Material World: The Six Raw Materials That Shape Modern Civilization

by Ed Conway

The main ingredient in most sands is silica—silicon dioxide or, as it’s sometimes known, quartz. And since glass is, for want of a better phrase, melted sand, silica is also the primary ingredient in glass. But that silica content can vary significantly. The glasses we drink from or have in our window panes typically have about 70 per cent silica. The silica content of obsidian and of most tektites is generally between 65 per cent and 80 per cent. The silica content of Libyan desert glass, on the other hand, is a staggering 98 per cent. Not only did this make it the purest naturally occurring glass to have been discovered anywhere, it was purer than anything humankind could create—for the time being at least.[2]

While most sands are primarily silica, some, especially the beautiful white ones on tropical beaches, are composed mainly of something else: the ground-down remains of seashells and corals. Indeed, if you’re on a pristine beach in the Caribbean or Hawaii, the chances are that your feet are probably sinking into parrotfish excrement: the fish eat the corals, extract the nutrients, and poop the remaining calcium carbonate on to the seabed. For the most part, the whiter and warmer the beach, the more likely it is to have come out of the bottom of a parrotfish. The question of what sand

The challenge in making glass is that sand’s main ingredient, silica (silicon dioxide), melts at extremely high temperatures— over 1,700°C, far higher than anything an open fire or primitive furnace could manage. However, add a so-called “flux” to the mix and you can persuade the silica to melt and flow (flux) at much lower temperatures. Indeed, choose the right flux and not only will it reduce silica’s effective melting temperature, it will also sop up the impurities in the glass, helping improve the final product.

The structure of glass is, frankly, a bit of a mess. Clear and perfect as it might appear to our eyes, down at the molecular level, glass looks more like a random ball-pit of atoms. The technical term for this jumble depends on who you’re asking: for some scientists it is an “amorphous solid,” for others a “supercooled liquid.” In theory it’s both liquid and solid, though, given the way it behaves, in practice it’s really the latter.

Philip Anderson, who won the Nobel Prize in Physics in 1977, wrote a couple of decades later: “The deepest and most interesting unsolved problem in solid state theory is probably the theory of the nature of glass and the glass transition.” It remains unsolved to this day.

Silica sands, which is to say sands with more than 95 per cent silica, have plenty of uses. We need them to help filter our water and to make foundry moulds into which you can pour molten metals. Without silica sands the rail system would grind to a halt, or rather it would fail to grind to a halt, since these sands are used in modern trains’ braking systems. But, most of all, silica sand is the primary ingredient in the manufacture of glass. And if you want the very clearest, finest glass, you need the very purest of all silica sands, sometimes called silver sands.

The striking thing is how rarely the price of such goods ever reflects their importance. Look at the national accounts of any large nation and it’s staggering to consider how much raw materials are reflected in national GDP. Staggering, because it is so minuscule. There is a clear and cogent economic logic to this: statistics like gross domestic product are ultimately measures of how much people will pay for a given item, and 99 times out of 100 raw materials—be they metal or mineral or food—are pretty cheap. But price is not the same thing as value, and occasionally, in extraordinary circumstances such as war—“supply shocks,” economists call them—you end up where the British and Germans ended up in 1915: prepared to trade critical goods with which the better to kill each other. You end up with naval officers from different sides wielding binoculars made in the same factory, as happened in the North Sea in 1916, when the British and German fleets clashed in the Battle of Jutland. As the bombs rained down all around his ship, one British officer remarked on how “interesting” it was that the “Hun perched up in one of those distant masts” was watching the British fleet “through a pair of Zeiss binoculars and I was watching his ship through a similar pair of Zeiss.”[11]

Because light takes longer to pass through water than it does through air or a vacuum—1.33 times longer, to be precise. That number, the refractive index, is one of the most important numbers in science, since understanding how to bend light enables you to start bending it to your will.

This principle of combining glasses with different refractive indices remains one of the mainstays of modern optics. A fibre optic wire can successfully transmit light over vast distances because it combines an inner glass core with an outer glass cladding with a different refractive index. Rather than escaping when it hits the edge of the fibre, the light is bent back inside.

Cast your eyes back through history and you encounter an uncanny pattern: in each era the rule of thumb is that the ascendant economic power also tends to be the global centre of gravity for glassmaking. From the ancient Egyptians to the Romans to the Venetians in the thirteenth century and the Dutch in the sixteenth century, to the British and French in the eighteenth century, all the way to the Germans in 1914.

The Lochaline quartz sand mine was opened in the summer of 1940 and in the following years it would become a crucial cog in the British war machine, producing sand that was shipped south to optical factories, where it was melted and worked into lenses for binoculars, periscopes and gunsights. Long before anyone came up with the concept of critical minerals, Lochaline was a site of military and national importance—even if few people, then as now, knew of its existence.

For a fraction of Lochaline sand gets transported to Norway where it is used to make silicon carbide, a material that is rapidly becoming one of the most important ingredients in the electrification of the world’s vehicles. Silicon carbide inverters are capable of making cars like Tesla’s Model 3 run longer, charge faster and consume less power than their predecessors. This magic material could play an outsize role in saving the planet.

In 1934, while working in upstate New York at the labs at Corning, the glassmaking business, a young American chemist called James Franklin Hyde managed to make glass by synthesising it from chemicals, spraying silicon tetrachloride (a liquid formed by dissolving silica sand in chloride compounds) into the flame of a welder’s torch. The resulting glass was remarkable, not just for the way it was created—the first step-shift in glassmaking in thousands of years—but for its chemical composition. Hyde had manufactured a nearly immaculate form of silica glass, a glass as pure—indeed, purer—than Libyan desert glass. What nature had created with a meteor strike in the Great Sand Sea could now be produced in a laboratory.

strands of glass. An optical fibre is essentially a long wire made of glass, or rather two long glass wires, one inside the other—an inner core to transmit the information and an outer layer of glass to keep that light bouncing, refracting, onwards in the fibre rather than escaping. Creating one of these strands entails forging a thick two-layer tube, which looks like a giant glass canister—a preform, as it’s known—and stretching it under intense heat until it’s a nearly hair-thin diameter.

The problem at the time was that the best optical glasses made using conventional methods could only carry light for about 10 metres, so Kao went in search of an even clearer glass. He found it in the form of that ultra-pure fused glass first developed by James Franklin Hyde at Corning back in the 1930s. Kao calculated that light could travel for kilometres down such a glass with barely any data loss. And since the bandwidth of tiny fibres was so much greater than far thicker copper ones, even an incredibly thin strand could carry multiples more information.

Consider the world’s tallest building, the Burj Khalifa in Dubai, which is quite literally built atop shifting desert sands. In its case, the sands actually make it more stable, not less, for the foundations include 192 piles—long, round concrete pillars—that reach 47 metres down into the ground, and use the friction of these sands and sandstone bedrocks to anchor the building in place.

Given a country’s territorial waters are determined by its coastline, the emplacement of sand has become a new frontline of twenty-first-century diplomacy. The bigger those territorial waters are, the bigger the area a country can fish and drill and mine for offshore resources. Then there are the military consequences. Between 2006 and 2010 China reclaimed an average of 270 square miles a year on its coastline. Its dredging and land reclamation activities across islands in the South China Sea more recently are of such a scale that one U.S. admiral referred to them as representing a “Great Wall of Sand.” The Spratly Islands, a disputed archipelago variously claimed by China, Taiwan, Malaysia and a handful of other Asian nations, were once just reefs inhabited mostly by birds. Today they have been expanded and covered with concrete runways and military bases.[1]

Singapore—whose leaders have characterised land reclamation as a defence against climate change, rising sea levels and water insecurity—is the world’s leading sand importer. The country is growing at a rate of more than 10 square miles a decade and, since more land means more inhabitants, more parks, more room for medical centres, schools and so on, its hunger for sand is only increasing. And as Singapore grows, neighbouring countries are quite literally shrinking. Indonesia, which provided much of that sand to Singapore, recently warned that dredging and sand mining had become so extensive that it had lost a number of its islands entirely, affecting its maritime borders.

We tend to treat sand as a common, essentially infinite material, which makes lots of sense if you assume all sands are alike. But as we know, they are not: why else would a desert nation like Dubai import sand from Belgium, the Netherlands and even, believe it or not, the UK? The short answer is that some sands are more useful than others. There are plenty of grains of silica in the Great Sand Sea, but far fewer with the purity of those you’ll find at Lochaline or Fontainebleau.[3]

by 2020 the total weight of human-made products, from iron to concrete and everything else besides, was greater than the total weight of every natural living thing on the planet.

If you have to mine sand (and it’s hard to imagine the modern world without doing so), these active sedimentary systems are the wrong place to mine. Instead you should be looking for what are known as “fossil deposits”: sands that were once part of an active river or coastal system but now, hundreds of millions of years later, lie inert. And that’s precisely where most sand comes from in developed nations, where sand removal and land reclamation are heavily regulated. But elsewhere there is strong evidence that sand is being mined from active systems.

A few years ago, Mexico began providing families with the cement to pave over dirt floors, with the consequence that parasitic infections dropped by 78 per cent. The number of children with diarrhoea dropped by half; those with anaemia dropped by four-fifths.

The concrete is still curing, the stones still reacting with the environment. Even the Hoover Dam, the enormous Depression-era project on the Colorado River, is still thought to be curing today—gaining ever so slightly in strength with each year that passes.

The cement itself is a powder formed when you roast and crush limestone or chalk along with clay, sand and, occasionally, some other additives such as iron oxide. When you add water, the calcium and silicon in the cement react with it to form a gloopy grey gel, inside of which are millions of microscopic stony tendrils. These tendrils, crystals of calcium silicate hydrate, grow and mesh and spread their fingers throughout the gel, locking in the water and forming a kind of skeletal stone-like structure. Add some gravel and sand to the initial mix and, rather than just bonding to themselves, these tendrils will glue themselves around the stone and sand too, creating concrete. Liquid stone. A rock you can pour.[16]

Since it is ultimately just reconstituted grains of sand and stone, glued into a new stony formation with the help of that lime, it too is made from local grains. Lay your hand on a concrete block in Manchester and you are likely touching gravel and rocks from the Peak District. Concrete in New York often began its life as sand from Jamaica Bay in Long Island. London’s concrete has an even more mystical backstory, since much of the sand and aggregate used to make it is scraped up from a shallow-lying sandbank in the North Sea called Dogger Bank. Back in the last Ice Age when sea levels were lower, this was a land bridge connecting the UK and Europe. This submerged land—Doggerland, as it is sometimes called—has also provided much of the sand used to shore up the Netherlands, and for that matter to create London’s Canary Wharf. Next time you see a concrete block in Britain’s capital and are tempted to dismiss it as a modernist monstrosity, ponder for a moment that it may well have been made of the sands from this mysterious drowned world, out in the middle of the North Sea.

But the recipe for the cement we mostly use today was patented in 1824 by a man called Joseph Aspdin.

it was Edison who perfected the mass production of concrete.

From concrete to copper, from iron to lithium, it is our ability not merely to invent the future but to mass produce and disseminate these materials that has enhanced lives and lifted millions from poverty.

Despite the fact that we only began mass producing this mixture of sand, aggregates and cement just over a century ago, there are now more than 80 tonnes of concrete on this planet for every person alive—around 650 gigatonnes in total. To put that slightly meaningless number into perspective, it is considerably more than the combined weight of every single living thing on the planet: every cow, every tree, every human, plant, animal, bacterium and single-celled organism. Each year we produce enough concrete around the world to cover the entire landmass of England.[20]

In the time it takes you to read this page, more than 120,000 wheelbarrows’ worth of concrete will have been poured in China. In the three years between 2018 and 2020, China poured more concrete than the U.S. had in its entire existence, from 1865, when it opened its first plant producing Portland cement—that variety patented by Joseph Aspdin—via the construction of the Hoover Dam, the U.S. highway system, Manhattan and everything else through to the present day.

The Pantheon may have lasted more than 2,000 years, but the quality of some newly constructed concrete housing in China is so poor that it has an average lifespan of about 20 years.

According to the Federal Highway Administration, nearly one in ten bridges in the U.S. is structurally deficient, with the proportion in Rhode Island and West Virginia over one in five. In the UK the figure may be even higher, with nearly half of all bridges on motorways or A-roads showing evidence of defects.

For the other curse of concrete is that it is one of the biggest emitters of carbon on the planet. For all the attention lavished on other sources of greenhouse gases such as aviation or deforestation, the production of cement generates more CO2 than those two sectors combined. Cement production accounts for a staggering 7–8 per cent of all carbon emissions.

At the time of writing, those global emissions were split roughly 60:40 between the chemical reaction occurring in chalk or limestone as it burns off its carbon in the process of becoming cement, and the energy needed to heat the kiln.

concrete use alone accounts for around a tenth of the world’s industrial water use.

There are many exciting start-ups aiming to mass produce carbon-negative concretes, which sounds mind-bending until you realise that even Portland cement slowly absorbs carbon dioxide from the air as it cures—a kind of reversal of the chemical reaction in the kiln during its creation.

Semiconductors are anomalous materials that for many decades did little more than perplex scientists. They didn’t conduct electricity like copper does, but nor did they insulate its current like, well, glass. No one could quite think what to do with them, but eventually they discovered that they worked brilliantly as a kind of switch. The first such switch, or transistor, as it was named, was made a couple of days before Christmas 1947 by Walter Brattain and John Bardeen, two physicists working under William Shockley at Bell Labs in the U.S. When you see it today (there is a replica in the Smithsonian Museum in Washington, DC) it looks somewhat ungainly, like a soldering experiment gone wrong. It dangles from a Perspex block: a mess of tangled wires plugged into what looks like a wedge-shaped piece of plastic, all sitting atop a dirty-looking slab of dark metal—in this case germanium. But this first solid state switch represented a revolution.

For, by combining enough of these switches, each one a tiny, physical manifestation of the binary code, zero or one, you could create a computer on a tiny piece of silicon, chipped off a circular wafer (hence “chips”). These leaps of innovation, from the switch itself to the “integrated circuit,” the first of which was etched on to silicon by Robert Noyce at Fairchild Semiconductor in 1959, represented the physical foundation of the computing age.

That first contraption from 1947 was about the size of a small child’s hand, but the part that really matters, the transistor itself, was perhaps about a centimetre. Now consider what happened in the following 75 years. By the time of the Intel 4004, the first modern computer chip, in 1971 there were just over 2,000 transistors crammed into roughly the same area, each single one about the size of a red blood cell. Roll forward to the early 2020s and smartphone processors could fit around 12 billion transistors into an area slightly smaller than a square centimetre.

It is tempting to describe these features as microscopic but, at the risk of being pedantic, today’s transistors are even smaller than the wavelength of visible light and are thus totally indiscernible to the naked eye through even the most powerful conventional microscope. They

The company that owns the mine is Ferroglobe, a Spanish business which is the world’s biggest silicon metal producer outside of China.

Indeed, the vast, vast majority of polysilicon goes towards solar panels, and the vast, vast majority of that is made in China. What is striking, however, is that China has yet to master the manufacture of the pièce de résistance of the silicon world: semiconductor grade polysilicon. This can have as many as ten nines (99.99999999 per cent purity), where for every impure atom there are essentially 10 billion pure silicon atoms.

Most of the initial efforts to create transistors were hamstrung not by a shortage of brainpower or a lack of imagination but by the absence of truly reliable materials. That very first transistor made by Brattain and Bardeen at Bell Labs in 1947 was made not of silicon but germanium. However, germanium was ill-suited for use as a transistor. It didn’t function very well at high temperatures, something that is deeply inconvenient given semiconductors can, as you’ll know if you’ve worked your laptop hard while perching it on your knees, get very hot. Silicon, with its high melting point, was a far more attractive material, at least in theory. But this was back before the Siemens process, so, frankly, those first semiconductors were impure and hence a bit rubbish.[3][*1]

There are many weird and wonderful manufacturing techniques in the Material World but the Czochralski process is among the most captivating. The polysilicon is tipped into a quartz crucible (the crucible must be incredibly pure, or else it may introduce impurities back into the silicon) and heated up to just under 1,500°C. A seed crystal, a pencil-sized rod of silicon, is dipped into the melt and is then slowly pulled upwards, rotating slightly. Gradually, a perfect, solid ingot, a boule, begins to form out of the melt.[*2]

the Czochralski process occurs inside a chamber filled only with argon gas. The boule slowly turns and is gradually lifted, until eventually above the crucible there is an extraordinary-looking, long, shiny, dark metallic cylinder hanging by a thread only a few millimetres thick. But only when you use X-ray diffraction to examine this torpedo, this silicon sausage, do you see the most extraordinary thing: the atoms are arranged into a quite perfect crystal.

For those crucibles in which Shin-Etsu melts the hyper-pure silicon before pulling it into that perfect boule and slicing it into wafers are all made—every single one of them—out of a very particular type of quartz, one you can only get in a single place in the world. It is rare, unheard of almost, for a single site to control the global supply of a crucial material. Yet if you want to get high-purity quartz—the kind you need to make those crucibles without which you can’t make silicon wafers—it has to come from Spruce Pine, a small town on the Blue Ridge escarpment in North Carolina. For a long time, this mine—and by extension the entire global supply of high-purity quartz—was operated by a single company, a secretive Belgian business called Sibelco.

It turns out that while it is pretty hard to find sources of quartz as pure as the snow-white rocks of Serrabal, it is nigh on impossible to find quartz as pure as that of Spruce Pine.

There are a few micro producers in India and Siberia, but nothing to rival the consistency and quality of the two mines in Spruce Pine, which raises some unsettling questions. What if something happened to those mines? What if, say, the single road that winds down from them to the rest of the world was destroyed in a landslide? Short answer: it would not be pretty.

No high-purity quartz means no Czochralski crucibles, which means no monocrystalline silicon wafers, which means, well, the end of computer chip manufacture as we know it. We would adapt; find a new process

TSMC. Here is a business whose sole purpose is to manufacture the processors dreamed up by Apple or Tesla, or “fabless” chip companies like Nvidia and Qualcomm (“fab” being short for fabrication plant). Obscure enough that few outside the computing sector have heard of it, it has nonetheless pushed the boundaries of physics, in the process becoming one of the world’s most valuable, and most important, companies.