the roughness of the real surface, reported flerovium sticking to gold less strongly than mercury does and more strongly than radon does, and read the lower limit as a metal-to-metal bond: the least reactive member of group 14, and still a metal. Three atoms at a time is the sample size the whole argument rests on.

Flerovium-289 lasts about 2 seconds.
🔵

Copernicium, mercury's neighbour that is not one

Element 112 sits under mercury. Kenneth Pitzer suggested in 1975 that relativity would make it behave like a noble gas instead, and the 2019 calculation by Mewes, Smits, Kresse and Schwerdtfeger came out on his side: a band gap of 6.4 electronvolts, held together by the same weak attraction that condenses argon, melting at 10 degrees Celsius plus or minus 11 (283 kelvin) and boiling at 67 plus or minus 10 (340 kelvin), with a density close to mercury's.

The 2008 gold-surface experiment by Eichler and colleagues measured a real interaction with the surface and read it as metallic. The 2019 paper reads the same energy, minus 0.37 electronvolts, as the sort of attraction closed shells have anyway. Same measurement, two readings, and the way to settle it is more atoms.

Copernicium-285 lasts about 30 seconds.

Put those together and the rule for the bottom of the table is plain enough: the column a superheavy element sits in is a suggestion, not a promise. It is drawn from how many electrons are in the outer shell, and by element 112 the relativistic pull on those electrons is doing more to set behaviour than the shell count is. The table still puts these elements in the right box for the right reason. It just stops predicting what happens inside the box.


The same element, wearing different matter

Pressure and temperature are two dials. There are others, and each one produces something with the same proton count and almost nothing else in common with the entry in the table.

An atom the size of a sand grain

Push a single electron into a very distant orbit and the atom swells, because its reach grows as the square of the orbit number. Stephen Hogan, Yhoshua Houston and Baochen Wei at University College London photoexcited helium past orbit number 400 with two colours of continuous-wave laser light and published it in 2018: that atom is about 17 micrometres across, a quarter of a hair's width, big enough to see under an ordinary microscope.

Nature does better. Sergei Stepkin, Alexander Konovalenko and colleagues picked up radio absorption near 26 megahertz from carbon atoms in a cold cloud in the Perseus arm, in front of the Cassiopeia A supernova remnant, sitting at orbit number about 1,009. Those atoms are 0.11 millimetres across, a grain of very fine sand, one and a half hair widths. If an ordinary carbon atom were a 1 millimetre bead, that one would be a bubble a kilometre across, and it is still carbon: six protons, doing carbon's job in every other respect.

A cloud of atoms behaving as one

Cool a dilute vapour far enough and the atoms stop being separate. Eric Cornell, Carl Wieman and their colleagues Michael Anderson, Jason Ensher and Michael Matthews did it first at Boulder in June 1995 with about 2,000 rubidium-87 atoms at 170 nanokelvin, which is 170 billionths of a degree above absolute zero, held for more than 15 seconds, and reported it in Science. Wolfgang Ketterle's group at MIT reached the same state in sodium later that year with far more atoms. The three shared the 2001 Nobel Prize in Physics.

What sits in the trap has 37 protons per atom and is unmistakably rubidium. It also has one shared wave for the whole cloud, and it will interfere with another cloud the way two torch beams will not. Same element. A different kind of matter.

A nucleus holding charge

Hafnium-178 has a state 2.446 million electronvolts above its ground state that cannot easily let go, because shedding the energy would mean shedding sixteen units of spin at once. It sits there for 31 years. One gram holds about 1.33 billion joules, which is roughly 2,300 kettles brought to the boil, or the energy in about 317 kilograms of TNT, packed into a gram.

Carl Collins and colleagues at the University of Texas at Dallas reported in 1999 that a beam of x-rays had prompted some of that energy out early. Other groups have looked for the same effect and not seen it, and the question of whether a stored nuclear state can be tapped on demand is open. It moved recently in a different element: in 2024 Johannes Tiedau's group at Germany's national metrology institute, working with the Vienna University of Technology, drove a thorium-229 nucleus with a tabletop laser at 148 nanometres, the first time a nucleus has been switched by ordinary laser light.

A cluster acting like an element that is not on the shelf

Thirteen aluminium atoms in a compact cluster carry 39 outer electrons, one short of a count that closes a shell. That makes the cluster hungry for one electron in the same way a halogen atom is. Its electron affinity is 3.57 electronvolts against chlorine's 3.62, and Denis Bergeron, A. Welford Castleman, Shiv Khanna and colleagues showed it forming the aluminium equivalent of a polyhalide and reported it in Science in 2005.

Aluminium is in group 13. This is not aluminium behaving as a metal, and it is not a compound of a halogen: it is a lump of one element doing another element's chemistry because of how many electrons it happens to hold. The count that decides the behaviour here is a count of electrons in a cluster, not protons in a nucleus.

And then there is the end of element identity altogether. Press matter past the point where nuclei touch, around 2.7 × 1017 kilograms per cubic metre, and the nuclei stop being separate objects: the material becomes one continuous liquid of neutrons with a few per cent protons and electrons. A teaspoon of it weighs about 1.3 billion tonnes, roughly two and a half Sydney Harbours of water in five millilitres. Squeeze further and the calculations say the neutrons themselves come apart into quarks. There is no proton count to read there and no element to name, which sets a plain boundary on where the table's question even applies.


What this opens

Not a conclusion. Edges worth pushing, each with the instrument that would push it and the quantity it would return.

💎

Take the anvils higher, and hold them

The toroidal diamond anvil cell that reached 425 gigapascals and the two-stage cells behind the 495 gigapascal reading both work on samples a few tens of micrometres across, and the diamonds themselves are the limit. Larger samples at those pressures would let x-ray diffraction identify the structure directly rather than inferring it from how the sample reflects light, which is precisely what the arguments over hydrogen turn on.

Returns: a diffraction pattern, so a crystal structure
📏

Nail down the pressure scale itself

Every diamond cell reading is calibrated against a standard material. Dayne Fratanduono, Marius Millot and colleagues compressed gold and platinum without a shock at the National Ignition Facility and the Z machine and fixed their pressure and density relationship out to one terapascal, nearly three times the pressure at Earth's centre, reported in Science in 2021. Every future number in this field is quoted against work like that.

Returns: pressure against density, absolutely
⚛️

Make 119 and 120

The titanium-50 beam that produced livermorium in 2024 is the route. Element 120 is expected to be 10 to 20 times harder to make than element 116 was, which is a several-year search rather than a closed door. Beams are running at RIKEN in Japan, at Dubna, at Berkeley and at Lanzhou.

Returns: a decay chain, so an atomic number
🕸️

Hold single atoms still and look at them

Optical lattices, which are grids of light that trap atoms one to a well, already hold Bose-Einstein condensates and single-atom arrays. Applied to the short-lived heavy elements they would let a handful of atoms be interrogated repeatedly rather than once, which is the fix for a chemistry argued from three atoms.

Returns: spectra from single atoms, repeatedly
🧲

Turn the magnetic dial

The third dial nobody has driven far. Continuous fields have reached 45.5 tesla and a pulsed field at the University of Tokyo hit 1,200 tesla for 40 microseconds in 2018, blowing the doors off the containment chamber. Neutron star surfaces run many orders higher, and at those strengths calculations have atoms squeezed into needles and chemistry rewritten. The gap between what has been reached and what would matter is the size of the opportunity.

Returns: spectra and structure under a field
🔭

Read the sky as a pressure cell

The interior of Jupiter, the crust of a neutron star and the cores of white dwarfs sit at pressures no bench reaches, and they are observable. Seismology on the Sun and Jupiter, cooling curves of white dwarfs and the sizes of neutron stars all return numbers that a laboratory equation of state has to match.

Returns: mass, radius and cooling rate

The thing worth keeping from all of this is not that the periodic table is wrong. It is right, and it is one of the best pieces of pattern recognition anyone has ever done. What the measurements say is that it is a map of one layer, drawn from a premise about which electrons matter, and that the premise has a range. Take an element outside that range and it does something else, and the something else is not chaos: it follows patterns of its own, which people are drawing right now, one anvil cell and one atom at a time.



Sources

  • F. Calvo, E. Pahl, M. Wormit and P. Schwerdtfeger, "Evidence for low-temperature melting of mercury owing to relativity", Angewandte Chemie International Edition 52, 7583, 2013 · onlinelibrary.wiley.com/doi/abs/10.1002/anie.201302742
  • N. E. Christensen and B. O. Seraphin, "Relativistic band calculation and the optical properties of gold", Physical Review B 4, 3321, 1971 · journals.aps.org/prb/abstract/10.1103/PhysRevB.4.3321
  • R. Ahuja, A. Blomqvist, P. Larsson, P. Pyykkö and P. Zaleski-Ejgierd, "Relativity and the lead-acid battery", Physical Review Letters 106, 018301, 2011 · arxiv.org/abs/1008.4872
  • Y. Ma, M. Eremets, A. R. Oganov and colleagues, "Transparent dense sodium", Nature 458, 182, 2009 · arxiv.org/abs/0911.3190
  • E. Gregoryanz and colleagues, "Melting of dense sodium", Physical Review Letters 94, 185502, 2005 · link.aps.org/doi/10.1103/PhysRevLett.94.185502
  • X. Dong, A. R. Oganov, A. F. Goncharov and colleagues, "A stable compound of helium and sodium at high pressure", Nature Chemistry 9, 440, 2017 · arxiv.org/abs/1309.3827
  • K. A. Goettel and colleagues, "Optical evidence for the metallization of xenon at 132(5) GPa", Physical Review Letters 62, 665, 1989; R. Reichlin and colleagues, same issue, 669.
  • J. S. Caldwell and J. H. Nguyen, "Structure, bonding, and geochemistry of xenon at high pressures", Science 277, 930, 1997.
  • A. Dewaele, N. Worth, C. J. Pickard and colleagues, "Synthesis and stability of xenon oxides Xe2O5 and Xe3O2 under pressure", Nature Chemistry 8, 784, 2016 · nature.com/articles/nchem.2528
  • E. Wigner and H. B. Huntington, "On the possibility of a metallic modification of hydrogen", Journal of Chemical Physics 3, 764, 1935.
  • P. Loubeyre, F. Occelli and P. Dumas, "Synchrotron infrared spectroscopic evidence of the probable transition to metal hydrogen", Nature 577, 631, 2020 · arxiv.org/abs/1906.05634
  • R. P. Dias and I. F. Silvera, "Observation of the Wigner-Huntington transition to metallic hydrogen", Science 355, 715, 2017, with the published comments and the erratum at science.org/doi/10.1126/science.aao5843
  • P. Dalladay-Simpson, R. T. Howie and E. Gregoryanz, "Evidence for a new phase of dense hydrogen above 325 gigapascals", Nature 529, 63, 2016.
  • D. N. Polsin and colleagues, "Measurement of body-centered-cubic aluminum at 475 GPa", Physical Review Letters 119, 175702, 2017 · link.aps.org/doi/10.1103/PhysRevLett.119.175702
  • D. E. Fratanduono, M. Millot and colleagues, "Establishing gold and platinum standards to 1 terapascal using shockless compression", Science 372, 1063, 2021.
  • O. R. Smits, J.-M. Mewes, P. Jerabek and P. Schwerdtfeger, "Oganesson: a noble gas element that is neither noble nor a gas", Angewandte Chemie International Edition 59, 23636, 2020 · pmc.ncbi.nlm.nih.gov/articles/PMC7814676
  • J.-M. Mewes, O. R. Smits, G. Kresse and P. Schwerdtfeger, "Oganesson is a semiconductor: on the relativistic band-gap narrowing in the heaviest noble-gas solids", Angewandte Chemie International Edition 58, 17964, 2019.
  • J.-M. Mewes, O. R. Smits, G. Kresse and P. Schwerdtfeger, "Copernicium: a relativistic noble liquid", Angewandte Chemie International Edition 58, 17964, 2019 · pmc.ncbi.nlm.nih.gov/articles/PMC6916354
  • P. Jerabek, B. Schuetrumpf, P. Schwerdtfeger and W. Nazarewicz, "Electron and nucleon localization functions of oganesson: approaching the Thomas-Fermi limit", Physical Review Letters 120, 053001, 2018.
  • A. Yakushev and colleagues, "On the adsorption and reactivity of element 114, flerovium", Frontiers in Chemistry 10, 976635, 2022 · frontiersin.org/journals/chemistry/articles/10.3389/fchem.2022.976635/full
  • R. Eichler and colleagues, "Chemical characterization of element 112", Nature 447, 72, 2007, and the 2008 follow-up on its thermochemical properties.
  • J. M. Gates, R. Orford and colleagues, "Toward the discovery of new elements: production of livermorium (Z=116) with 50Ti", Physical Review Letters 133, 172502, 2024 · arxiv.org/abs/2407.16079
  • S. D. Hogan, Y. Houston and B. Wei, "Laser photoexcitation of Rydberg states in helium with n > 400", Journal of Physics B 51, 145002, 2018 · discovery.ucl.ac.uk/10052702
  • S. V. Stepkin, A. A. Konovalenko and colleagues, "Radio recombination lines from the largest bound atoms in space", Monthly Notices of the Royal Astronomical Society, 2007 · arxiv.org/pdf/astro-ph/0610407
  • M. H. Anderson, J. R. Ensher, M. R. Matthews, C. E. Wieman and E. A. Cornell, "Observation of Bose-Einstein condensation in a dilute atomic vapor", Science 269, 198, 1995.
  • C. B. Collins and colleagues, "Depopulation of the isomeric state 178m2Hf", Physical Review Letters 82, 695, 1999.
  • J. Tiedau and colleagues on the laser excitation of the thorium-229 nuclear transition at 148 nanometres, 2024 · physics.aps.org/articles/v17/71
  • D. E. Bergeron, P. J. Roach, A. W. Castleman Jr, N. O. Jones and S. N. Khanna, "Al cluster superatoms as halogens in polyhalides and as alkaline earths in iodide salts", Science 307, 231, 2005 · science.org/doi/10.1126/science.1105820
  • S. Takeyama and colleagues, the 1,200 tesla pulsed field at the University of Tokyo, 2018 · u-tokyo.ac.jp/focus/en/press/z0508_00008.html
  • A. M. Dziewonski and D. L. Anderson, "Preliminary reference Earth model", Physics of the Earth and Planetary Interiors 25, 297, 1981. Depths on this bench come from integrating its density profile, which returns 135.8 GPa at the core and mantle boundary and 364 GPa at the centre.
  • Fine structure constant reciprocal 137.035999177(21), CODATA 2022. Element figures throughout come from the atlas dataset.