Silicate-derived calcium as a pathway to low-carbon Portland cement

To construct the built environment, modern society uses limestone (CaCO₃) to produce the polycalcium-silicate binder known as Portland cement (PC). PC was invented in the 19th century and has remained the preferred cement for >99% of concrete construction globally because of its ease-of-use, cross-material compatibilities, centuries of data for companies to build strategies and policies around, acceptance in project specifications, trillions of dollars of existing PC-based concrete production and utilization infrastructure, and an existing skilled labor force^1,2. PC production currently accounts for about 4.4% of total global greenhouse gas (GHG) emissions, on the same order as all light-duty automobiles (2.4 Gt CO₂ versus 3.0 Gt CO₂ per year, respectively)³. 1.5 Gt of that CO₂ are direct emissions from limestone calcination (the heating of CaCO₃ to produce CaO and CO₂), with the remainder primarily attributable to burning fossil fuels needed to achieve the high temperatures ( > 1500 °C) that drive the calcination reactions^3,4.

The amount of carbon emitted in the production of each tonne of cement has dropped by 23% compared to the 1990 baseline⁵, but reaching net zero emissions will require a new approach. So far, emissions reductions have occurred by increasing the energy efficiency of cement plants, using more alternative (non-fossil) fuels, and reducing the amount of CaO that is used in cement by blending in supplementary cementitious materials (SCMs)⁵. However, the energy efficiency of cement production is close to its thermodynamic minimum⁶, limiting the potential carbon abatement from energy efficiency. And even if cement production eliminated energy emissions by using nuclear or renewable energy, traditional cement production would still release 1.5 GtCO₂ y⁻¹ from limestone. This is because all hydraulic cements (including PC and common alternative cements, like calcium sulfoaluminate cement (CSA) and calcium aluminate cement (CAC)) are comprised primarily of calcium, and our current source of calcium (limestone) is intrinsically carbon-intensive^7,8. Blended cements (like slag cement and limestone calcined clay cement, LC3) can reduce the demand for calcined limestone by half, but even if these materials were adopted globally, their direct emissions would still be 0.75 GtCO₂ y⁻¹. Recycled cement could be separated from concrete waste to supply some of the calcium needed for PC⁹, but if this technology scales, it would supply only a minority fraction of the needed calcium, at best¹⁰.

Because cement’s current calcium source is so carbon-intensive, many consider carbon capture and sequestration (CCS) to be the most realistic solution for decarbonization^7,11, but high costs and reliability have so far limited its adoption¹². The 2.4 billion tons of CO₂ that the cement industry releases from limestone and fossil fuels every year is expensive to capture and put back into the ground. The International Energy Agency estimates a levelized cost of US$60 to $120 to capture each tonne of CO₂ for new CCS-equipped plants with an additional cost up to $55 per tonne CO₂ for storage¹³. At that cost and the current rate of cement production, capturing and storing all the CO₂ we release during cement production would cost at least $276B every year (for comparison, the current annual investment in decarbonization across all sectors is roughly $700B¹⁴). This cost may be a worthwhile exchange to reduce the impacts of climate change on human and ecological health, but so far, and despite CCS investments increasing, the pace of CCS deployment is chronically far behind the rates required in most net-zero plans¹².

We need to use a different source of calcium to accelerate and lower the cost of cement decarbonization. In 2023, a start-up (Brimstone) made PC made using calcium only from carbon-free silicate rocks (basalt and others) at the lab-bench scale, opening a promising new pathway¹⁵. If this technology were to scale, process emissions from limestone would be eliminated. In this paper, we evaluate the viability of sourcing calcium from silicate rocks to decarbonize cement, using four criteria:

1. 1)

   Are there sufficient feedstocks of calcium-rich silicate rocks to supply calcium to the global cement market? We map the estimated abundance of calcium in silicate rocks at Earth’s surface in comparison to the calcium demanded by local cement consumption.

2. 2)

   Is it possible that there exists an energy-efficient process to make PC from silicate rocks? We evaluate the thermodynamic minimum energy needed to convert various feedstocks to PC and compare their associated emissions with a range of energy sources. We also evaluate a process using only proven technologies that can reduce the energy demand of making PC and SCM from silicate minerals while eliminating process emissions.

3. 3)

   What are the potential byproducts and coproducts of silicate-derived calcium? We compare the amount of iron, aluminum, and silica in typical basalts against the global demand for these materials to assess how valuable co-products could meet global demand alongside cement production.

4. 4)

   Can we further reduce the energy demand and total rock extraction by making an alternative (non-PC) cement with silicate-derived calcium? We evaluate the potential advantages of a theoretical novel material that requires less calcium, while also quantifying the cost of increased failure risk that comes with using an unfamiliar material.

We conclude that producing PC from silicate rocks has the potential to completely decarbonize cement while costing less than PC made from limestone. This is because silicate rocks are an abundant source of calcium and other valuable co-products, and our thermodynamic analysis shows that it could be less energy-intensive to produce PC from these rocks if the right process pathway is unlocked. Research into silicate-derived PC has so far been extremely limited, and we discuss the research that is needed to help make carbon-free PC a real cost-competitive solution, potentially reducing global GHG emissions by 4.4%.

Silicate feedstocks are widely available

Historically, PC has used calcium from limestone because the rock is abundant, its calcium concentration is high (up to 50% CaO by mass), the process to make PC requires only heat and the addition of clay, and its impurity concentration is low, eliminating the need for purification⁸. However, nearly 80% of the calcium in Earth’s crust is in silicate rocks, like basalt, that are carbon-free (“Methods”).

Silicate rocks are defined compositionally as those composed of minerals containing silicon and oxygen, along with a wide array of other elements. Silicates include virtually all magma-derived (igneous) rock, as well as the sedimentary and metamorphic rocks that form from the transformation of igneous rocks. The common silicate rocks that contain a substantial amount of calcium are basalt and gabbro, which typically contain between 9% and 12% CaO by mass^16,17 (Table 1). Both of these rock types are composed primarily of the silicate minerals plagioclase and pyroxene, and sometimes contain smaller amounts of olivine and other minerals. Both plagioclase and pyroxene can vary in composition, but basalts and gabbros tend to be enriched in the calcium end-members of these minerals: diopside for pyroxene (CaMgSi₂O₆) and anorthite for plagioclase (CaAl₂Si₂O₈)¹⁸. However, pyroxene and plagioclase rarely occur in these pure end-member forms and metals like sodium and iron substitute for calcium in plagioclase and pyroxene, respectively. So, while the median composition of mafic rocks is shown in Table 1, their elemental composition can vary substantially. This means that cement production can focus on sourcing basalt and gabbro with higher-than-average calcium content, up to 15% CaO (e.g., ref. ¹⁹).

The calcium in these silicate minerals is nearly as abundant as the calcium in limestone at Earth’s surface (“Methods”). Fig. 1b shows the global distribution and abundance of common calcium-bearing rock types that are exposed on Earth’s land surface: limestone and mafic silicate rocks (predominantly basalt and gabbro). We use this abundance to estimate the longevity of calcium resources for feeding the global demand of calcium for consumption in PC (“Methods”). Limestone is considered to be a practically limitless source of calcium for PC^8,20, and our analysis shows why: exposed limestone is capable of feeding cement production at current rates for millions of years in many countries (Fig. 1d). Basalts and gabbros, despite their smaller calcium concentration and lower abundance at Earth’s surface, have a similarly limitless capacity to feed cement production. Exposed mafic silicate rocks have enough calcium to feed cement production for hundreds of thousands of years in most countries (Fig. 1c), presenting no practical limitation to feedstock supply.

The alternative text for this image may have been generated using AI.

Full size image

A shows the average annual PC demand by country between 1998 and 2021 (“Methods”). B shows the extent of calcium-rich silicates (mafic silicate rocks) and carbonates (of which limestone is most common), and exposed at Earth’s surface, based on the GLiM global lithologic map⁶³. Although mafic silicates are most common in oceanic crust and beneath the Earth’s sedimentary shell, they are also abundant at the surface of continents. C, D show the estimated longevity of the calcium supply from silicate and carbonate rocks, respectively. To calculate longevity, we assumed that all rock deposits shown in (B) are 100 m thick. We calculated CaO abundance for each country, assuming that silicate rocks have a density of 2700 kg m⁻³ and are 10% CaO by mass, and that carbonate rocks have a density of 2600 km/m³ and are 52% CaO by mass (Table 1). Longevity is estimated by dividing CaO abundance by demand (A).

Silicate rocks have the potential to make Portland cement with lower energy requirements

The challenge of using silicates as the sole source of calcium for PC is that common calcium-rich silicate rocks contain maximum 15% CaO, by mass, while PC clinker is approximately 68% CaO by mass (Table 1). Thus, to make PC from silicates, it is necessary to concentrate the calcium from silicate rocks. Heat alone will not separate CaO from silica and the other metals in silicate rocks, so a more complex process is required. In 2023, researchers produced PC clinker from calcium-rich silicate rocks using a series of hydro- and pyrometallurgical processes¹⁵, thereby proving that this technology is possible. However, the process to make PC from silicate rocks is nascent in its development, and the early attempts were far from optimized in terms of economics and energy. In contrast, the technology to produce PC from limestone has been refined over the past two centuries¹. In just the past 50 years, process improvements have cut the energy intensity of producing PC from limestone in half²¹. To assess the potential for engineering advancement in making PC from silicate rocks, we consider the thermodynamic limits of the process, in comparison to the limits of the limestone process.

We evaluated the minimum energy needed to produce PC from various rock source materials by calculating the change in enthalpy between the starting Ca-rich mineral and the final products (PC and stoichiometric byproducts), a methodology used previously to evaluate the emissions potential of alternative cements²². The enthalpies of reaction of turning calcite (carbonate), wollastonite (silicate), anorthite (silicate), and diopside (silicate) into a standard PC are shown in Table 2 and the simplified reactions are listed in “Methods”. The theoretical minimum energy needed to convert any of the three silicate minerals into PC is less than half of the theoretical energy needed to convert calcite to PC, and about a third of the energy used to make PC in practice (Table 2)²³.

These results highlight the extraordinary potential for engineering advancement in making PC from silicate rocks: not only does silicate-derived PC eliminate process emissions, but there is potential to reduce the energy intensity of PC production, as well. In Fig. 2, we show the theoretical minimum carbon emissions for limestone-derived PC (ordinary and blended) along with the theoretical minimum carbon emissions of silicate-derived PC, broken into process and energy emissions (assuming natural gas as the energy source). There is an enormous gap in the minimum emissions needed to produce one tonne of PC from silicate rocks (49 kg CO₂) versus limestone (609 kg CO₂). Figure 2 also shows the theoretical minimum emissions for several blended cements. Generally, replacing cement with SCM results in a proportional decrease in emissions. As shown in Fig. 2, common cement blends can achieve a maximum of 40% reduction in total emissions. If we instead blended SCM with silicate-derived PC, we would see a similar proportional decrease in the energy emissions of that blended material, bringing us closer to zero emissions, even with fossil energy.

The alternative text for this image may have been generated using AI.

Full size image

The minimum emissions needed for cement production are broken into limestone decarbonation (process) and natural gas burning (energy). Solid circles represent pure PC and correspond to the natural gas emissions values in Table 2. Open circles represent PC blended with SCMs to reduce emissions and cost: 10% limestone (PLC), 20% fly ash (fly-ash cement), 40% ground granulated blast furnace slag (slag cement), and the LC3 blend (30% calcined clay and 15% limestone added to PC). All cements contain 5% gypsum. This plot shows the limits of cement decarbonization with a limestone feedstock. Even an optimized process and high SCM content will emit 360 kg CO₂ for every tonne of cement.

A lower-energy process to manufacture PC from silicate rocks would also pave the way to eliminate cement emissions entirely. Energy emissions from cement production could be eliminated today with existing technology through the use of renewable fuels²⁴, clean electricity²⁵, and fossil fuels combined with CCS^7,11. However, each of these energy decarbonization solutions has its own set of challenges and costs more than the unabated burning of fossil fuels (Table 2), which is why none have meaningfully reduced carbon emissions from cement production^12,24,25. Lower energy intensities would help ease the transition to decarbonized energy in cement production. Based on calculations in Table 2, producing carbon-free PC from basalt using electricity could be cheaper than producing carbon-intensive PC from limestone using natural gas (Table 2). With an optimized process, it is theoretically possible to completely decarbonize cement with lower energy costs.

Feasibility of silicate-based PC with proven technologies

The range of process pathways to extract and purify calcium, iron, and aluminum from silicate rocks is broad, and the aim of this article is to highlight the potential of silicate rock refining, not to prescribe any particular process pathway. However, to provide some baseline estimates of practical energy requirements (on which future technology development can improve), we consider a simple processing pathway using proven technologies. Silicate rocks theoretically require less energy than calcite to make PC (Table 2), but obvious process pathways using existing technologies have higher energy requirements, if comparing to processes that make cement alone. As discussed below, any excess economic cost and emissions of even obvious processes can be compensated through the co-production of SCM, aluminum, steel, or other co-products. Furthermore, there is massive room for innovation to optimize and develop non-obvious processes for this young technology.

Most options for purifying the metals in silicate rocks will include one or more unit processes to achieve each of the following: leaching metals from silicate minerals, precipitation of metals or metal compounds, and regeneration of any reagents used for leaching and precipitation. Metals can be leached into solution with a broad range of reagents. Precipitation of those metals can be achieved by adding a second reagent (CO₂, H₂SO₄, CaO, NaOH, or numerous others to form carbonates, sulfates, hydroxides, or others, respectively), using electrolysis, or varying pressure and/or temperature. The leaching and/or precipitation steps must be selective for each of the metals to avoid the need for additional purification steps. Finally, in most hydrometallurgical processes, the metals will be bound to an ion of one of the reagents, and additional processing steps are needed to regenerate the reagents and produce a metal as an oxide or hydroxide.

Any process will require energy in the form of heat and/or electricity to make PC from silicate minerals. Energy requirements span an infinite range, with the minimum energies defined in Table 2. As a baseline, we evaluate the practical energy needed to (1) leach calcium from plagioclase with hydrochloric acid (HCl), (2) precipitate Ca(OH)₂ using electrolytically generated base, and (3) calcine PC in a heating step similar to current PC production from CaCO₃ (Fig. S1).

Silicate minerals will generally leach cations at low pH, but reactivity varies between cations and is sensitive to position in the mineral lattice and acid stoichiometry. This sensitivity can be used to selectively leach and separate different cations. Plagioclase exists as a solid solution of calcic and sodic endmembers (CaAl₂Si₂O₈ and NaAlSi₃O₈). Leaching plagioclase with HCl will dissolve Ca²⁺, Al³⁺, and Na⁺. Thermokinetic modeling predicts that the stability of these cations is sensitive to acid loading²⁶. At low concentrations of HCl, only Na is leached. At moderate concentrations, Na and Ca are leached, leaving an aluminosilicate as a solid residue. This leaching will proceed at ambient temperatures but will be faster if heated, following an Arrhenius relationship²⁶. Thus, the leaching step can be achieved with minimal energy input (notwithstanding a tradeoff between reactor output and energy input). The HCl consumed in leaching will be recovered in the precipitation step.

Precipitating Ca(OH)₂ can be achieved by adding NaOH (a strong base) to the solution²⁷:

Any NaCl that was introduced by leaching Na from plagioclase will remain in solution. Solid Ca(OH)₂ precipitate can be used for clinkering. After calcium precipitation, the sodium chloride brine can be electrolyzed to recover both HCl and NaOH for subsequent leaching and precipitation:

Bipolar membrane electrodialysis has been used to perform this reaction for desalination at pilot scale, where NaCl is split with an efficiency of 202 kJ mol⁻¹²⁸. One tonne of PC contains roughly 11,000 moles of CaO, so splitting enough NaCl to leach and precipitate Ca(OH)₂ would require 4.5 GJ. It should be noted that leaching sodium from the rock as NaCl will require additional salt splitting to recover the HCl, which requires energy but will produce excess NaOH that can be sold as an additional co-product²⁹.

Once calcium hydroxide (portlandite) has been produced, it can be calcined and clinkered, similar to traditional PC production, but with more favorable thermodynamics (Table 2). The thermodynamic minimum energy needed to produce PC from portlandite is only 58% of the minimum energy needed to produce PC from calcite (1.1 GJ vs. 1.9 GJ per tonne PC). The actual energy needed to produce one tonne of PC from calcite is 2.9 GJ³⁰, meaning that the process has a thermodynamic efficiency of 66%. If we assume that clinkering portlandite through a similar heating step has the same efficiency as using calcite, then the practical energy requirement for producing PC from portlandite is 1.7 GJ per tonne PC. Adding this clinkering energy and the energy needed to make Ca(OH)₂ while regenerating acid and base (4.5 GJ/tPC), we arrive at a final baseline energy consumption of 6.2 GJ per tonne PC. This energy consumption is six-times the thermodynamic minimum needed to make PC from anorthite and twice the energy that is currently used to make PC from carbonates (Table 2). However, even this inefficient process to make PC from anorthite can dramatically reduce emissions because carbon is not released from limestone (Fig. 3). Moreover, the process is electric and therefore inherently compatible with renewable electricity sources. Additionally, processes that are currently operating at lab-scale, such as hydrogen-depolarized anodes, have the potential to reduce the energy for acid and base regeneration by several GJ per tonne PC³¹.

The alternative text for this image may have been generated using AI.

Full size image

Stacked bars show the emissions sources for producing PC and blended cements with limestone versus silicate feedstocks under more realistic conditions than Fig. 2. Production from limestone is assumed to use state-of-the-art dry-process technology⁶, and production from silicates is assumed to use the baseline process identified in this study. We assumed the worst-case energy supply for both processes: coal is a worse polluter than natural gas but is still used for most PC production⁶⁴, and grid electricity is still mostly generated from burning fossil fuels in the US and globally⁴⁰. The bar charts break energy emissions into two components: the minimum energy needed to produce PC from the feedstock is shown in solid color and the additional energy needed with existing technology is shown in hatched colors (brown for coal burning and blue for grid electricity usage). Creating PC from silicate minerals has three distinct advantages: (1) there are no emissions from limestone decarbonation (red), (2) there is much more room for emissions reduction through technological innovation (larger hatched bars), and (3) because all emissions are from energy, emissions can be eliminated by using non-fossil energy. The main challenge of using a process that relies on electrochemistry is cost (Table 2), but future technologies that improve energy efficiency and rely less on expensive electricity could bring down costs while further reducing emissions.

The solid aluminosilicate residues from anorthite dissolution provide enriched feedstocks for producing aluminum and silica-based products²⁶, mitigating the energy (and thus emissions and cost) needed to make those products. Even if the aluminosilicate residue is not refined further, it can be used directly as SCM without any additional processing or energy³². Because plagioclase is roughly 80% SiO₂ and Al₂O₃, making 1 tonne of PC (65% CaO) would produce at least 6 tonnes of SCM without any additional energy. Currently, producing metakaolin, a popular carbon-free SCM, requires 1.5 GJ per tonne³³. Thus, producing 1 tonne of PC and 6 tonnes of SCM from plagioclase would displace 8.9 GJ of energy consumption. Using anorthite as the calcium feedstock would therefore cut 2.7 GJ (30%), relative to conventional processes, while cutting out process emissions entirely. A ratio of SCM to PC of 6:1 cannot scale globally because cement blends are unlikely to contain more than 50% of SCM in general purpose construction. However, as fly ash and blast furnace slag become increasingly scarce^34,35, the production of excess SCM in the first generation of decarbonized PC plants could help fill the existing gap in SCM supply. Future plants can focus on further refinement of the solid residue to produce aluminum and other silica-based products.

The potential to cogenerate aluminum and steel

The coproduction potential of silicate-sourced PC extends far beyond SCM. Another advantage of using calcium from silicate rocks instead of from limestone is the potential for cogeneration of industrial metals. Calcium-rich silicate rocks (like basalt) are composed almost entirely of silica and metal oxides, including aluminum, iron, and magnesium oxides (Table 1). Silica, aluminum, and iron are currently added to limestone-based PC from materials mined offsite³⁶. Silicate rocks contain the necessary metals to produce PC from a single rock, with substantial quantities of valuable industrial materials as co-products (Table 1). If we produced all PC using calcium from typical basalts, those same rocks could be processed to meet the entire global demand for steel, aluminum, and SCM (Table 1)³⁷. The mass ratio of global consumption for calcium (in cement) and iron (in steel and cement) almost exactly matches the mass ratio of those metals in basalt (Table 1), making widespread co-production from basalt economically attractive. Moreover, magnesium oxide can readily mineralize CO₂ into magnesium carbonate (magnesite), meaning that silicate-derived cement could provide a source of carbon removal from the same rock.

Iron, aluminum, and magnesium oxides occur in the same mineral phases as calcium (primarily pyroxene and plagioclase), so a process that extracts calcium will leave a residue enriched in these valuable metals. The potential to extract and sell several products from the same rock could reduce the overall cost of production for cement, steel, and aluminum. The co-production of these three major industrial materials in a rock refinery process also has the potential to reduce the number of mines and the amount of mine waste currently produced from these industries. Producing 1 tonne of aluminum or iron ore requires mining ~10 tonnes rock, most of which is waste³⁸. Currently, 13 Gt of rock is mined annually to produce 1.1 Gt of iron (equivalent to 1.4 Gt FeO)³⁸. Separate mines extract limestone for PC (at least 3.2 Gt of rock) and bauxite for aluminum (0.6 Gt of rock)³⁸. Byproducts from coal-fired electricity generation currently supply much of the world’s SCM, and that industry extracts roughly 12 Gt of rock annually (10⁷ GWh of electricity at 1200 t-rock GWh⁻¹)^39,40. The 16 Gt of basalt needed to supply all the CaO for PC could replace the 16.8 Gt of rock currently mined for steel, calcium, and aluminum, while reducing reliance on the coal industry for SCM (Table 1). If achieved, refining basalt into calcium, iron, aluminum, and silicon oxides could meaningfully reduce global mine waste. It should be noted that, to our knowledge, an efficient rock-refinery process has not been publicly demonstrated, but is an area of immense research opportunity.

The opportunities and challenges of making alternative cements

The previous sections describe the advantages of making PC using calcium from silicate rocks. If we remove the constraint of making PC and instead consider making alternative cement using silicate rocks, then there is even greater potential to reduce carbon emissions. Alternative cements can have two advantages over PC for decarbonization: (1) they can contain less calcium, thereby reducing the need to concentrate the calcium from silicate rocks (e.g., ref. ⁴¹); and (2) some can be produced without conventional high-temperature clinkering, thereby reducing energy requirements of cement production (e.g., ref. ²²). However, alternative cements have historically not been adopted for general use in construction¹, and relying on alternative cements for decarbonization poses the risk that the construction industry will simply not use them. Here, we briefly seek to understand the major market barriers to adopting alternative cements for general use and what is needed to overcome those barriers in order to minimize the calcium and energy demands of cement production.

Several alternative cements have been invented in the two centuries since the invention of PC. Many of them have some advantage(s) over PC in terms of strength, curing time, curing temperature, workability, or emissions^42,43. Alternative cements include clinkered materials (e.g., CAC, CSA), calcined materials (e.g., magnesium oxychloride cement, magnesium phosphate cement), or non-clinkered materials (e.g., alkali-activated cements, precipitated calcite cements, and biologic cements); and they can be hydraulic (meaning that they harden when combined with water) or nonhydraulic. While some alternative cements like alkali-activated cements are promoted for general use concrete⁴³, others are often limited to targeted applications due to differences in concrete behavior compared to PC. For example, CAC is typically used for refractory concrete, while CSA is typically used for rapid strength concrete⁴¹. Some of these alternative cements have a large emissions advantage, emitting as little as 44% of the carbon emitted during PC production⁴². Unfortunately, none of these reduced-carbon materials have gained relevant ( > 0.1%) market share, and thus their impact on cement decarbonization has been minimal.

One explanation for the slow adoption of alternative cements is the lack of centuries-old buildings that demonstrate their serviceability and durability, and the associated risks of these unknowns is difficult to mitigate without long-term observations of many buildings. We estimated the potential costs of unknown risks using a simple probabilistic model (“Methods”). Assessments of building collapse events indicate that structural failure and subsequent collapse of a typical commercial building costs tens of millions of U.S. dollars⁴⁴. When observational data is limited and confidence in new materials is low, the perceived cost of the unknown risk can be prohibitively high⁴⁵. As the number of buildings standing with a novel material n grows, the empirical probability of collapse from unknown risks will decrease as \(\frac{1}{n}\). For the market to adopt the new material, we assume that customers will not freely adopt it until the price of the unknown risk is demonstrated to be negligible. Based on a range of prices that customers are willing to pay for cement in the same market⁴⁶, we estimate that the market will tolerate a risk cost of $10 per tonne PC. Our analysis suggests that this level of confidence in the risk is achieved only after the construction of approximately 13,000 buildings (Fig. S1) or approximately $50B of construction. Because each novel cement will have its own potential for new unknown risks, each new material will need its own set of demonstration projects. Additionally, for customers to be confident in the long-term serviceability of the material, they will likely want observational data for the entire lifecycle of their building, which can be on the order of 50 years or more⁴⁷.

This analysis represents an upper-bound estimate of the market barriers to adopting alternative materials. Governments could accelerate the adoption of novel materials by some combination of mandates and financial incentives, eliminating the market barriers. Also, some adopters may discount these risks. However, the reality that buildings contain people and building collapse may cause loss of life may help to explain why essentially zero consumers have been willing to adopt an alternative cement as a complete replacement for PC.

Greater risk of structural failure is only one of the challenges with using novel materials. Other factors, including construction delays, workability, availability, safety requirements, and maintenance, are all potential challenges that can be translated into potential financial costs, which can help us understand why they matter to the businesses responsible for building with concrete⁴⁵. A compounding issue is that alternative cements are not as widely available as PC and are seldom specified in project specifications, making their inclusion into structures slow and further delaying the growth of needed empirical observations and data. Unfortunately, if society waits decades for observational data from thousands of demonstration projects, alternative cements will not be widely adopted with enough time to contribute substantially to 2050 decarbonization goals.

Next steps for cement decarbonization

Mafic silicate rocks are an abundant source of calcium (Fig. 1) with favorable thermodynamics for making PC (Table 2 and Fig. 2). However, research into silicate-derive PC has been extremely limited. Research is urgently needed to develop energy-efficient chemical processes that recombine calcium, silicon, aluminum, and iron from silicate rocks into the correct proportions needed for PC (Table 1). Additional work to further refine aluminum and iron from silicate residues could help accelerate the market viability of silicate-based PC and help decarbonize those industries, as well. Hydro- and pyrometallurgy are obvious process pathways for developing cost-competitive cement from silicate rocks, owing to the low cost of heat and the abundance of affordable industrial reagents (Table 2). And, while most industrial heating is currently accomplished by burning fossil fuels due to cost, there are several options for renewable heat sources, from burning waste or biomass to harnessing renewable electricity through heat pumps, resistive heating, or electric arc furnaces^24,25. Electrochemical processes like the one identified in this study offer an alternative pathway that is inherently compatible with electricity sources but is also likely to be more expensive because of the high costs of electricity relative to fuel (Table 2). However, electrochemical or electrothermal processes could be cost-competitive with traditional PC if there was a creative solution to convert silicate rocks to PC with electricity requirements close to the thermodynamic minimum, if the relative cost of dispatchable electricity were to decrease, and/or if the process were able to produce other valuable co-products. There are innumerable potential process pathways to make PC from silicate rocks, warranting wide research attention.

Combining cement decarbonization solutions

While we have demonstrated the potential for making cost-competitive, decarbonized PC from silicate rocks, this solution is compatible with other decarbonization technologies. Silicate-derived PC can be blended with SCMs, clinkered with renewable energy sources, used in leaner building designs, and recycled exactly the same as traditional PC^9,48. And if ten thousand demonstration projects for alternative cements are built over the next several decades to overcome adoption barriers, they are still likely to require large amounts of calcium. Silicate rocks can provide the calcium for these alternative cements through the metallurgical processes developed to produce PC.

In short, technologies can be combined to achieve net-zero emissions and decarbonization could be considered to have three fronts: (1) developing the technology to source calcium from silicate rocks to produce PC, (2) further testing and demonstration projects to understand the risks of lower-calcium alternative cements, and (3) using renewable sources of energy (either fuel or electricity) to fully decarbonize cement. CCS could prove to be an expensive stopgap solution while these other solutions are in development, but it comes with risks: retrofitting cement plants to accommodate CCS would lock markets into high cement prices for decades to cover the cost of capital expenditures⁴⁹ and could delay the rollout of new technologies aiming to replace aging cement plants. In contrast, investing into the development of silicate-derived PC using renewable energy could unlock a cheaper way to build while reducing carbon emissions.

Abundance and longevity of calcium-bearing rocks

In order to assess any potential resource limitations with supplying calcium from silicate rocks, we estimated how calcium is distributed in different rocks within Earth’s crust, with particular attention to the exposure of calcium-bearing rocks at Earth’s surface.

We estimate that 80% of the crust’s calcium is in silicate rocks by estimating the total calcium content of the crust and how much of that calcium is in non-silicate rocks. Based on the CRUST2.0 Model, the total mass of Earth’s crust is 2.77 × 1013 Mt⁵⁰. The crust is 4.15% calcium by mass (or 1.1 × 1012 Mt Ca) based on chemical analyses of glass produced during the impact large meteorites that locally melted and homogenized the upper and lower crust⁵¹. That 10¹² Mt of calcium is distributed predominantly in silicate rocks (like basalt and gabbro), carbonate rocks (like limestone), and evaporites (like gypsum). The total mass of carbon in carbonate rocks is estimated to be 6.53 × 1010 Mt C⁵². If we assume that all carbonate is CaCO₃ (which is an upper-bound, because MgCO₃ is also common), then there is roughly 2.2 × 1011 Mt Ca in carbonate rocks. According to Ronov (1982)⁵³, there is 23-times as much carbonate rocks as chlorites and sulfites, so even if we assumed that all sulfites were gypsum, there would be a maximum of 5.6 × 109 Mt Ca in evaporites, which is an extreme overestimate. Thus, calcium is roughly partitioned in Earth’s crust with 20% in limestone and 80% in silicate rocks.

However, limestone primarily occurs at Earth’s surface, while silicate rocks are distributed throughout Earth’s crust, with higher calcium concentrations in the deep continental crust and oceanic crust. Thus, much of silicate calcium is inaccessible. For a more practical analysis of calcium distribution, we rely on geologic maps and estimates of rock composition for mapped rock types. The global lithologic map (GLiM) provides a synthesized geologic map with broad rock classifications, including “carbonate sedimentary rocks” (predominantly limestone), “basic volcanic rocks” (predominantly basalt), and “basic plutonic rocks” (predominantly gabbro). We combined “basic volcanic rocks” and “basic plutonic rocks” to estimate the abundance and distribution of Ca-rich silicate rocks, assuming that the average composition of these rocks is equal to the standardized basalt defined in Table 1. We assumed that the average composition of “carbonate sedimentary rocks” is equal to limestone in Table 1. The distribution of these rocks is mapped in Fig. 1b. We assumed deposit thicknesses of 100 m for both limestone and silicates. Deposits can easily be much thicker than this, especially for silicate rocks⁵⁴, but we limited the analysis to 100 m to reflect reasonable rock quarry depths. To estimate the mass of accessible calcium from both calcium-rich silicates and limestone, we calculated the product of areal extent, characteristic quarry depth (100 m), respective rock density (2.6 t m⁻³ for limestone and 2.7 t m⁻³ for Ca-rich silicates), and respective calcium grade (Table 1). We divided these calcium abundances by country.

In order to estimate the longevity of these calcium sources for PC demand, we compiled data on global PC production from the USGS Mineral Commodity reports spanning 1998 to 2021³⁷, and global PC trade from the UN Comtrade Database⁵⁵. The UN Comtrade Database lists cement under 3 commodity IDs: “252310” (clinker), “252329” (cement, portland), and “252390” (cement, hydraulic). After data cleaning to remove outlier values, net imports of cement (imports minus exports) were calculated for every country and every year available between 1998 and 2021. For each country, we calculated the average PC demand by summing average production values and average net import values (Fig. 1a). Using the calcium abundances from carbonates and Ca-rich silicates, we roughly estimated the longevity of calcium supplies from those two rocks by dividing calcium abundance by the calcium consumed by PC demand (Fig. 1c, d).

Minimum energy calculations

To calculate the minimum energy needed to make PC with different minerals as the primary calcium source, we calculated the change in enthalpy between the final products and the initial reactants. We assumed the same final composition of PC for each starting material: 35.6% Ca₃SiO₅, 27.6% Ca₂SiO₄, 8.6% Ca₃Al₂O₆, and 27.3% Ca₄Al₂Fe₂O₁₀ (all weight percent). To meet these mass fractions, we added reactants to the calcium source, including kaolinite (Si₂Al₂O₅(OH)₄), alumina (Al₂O₃), quartz (SiO₂), and hematite (Fe₂O₃). For non-silicate calcium sources, kaolinite is first added to balance stoichiometry with respect to aluminum. For all minerals, we then added alumina, quartz, and hematite to balance the stoichiometry with respect to aluminum, silica, and iron in PC. This results in the following chemical reactions to produce 950 kg of PC clinker from each mineral feedstock (resulting in 1 tonne of PC after the addition of 50 kg gypsum):

Calcite

Portlandite

Wollastonite

Anorthite

Diopside

We then use molar enthalpy values at standard temperature and pressure from the literature⁵⁶, reported in Table S1, to calculate the change in enthalpy needed to make 1 tonne of PC from each feedstock, using the following equation:

where n is the number of moles, M is the molar mass (Table S1), and \({\triangle H}_{f}^{0}\) is the enthalpy of formation (Table S1).

Calculating the threshold for market adoption of novel materials

To quantify the confidence level of building with a new material and estimate timescales of adoption, we developed a simple probabilistic model to estimate the cost of unknown risk. We can estimate the average cost of material failure in terms of failure probability (P_f) and the ratio of total economic loss (c_T) to building replacement cost (c_b): \(R=\frac{{C}_{T}}{{C}_{b}}\), where c_T includes damage to surrounding properties, casualties, business interruption, and litigation costs. For reference, the Twin Tower and Pentagon building collapses during the 9/11 U.S. terror attacks had R values of 20 and 40, respectively⁴⁴, while we estimate that the Surfside condominium collapse had a value of R = 70. The cost of the risk, per tonne of PC used in construction is:

where c_b is the cost of construction per unit area and u is the mass of cement used per unit area. Probability of failure is difficult to assess for a novel material, owing to the potential for unknown risks, as evidenced by the failures of CAC structures^57,58. Consequently, we conservatively estimate that the risk of failure from unknown material behavior is the observed probability of failure:

where n_f is the number of failures caused by an unknown material behavior and n_b is the number of buildings constructed with the new material. For a material to continue to be used, we assume that any building failure must be understood and remedied in future use, meaning that n_f must remain zero for all previous buildings constructed and takes a maximum value of n_f = 1 for the possibility of failure of the next building. Consequently, the risk cost decreases as \({c}_{r}\propto \frac{1}{{n}_{b}}\) (Fig. S2). Substituting these values into Eq. (2) and solving for n_b:

If we use typical values for the construction of buildings in the US (u = 0.38 tPC m⁻² and c_b = $2500 m⁻²), a conservative economic loss ratio of R = 20, a tolerated cost of risk based on the existing range of prices for cement within a given market c_r = ~$10 tPC⁻¹, we estimate that n_b = 13,158 buildings would be needed to fully demonstrate a negligible risk for using a new material. This represents an upper-bound estimate from a very simple model.

In reality, the factors driving the adoption of new materials are complex. The risk tolerance will vary between building types and customer. Government subsidies, government regulations, or cheaper material costs could add incentives to adopt the material faster. And non-economic incentives, including the desire to meet decarbonization goals, could also incentivize early adoption. This model is nevertheless useful for quantifying the potential market barriers to the adoption of a new material.

All data presented in this manuscript (Tables 1, 2, and S1 and Figs. 1–3) are available here: https://doi.org/10.6084/m9.figshare.31667374.

1. Scrivener, K. L. & Snellings, R. The rise of Portland cements. Elements 18, 308–313 (2022).

   Article  Google Scholar 

2. Bernal, S. A., Rodríguez, E. D., Kirchheim, A. P. & Provis, J. L. Management and valorisation of wastes through use in producing alkali-activated cement materials. J. Chem. Tech. Biotech. 91, 2365–2388 (2016).

   Article  Google Scholar 

3. International Energy Agency. World Energy Outlook https://www.iea.org/reports/world-energy-outlook-2023 (Paris, 2023).

4. Bajželj, B., Allwood, J. M. & Cullen, J. M. Designing climate change mitigation plans that add up. Environ. Sci. Technol. 47, 8062–8069 (2013).

   Article  Google Scholar 

5. Global Cement and Concrete Association (GCCA). GNR 2.0 - GCCA in Numbers https://gccassociation.org/sustainability-innovation/gnr-gcca-in-numbers/.

6. DOE Advanced Manufacturing Office. Bandwidth Study on Energy Use and Potential Energy Savings Opportunities in U.S. Cement Manufacturing. https://www.osti.gov/biblio/1512370 (DOE/EE-1660, DOE Energy Efficiency and Renewable Energy, United States, 2017).

7. Goldman, S. et al. DOE Pathways to Commercial Liftoff: Low-Carbon Cement https://liftoff.energy.gov/industrial-decarbonization/low-carbon-cement/ (US Department of Energy, 2023).

8. Oates, J. A. H. Lime and Limestone: Chemistry and Technology, Production and Uses (Wiley-VCH, Weinheim, 1. Auflage, 2008).

9. Dunant, C. F., Joseph, S., Prajapati, R. & Allwood, J. M. Electric recycling of Portland cement at scale. Nature 629, 1055–1061 (2024).

   Article  Google Scholar 

10. Miller, S. A. Combined cement and steel recycling could cut CO₂ emissions. Nature 629, 1008–1009 (2024).

    Article  Google Scholar 

11. Bouckaert, S. et al. Net Zero by 2050: A Roadmap for the Global Energy Sector. International Energy Agency (2021).

12. Martin-Roberts, E. et al. Carbon capture and storage at the end of a lost decade. One Earth 4, 1569–1584 (2021).

    Article  Google Scholar 

13. International Energy Agency. Is Carbon Capture Too Expensive? https://www.iea.org/commentaries/is-carbon-capture-too-expensive (Paris, 2021).

14. International Energy Agency. Government Energy Spending Tracker https://www.iea.org/reports/government-energy-spending-tracker-2 (Paris, 2023).

15. Brimstone carbon-negative cement meets ASTM standard. International Cement Review. https://www.cemnet.com/News/story/175208/brimstone-carbon-negative-cement-meets-astm-standard.html (2023).

16. Ewart, A. Mineralogy and chemistry of modern orogenic lavas—some statistics and implications. Earth Planet. Sci. Lett. 31, 417–432 (1976).

    Article  Google Scholar 

17. Gale, A., Dalton, C. A., Langmuir, C. H., Su, Y. & Schilling, J. The mean composition of ocean ridge basalts. Geochem. Geophys. Geosyst. 14, 489–518 (2013).

    Article  Google Scholar 

18. Gill, R. Igneous Rocks and Processes: A Practical Guide (Wiley, Somerset, 2011).

19. Re, G., Palin, J. M., White, J. D. L. & Parolari, M. Unravelling the magmatic system beneath a monogenetic volcanic complex (Jagged Rocks Complex, Hopi Buttes, AZ, USA). Contrib. Miner. Pet. 172, 94 (2017).

    Article  Google Scholar 

20. Scrivener, K., Martirena, F., Bishnoi, S. & Maity, S. Calcined clay limestone cements (LC3). Cem. Concr. Res. 114, 49–56 (2018).

    Article  Google Scholar 

21. Worrell, E., Kermeli, K. & Galitsky, C. Energy Efficiency Improvement and Cost Saving Opportunities for Cement Making: An ENERGY STAR® Guide for Energy and Plant Managers (Document 430-R-13–009, US Environmental Protection Agency, 2013).

22. Miller, S. A. & Myers, R. J. Environmental impacts of alternative cement binders. Environ. Sci. Technol. 54, 677–686 (2020).

    Article  Google Scholar 

23. International Energy Agency. Global Thermal Energy Intensity of Clinker Production by Fuel in the Net Zero Scenario, 2010-2030 https://www.iea.org/data-and-statistics/charts/global-thermal-energy-intensity-of-clinker-production-by-fuel-in-the-net-zero-scenario-2010-2030 (2023).

24. Luo, Z., Song, H., Huang, Y. & Jin, B. Recent advances on the uses of biomass alternative fuels in cement manufacturing process: a review. Energy Fuels 38, 7454–7479 (2024).

    Article  Google Scholar 

25. Quevedo Parra, S. & Romano, M. C. Decarbonization of cement production by electrification. J. Clean. Prod. 425, 138913 (2023).

    Article  Google Scholar 

26. Neron, T. et al. Thermo-kinetic modelling of the acidic leaching of anorthosite: Key learnings toward the conception of a sustainable industrial process. Miner. Eng. 180, 107500 (2022).

    Article  Google Scholar 

27. Um, N. & Hirato, T. Precipitation behavior of Ca(OH)2, Mg(OH)2, and Mn(OH)2 from CaCl2, MgCl2, and MnCl2 in NaOH-H2O solutions and study of lithium recovery from seawater via two-stage precipitation process. Hydrometallurgy 146, 142–148 (2014).

    Article  Google Scholar 

28. Cassaro, C. et al. Electrodialysis with bipolar membranes for the sustainable production of chemicals from seawater brines at pilot plant scale. ACS Sustain. Chem. Eng. 11, 2989–3000 (2023).

    Article  Google Scholar 

29. Contreras, R. R., Almarza, J., Rincón, L. & Ruiz, C. Green sodium hydroxide for industrial purposes. A short review. J. Environ. Chem. Eng. 13, 114972 (2025).

    Article  Google Scholar 

30. Mokhtar, A. & Nasooti, M. A decision support tool for cement industry to select energy efficiency measures. Energy Strategy Rev. 28, 100458 (2020).

    Article  Google Scholar 

31. Liu, J. et al. Hydrogen-motivated electrolysis of sodium carbonate with extremely low cell voltage. Chem. Commun. 54, 3582–3585 (2018).

    Article  Google Scholar 

32. Shah, V. & Scott, A. Pozzolanic characteristics of silica recovered from olivine. Constr. Build. Mater. 332, 127378 (2022).

    Article  Google Scholar 

33. Martinez, D. M., Horvath, A. & Monteiro, P. J. M. Comparative environmental assessment of limestone calcined clay cements and typical blended cements. Environ. Res. Commun. 5, 055002 (2023).

    Article  Google Scholar 

34. Diaz-Loya, I., Juenger, M., Seraj, S. & Minkara, R. Extending supplementary cementitious material resources: reclaimed and remediated fly ash and natural pozzolans. Cem. Concr. Compos. 101, 44–51 (2019).

    Article  Google Scholar 

35. Knight, K. A., Cunningham, P. R. & Miller, S. A. Optimizing supplementary cementitious material replacement to minimize the environmental impacts of concrete. Cem. Concr. Compos. 139, 105049 (2023).

    Article  Google Scholar 

36. Bye, G. C. Portland Cement: Composition, Production and Properties 2nd edn (Thomas Telford, London, 1999).

37. USGS. Mineral Commodity Summaries 2023 https://doi.org/10.3133/mcs2023 (2023).

38. Nassar, N. T., Lederer, G. W., Brainard, J. L., Padilla, A. J. & Lessard, J. D. Rock-to-metal ratio: a foundational metric for understanding mine wastes. Environ. Sci. Technol. 56, 6710–6721 (2022).

    Article  Google Scholar 

39. Wang, S., Cook, P., Stein, A., Lloyd, J. & Smith, C. Updated mining footprints and raw material needs for clean energy. Breakthrough Institute https://thebreakthrough.org/issues/energy/updated-mining-footprints-and-raw-material-needs-for-clean-energy (2024).

40. International Energy Agency. Electricity 2025 https://www.iea.org/reports/electricity-2025/emissions (Paris, 2025).

41. Hanein, T., De La Torre, A. G., Zhang, Z. & Provis, J. L. Alternative non-Portland binders. Elements 18, 314–320 (2022).

    Article  Google Scholar 

42. Burris, L., Kurtis, K. & Morton, T. Novel Alternative Cementitious Materials for Development of the Next Generation of Sustainable Transportation Infrastructure (TechBrief FHWA-HRT-16-017, US Federal Highway Administration, 2015).

43. Singh, N. B. & Middendorf, B. Geopolymers as an alternative to Portland cement: an overview. Constr. Build. Mater. 237, 117455 (2020).

    Article  Google Scholar 

44. Stewart, M. G. Risk of progressive collapse of buildings from terrorist attacks: are the benefits of protection worth the cost? J. Perform. Constr. Facil. 31, 04016093 (2017).

    Article  Google Scholar 

45. Nguyen, H. D. & Macchion, L. Risk management in green building: a review of the current state of research and future directions. Environ. Dev. Sustain. 25, 2136–2172 (2023).

    Article  Google Scholar 

46. Miller, N. H. & Osborne, M. Spatial differentiation and price discrimination in the cement industry: evidence from a structural model. RAND J. Econ. 45, 221–247 (2014).

    Article  Google Scholar 

47. Palacios-Munoz, B., Peuportier, B., Gracia-Villa, L. & López-Mesa, B. Sustainability assessment of refurbishment vs. new constructions by means of LCA and durability-based estimations of buildings lifespans: a new approach. Build. Environ. 160, 106203 (2019).

    Article  Google Scholar 

48. Olsson, J. A., Miller, S. A. & Alexander, M. G. Near-term pathways for decarbonizing global concrete production. Nat. Commun. 14, 4574 (2023).

    Article  Google Scholar 

49. Hughes, S. et al. Analysis of Carbon Capture Retrofits for Cement Plants https://doi.org/10.2172/1970135 (DOE/NETL–2023/3856, 1970135, 2023).

50. Peterson, B. T. & Depaolo, D. J. Mass and Composition of the Continental Crust Estimated Using the CRUST2.0 Model V33A-1161 https://ui.adsabs.harvard.edu/abs/2007AGUFM.V33A1161P (San Francisco, CA, USA, 2007).

51. Rumble, J. R., Brunno, T. J. & Doa, M. J. (eds) CRC Handbook of Chemistry and Physics: A Ready-Reference Book of Chemical and Physical Data 2023–2024 (CRC Press, Boca Raton, London, New York, 104th edn, 2023).

52. Li, Y.-H. A Compendium of Geochemistry: From Solar Nebula to the Human Brain (Princeton University Press, Princeton, N.J, 2000).

53. Ronov, A. B. The Earth’s sedimentary shell (quantitative patterns of its structure, compositions, and evolution): the 20th V.I. Vernadskiy Lecture, March 12, 1978. Int. Geol. Rev. 24, 1313–1363 (1982).

    Article  Google Scholar 

54. Ross, P.-S. et al. Mafic volcaniclastic deposits in flood basalt provinces: a review. J. Volcanol. Geotherm. Res. 145, 281–314 (2005).

    Article  Google Scholar 

55. United Nations Statistics Division. UN Comtrade: International Merchandise Trade Statistics (1998 to 2021) https://comtradeplus.un.org/TradeFlow.

56. Robie, R. A., Waldbaum, D. R. Thermodynamic Properties of Minerals and Related Substances at 298.15 K and 1 bar (105 Pascals) Pressure and at Higher Temperatures https://doi.org/10.3133/b1452 (1452, US Geological Survey, 1978).

57. Midgley, H. G. & Midgley, A. The conversion of high alumina cement. Mag. Concr. Res. 27, 59–77 (1975).

    Article  Google Scholar 

58. Scrivener, K. L. Calcium Aluminate Cements. In Advanced Concrete Technology 1: Constituent Materials (ed. Newman, J. & Choo, B. S.) 1–31 (Butterworth-Heinemann, 2003).

59. Global Cement and Concrete Association. Concrete Future – Roadmap to Net Zero (0.1, Global Cement and Concrete Association, 2021).

60. International Energy Agency. IEA Energy Prices and Taxes Statistics https://doi.org/10.1787/eneprice-data-en (OECD, 2023).

61. IRENA. Renewable Power Generation Costs in 2024 (International Renewable Energy Agency, Abu Dhabi, 2025).

62. U.S. Energy Information Administration - EIA - Independent Statistics and Analysis. https://www.eia.gov/environment/emissions/co2_vol_mass.php.

63. Hartmann, J. & Moosdorf, N. The new global lithological map database GLiM: a representation of rock properties at the Earth surface. Geochem. Geophys. Geosyst. 13, Q12004 (2012).

64. Clark, G., Davis, M., Shibani, A. & Kumar Assessment of fuel switching as a decarbonization strategy in the cement sector. Energy Convers. Manag. 312, 118585 (2024).

    Article  Google Scholar 

Download references

Author notes

1. These authors contributed equally: Jeff P. Prancevic, Cody E. Finke.

Authors and Affiliations

1. Earth Research Institute, University of California, Santa Barbara, Santa Barbara, CA, USA

   Jeff P. Prancevic

2. Brimstone Energy, Inc., Reno, NV, USA

   Jeff P. Prancevic, Cody E. Finke & Wilson Nguyen

3. Webcor Builders, San Francisco, CA, USA

   Eric Peterson

4. Department of Civil and Environmental Engineering, University of Virginia, Charlottesville, VA, USA

   Andres F. Clarens

5. Energy Systems Division, Interdisciplinary Science Department, Brookhaven National Laboratory, Upton, NY, USA

   Tatiana Pyatina

Contributions

C.E.F. conceived the study and analyzed thermodynamic minimum energies. J.P.P. performed all other analyses, produced the figures, and led manuscript writing. C.E.F. and J.P.P. jointly developed the study concepts. E.P. guided analysis and conceptualization regarding construction industry requirements. W.N. provided data and expertise on the composition and properties of ordinary and blended Portland cements. A.F.C. conceptualized the analysis of practical emissions of silicate-sourced PC. T.P. provided data and expertise on alternative cements. All authors contributed to manuscript writing and approved the final version.

Corresponding author

Correspondence to Jeff P. Prancevic.

Competing interests

C.E.F. is the founder and CEO of Brimstone, an industrial decarbonization company that is developing technology to use for the co-production of PC and other industrial metals. W.N. is currently an employee of Brimstone. J.P.P. was employed by Brimstone from May 2022 to February 2025 and owns shares of the company.

Peer review information

Communications Sustainability thanks Mai Uibu and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Jiachao Peng and Nandita Basu. A peer review file is available.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions