Perspective

The refinery of the future

https://doi.org/10.1038/s41586-024-07322-2 Eelco T. C. Vogt1 ✉ & Bert M. Weckhuysen1 ✉

Received: 13 April 2021

Accepted: 15 March 2024 Fossil fuels—coal, oil and gas—supply most of the world’s energy and also form the
Published online: 8 May 2024 basis of many products essential for everyday life. Their use is the largest contributor
to the carbon dioxide emissions that drive global climate change, prompting joint
Check for updates
efforts to find renewable alternatives that might enable a carbon-neutral society by as
early as 2050. There are clear paths for renewable electricity to replace fossil-fuel-
based energy, but the transport fuels and chemicals produced in oil refineries will
still be needed. We can attempt to close the carbon cycle associated with their use
by electrifying refinery processes and by changing the raw materials that go into a
refinery from fossils fuels to carbon dioxide for making hydrocarbon fuels and to
agricultural and municipal waste for making chemicals and polymers. We argue that,
with sufficient long-term commitment and support, the science and technology for
such a completely fossil-free refinery, delivering the products required after 2050
(less fuels, more chemicals), could be developed. This future refinery will require
substantially larger areas and greater mineral resources than is the case at present and
critically depends on the capacity to generate large amounts of renewable energy for
hydrogen production and carbon dioxide capture.

Population growth and the increasing consumption associated with materials and land resources sufficient for designing and construct-
a higher standard of living have led to increasing anthropogenic car- ing carbon-neutral refineries? (4) What are the costs associated with
bon dioxide (CO2) emissions that cannot be absorbed by the natural building future refineries and how can we make them economically
geological and biological carbon cycles. Because these emissions and socially viable?
change Earth’s climate with adverse effects such as more frequent In this Perspective, we attempt to provide some answers to the ques-
and severe droughts, heatwaves and rainfall, the aim is to reduce CO2 tions above by sketching a possible roadmap towards a refinery of the
in the atmosphere (about 424 ppm in May 2023 (ref. 1)), as set out future. We do this using order-of-magnitude calculations for a single
in the Paris Agreement and the Kyoto Protocol2,3. Applicable to us refinery plant that we assume to be located in Europe. We first discuss
is the European Climate Law, which provides a legal framework for the necessary building blocks for the anticipated refinery concept.
the rapid reduction of CO2 emissions and targets an economy that is We then evaluate the required land and offshore area, resources and
carbon-neutral by 2050 (ref. 4). Because a large fraction of current CO2 investments costs and consider the necessary scientific and techno-
emissions arises from the production and use of energy, achieving this logical developments that are needed to make this all possible. We
ambitious goal will probably involve an energy system based largely also highlight some possible show-stoppers, which include the need
on renewable electricity, with a prominent role for solar photovoltaics for large amounts of renewable energy and so-called critical chemical
(PV) and wind power5–7. elements to construct all the necessary hardware.
But there are less obvious solutions for how to reduce the emissions
of refineries, which—at present—process fossil-based resources to
make our transport fuels, performance chemicals and monomers for Building blocks of a refinery of the future
polymer production. Transport fuels, which are the main products of Crude oil refining as we know it today took off in the early 1900s
today’s refineries and contribute about 25% of present-day CO2 emis- because petroleum-based products were increasingly used for trans-
sions, can—in part—be replaced through transport electrification. But port8. Large-scale chemical processes, such as cracking, hydropro-
electrification will not be feasible for all transport modes and there will cessing, isomerization, reforming and alkylation, were developed
be an increasing need for chemicals and polymers as both the world to allow crude oil refineries to produce ever more complex product
population and the material needs of developing countries grow. slates that included transport fuels (that is, gasoline, kerosene and
Developing alternative, carbon-neutral refining concepts that focus diesel), heating oil and later on also petrochemicals such as ethylene,
on sustainability and circularity will thus be essential for transitioning propylene and benzene, toluene and xylenes9–13. Today’s refinery and
towards a carbon-neutral economy in 2050. The key questions that associated petrochemical complex produce essentially all transport
need to be answered for a carbon-neutral refining concept are as fol- fuels and the essential raw materials for polymers (plastics, resins,
lows. (1) What carbon input will replace the fossil-based resources to fibres), detergents, coatings, construction chemicals and medicine.
enable a carbon-neutral refinery? (2) Can science and technology pro- Along with changes to the product slate, economy of scale and pro-
gress rapidly enough so that we can replace in time existing conversion cess integration led to the construction of ever larger petrochemical
processes with more sustainable alternatives? (3) Are available energy, complexes and refineries14.

Inorganic Chemistry and Catalysis Group, Department of Chemistry, Faculty of Science, Utrecht University, Utrecht, the Netherlands. ✉e-mail: e.t.c.vogt@uu.nl; b.m.weckhuysen@uu.nl
1

Nature | Vol 629 | 9 May 2024 | 295

Perspective
a Refinery b Electricity generation c Hydrogen generation
2020 Hydrogen-based
Other
43% Gasoline Blue H2

Oil 86% Solar

150,000 bpd 28% Diesel

10% Jet fuel

Other 14%
19% Other
2050 (including polymers Wind
and chemicals)
Green H2
Plastic
waste 25% Diesel and jet fuel Hydropower
50%
Biomass 25%
81,000 bpd Polymers Nuclear
25% Grey H2
CO2 50%
25% Chemicals Fossil
2020 2050 2021 2050
26,745 TWh 71,135 TWh 94 million tons 528 million tons

Fig. 1 | The anticipated refinery of 2050 compared with the refinery of US Energy Information Administration157. b, Changes in electricity generation
2020. a, The carbon input stream changes from mostly oil in 2020 to CO2 and capacity by source for 2020 and 2050 (projected), based on 2021 data of the
agricultural and municipal waste, including biomass and plastics, in 2050. IEA 158. It is projected that electricity production will almost triple in the next
The ‘Other’ in the 2020 input comprises other liquids, such as fuel ethanol, 30 years from about 26 terawatt-hours (TWh) to about 70 TWh. c, Projection of
hydrocarbon gas liquids and blending components. The output changes from the development of hydrogen generation based on IEA data (2021 data from
mostly transport fuels with a main gasoline component in 2020 to a 50%/50% ref. 159, projected 2050 data from ref. 158). Grey hydrogen, hydrogen produced
mixture of fuels (limited to diesel and jet fuel) and chemicals and materials, from hydrocarbons without carbon capture and use/storage; blue hydrogen,
including performance chemicals and polymers. The average size of the hydrogen produced from hydrocarbons with carbon capture and use/storage;
proposed 2050 refinery is most probably smaller than the 2020 refinery in green hydrogen, hydrogen made with electrolysis using renewable electricity.
terms of carbon molecules processed. The 2020 numbers originate from the

Although historically the share of petroleum-derived chemicals The anticipated changes in the transport sector imply that we only
was roughly 5–10% in volume of all refinery products, their impor- need a relatively small part of the present-day refinery to satisfy the
tance in the product slate is increasing: recent developments suggest future demand for hydrocarbon fuels (Fig. 1a). Although we expect
that the share of chemicals (such as propylene, an important polymer that most of the naphtha (the crude oil fraction processed at present
precursor) may reach 30% and beyond and that the oil-to-chemicals into gasoline and H2) will no longer be needed, there are other valuable
approach could become viable in well-integrated oil refinery com- refinery cuts (that is, crude oil fractions such as asphalt, bitumen, lubes
plexes15. But irrespective of the nature of the refinery output, most and benzene, toluene and xylenes) that are not directly linked to fuel
of the carbon atoms that enter a refinery at present will eventually be production and will need to be obtained from other carbon sources,
emitted as CO2. This is because, even though only a small portion of such as from agricultural and municipal waste. Using these materials
the carbon atoms emitted as CO2 is lost in the refining process, most as alternative refinery input streams may seem straightforward, but
products are either burnt as transport or heating fuels or are inciner- will be challenging when considering that present-day refineries have
ated after being transformed into chemicals or materials and then evolved into highly integrated units that ensure optimal operation of
used. If we could convert some or all of this emitted CO2 back into use- the refinery as a whole, with complex interlocking of processes that
ful products, we could substantially contribute to a carbon-neutral supply and require heat and interlocking of processes that generate
society. For this reason, we analyse in this Perspective the options for and use particular molecules and thereby avoid waste. The result is
using CO2 as the carbon source for transport fuels and considering a highly optimized and very efficient overall processing system, and
plastic waste and biomass (that may already contain desired chemi- future refineries that use different input streams will have to be rede-
cal features) as raw materials for producing polymer monomers and signed to integrate the new conversion processes to ensure that they
chemicals. reach the same degree of optimization while also avoiding unnecessary
waste streams.

Changes in the carbon streams

Although refineries of today convert mostly crude oil, we predict a Current and new refinery schemes
gradual change to the refinery input stream so that only CO2 and agri- At the end of 2018, there were 615 crude oil refineries in the world, with a
cultural and municipal waste (which includes biomass and plastics) combined refining capacity of roughly 92 million barrels per day (bpd)17.
are used in 2050 (Fig. 1a). Refinery output streams are also expected to This equates to an average single refinery, based on crude oil processed,
change substantially in response to transport electrification and use of having a capacity of about 150,000 bpd, or about 20,700 tons per day
fuel cell vehicles, which could largely eliminate the demand for gasoline (based on 7.25 barrels per ton). We will consider several scenarios for
products by as early as 2050. It is expected that a few modes of trans- how such an average-sized single refinery, projected to use no crude oil
port, such as long-distance aviation and heavy and marine transport, while meeting the demands of 2050, could operate in the future. Box 1
will still require high-energy-density hydrocarbon fuels (although these summarizes the assumptions underpinning the 2050 refinery con-
might be gradually replaced by an alternative liquid energy source, such cept, with one key point being that it needs to produce only one-third
as ammonia or methanol). These shifts are expected to limit future of the hydrocarbon fuels that are generated in a refinery at present.
demand for hydrocarbon fuels to about one-third of the transport This amounts to approximately 5,600 tons per day of hydrocarbon
fuels that are produced today16. fuels. Extrapolating current trends in the amount of chemicals versus

296 | Nature | Vol 629 | 9 May 2024

 Box 1

Assumptions for designing hydrocarbon fuel production in the

refinery of the future
• The average refinery size is about 150,000 barrels per day (bpd), of • We need areas of 104 km2 for solar PV panels and 722 km2 for
which 81% results in transport fuels. 328 wind turbines (situated offshore or on land) if solar and wind
• Our future refinery produces only one-third of the hydrocarbon power each contribute 50% to the electricity for H2 generation.
transport fuels required today, or about 40,500 bpd. This equates • We estimate costs of 5–10 billion euros for solar PV and 3–6
to about 5,600 tons per day. billion euros for wind turbines for this scenario, as well as notable
• Only carbon from CO2 goes into the 40,500 bpd of hydrocarbon pressure on resources, including some critical raw materials.
fuels produced, which requires 18 kilotons (kt) (or 4.1 × 108 mol) of • H2 has to be stored for several days following production
CO2 per day. (for example, approximately 10 kt for 4 days).
• The overall conversion of CO2 into hydrocarbons follows
nCO2 + 3nH2 → (-CH2-)n + 2nH2O. This requires 3 mol of H2 per mol of Scenarios
C entering the refinery, which translates into 2.4 kt (or 1.2 × 109 mol) Several scenarios have been calculated to compare the raw materials
H2 per day. use (carbon and hydrogen) for the refinery operations and land area
• H2 will be produced from electrolysis. This requires 5.5 gigawatts for typical refineries. The following existing refineries/complexes
(GW) of electricity, which is to be generated from both wind power were used: Shell Pernis, the largest refinery in the Netherlands, at
and solar photovoltaics (PV) to solve intermittency issues. 404,000 bpd (ref. 17), and Shell’s Pearl GTL (gas-to-liquids) complex
• CO2 capture requires an extra 0.3 GW for capture from a in Qatar, at 140,000 bpd (ref. 37). More details can be found in
concentrated source such as flue gas or 1.5 GW for direct air the Supplementary information (Supplementary Table 1 and
capture. Supplementary Fig. 1).

Carbon in CO2 from process H2 in Refinery output Area for refining Area solar/wind

Shell Pernis
404,000 bpd
×

4.37 × 109 mol C 2.84 × 108 mol C 4.37 × 108 mol grey H 3.30 × 109 mol C fuels
5 km2
crude oil (6.5% of input) from hydrocarbons 7.76 × 108 mol C others

Shell Pearl
140,000 bpd
No data ×

1.52 × 109 mol C 6.06 × 109 mol grey H 1.52 × 109 mol C
3.5 km2
natural gas from hydrocarbons hydrocarbons

Refinery
of the future ×
81,000 bpd

8.2 × 108 mol C

4.08 × 108 mol C fuels
50% CO2 2.46 × 109 mol green H 104 km2 solar
2.04 × 108 mol C chemicals 1 km2
25% plastic waste from electrolysis 722 km2 wind
2.04 × 108 mol C polymers
25% biomass

materials versus fuels produced, we assume that the 2050 refinery for chemicals because we can then make efficient use of the chemical
will produce an equal quantity of chemicals and polymers (that is, functionalities already present in this raw material, and that plastic
2,800 tons each per day of chemicals and polymers). Because we aim waste would be a good carbon source for producing monomers that can
for overall carbon neutrality even though the hydrocarbon fuels gener- be polymerized to generate new materials. Figure 2 sketches the use of
ated in the refinery will be burnt and emit CO2 when used, the carbon CO2, biomass and plastic waste as carbon sources and the conversions
feedstock processed in our future refinery cannot include fossil fuels expected for the refinery of the future.
and should instead comprise a mixture of CO2 and waste. We opt for
CO2 as the carbon source for fuel production because the required
fuels are long-chain linear hydrocarbons that can easily be made from Need for increased carbon capture
CO2 through syngas routes that are established industrial processes. The CO2 required for future fuel production needs to be captured from
Chemical considerations also suggest biomass as a good carbon source either high-concentrated flue gas emitted from blast furnaces or during

Nature | Vol 629 | 9 May 2024 | 297

Perspective

H2O Electrolysis H2 FTS and HC

CxHy
rWGS x ≥ 12

Diesel/
CO2 Electrolysis CO/H2 kerosene
(and lubes)
DAC

Sabatier
CCU
CO2 CH3OH MTH CxHy
x≤3
Gasification

Pyrolysis Functional
CH4 molecules
Biomass Depolymerization Various processes
(HDO, isom, alkyl)

Carbo-
hydrate
platform
Pyrolysis molecules Monomers Polymers

Depolymerization
Plastic Naphtha
waste range
hydro- Coke Asphalt
carbons

Fig. 2 | Overview of the various conversion processes in the refinery of the Functional molecules are aromatics, oxygenates, amines, for example.
future. The starting materials are CO2, biomass and plastic waste. The last two Polymers are not necessarily produced in the refinery but are added in the
are taken as showcase examples of agricultural and municipal waste. The green process flow to clarify the plastic waste-to-polymers concept. alkyl, alkylation;
process line shows the main process route for converting CO2 to hydrocarbon CCU, carbon capture and utilization; isom, isomerization; MTH, methanol-to-
fuels advocated in this Perspective. The dashed lines indicate that biomass and hydrocarbons; FTS, Fischer–Tropsch synthesis; HC, hydrocracking; HDO,
plastic waste conversion may produce CO2 that can be used in the fuels train. hydrodeoxygenation.

cement production or other CO2-intensive activities that may remain

in place in 2050, or be obtained as one of the by-products of the pro- CO2 and CO conversion
cessing of agricultural and municipal waste inside the refinery itself Once captured, we need to convert CO2 into hydrocarbons28–31, which
(a variation on the conventional high-concentrated point sources), or requires about three times as much energy (for H2 generation and CO2
be pulled from the surrounding air by means of direct air capture (DAC). capture) as is generated when using the fuel32. Thermodynamically,
Capture from point sources (post-combustion capture) is much more this approach is not an ideal choice. But as stated above, some liquid
energy-efficient than DAC, simply because much less volume needs hydrocarbon fuels will be needed for long-distance flying, trucking and
to be treated to obtain the same amount of CO2 and extraction from a marine shipping, for which H2, CH3OH and NH3 are not (yet) appropri-
more concentrated source is less demanding. CO2 capture from point ate fuel choices33. For this reason and because CO2 has to be removed
sources is typically stored, not used as we propose here, and is already to reduce its concentration in the atmosphere to climatically less dis-
performed at the multimillion tons per year scale18. ruptive amounts, we are of the opinion that it is better to directly use
CO2 capture uses absorption by liquids such as amines or ionic captured CO2 (carbon capture and utilization (CCU)) than to store it
liquids or by solids such as activated carbon, zeolites, hydrotalcites, in the long run (carbon capture and storage (CCS)), despite the high
metal-organic frameworks or K2CO3 supported on alumina19. Although direct cost involved.
the field is rapidly growing20 and has seen substantial reductions in The hydrocarbons in the C12–C16 range needed as diesel and jet fuel
energy requirements18, opinions diverge about the overall impact of can be produced efficiently using a combination of Fischer–Tropsch
CO2 capture and especially about whether DAC is a viable CO2 miti- synthesis (FTS) and hydrocracking (HC), both established technologies
gation option in view of its energy and materials requirements21–23. that have been proved well beyond the required scale34,35. FTS produces a
Climeworks, one of the companies involved in DAC for storage, is mixture of mostly linear hydrocarbons following the Anderson–Schulz–
operating ORCA, a 4,000 tons per year DAC demonstration project in Flory distribution, and this mixture can then be cracked/oligomerized/
Iceland24 (for context, our refinery would need 18,000 tons per day or alkylated into the required carbon range36. FTS starts from synthesis
6,570,000 tons per year). Carbon Engineering, another company in the gas, a mixture of CO and H2 conventionally made from hydrocarbons37
DAC field, states developing DAC and associated fuels synthesis in the (as in the Shell Pearl complex described in Box 1). We anticipate that
million tons per year range25. Some reviews summarize developments the refinery of the future will generate synthesis gas either through
and challenges associated with DAC technology26,27, and one study has a process known as reverse water gas shift (rWGS) that uses CO2 and
evaluated the factors limiting the rate at which developed technolo- hydrogen produced by electrolysis or by electrocatalytic reduction
gies can be scaled up21. For our refinery, challenges with CO2 capture of CO2 to CO, as depicted in Fig. 3. That this approach is technically
are mitigated because it can be a combination of point source capture viable was illustrated by producing synthetic jet fuel from captured
(from plastic waste and biomass conversion in the refinery) and DAC. CO2 and green H2 using a process combining rWGS, FTS and HC32,38

298 | Nature | Vol 629 | 9 May 2024

 Green energy
Electrocatalysis

FTS HC
3.7 GW
Jet fuel
+ – H2 Diesel

2H2O 2H2 + O2

7.3 GW Synthesis gas

Electrocatalysis
1.5 GW
DAC 5.6 kt per day
2.1 GW CO 40,500 bpd
+ –
CO2
2CO2 2CO + O2

Fig. 3 | A scenario for hydrocarbon fuels production processes in the molecules of H2 per C atom, so only 3.7 GW of power is needed, instead of the
refinery of the future. This scenario for the process uses hydrogen from water 5.5 GW mentioned previously. However, to replace the third H2 molecule
electrolysis and CO from electrocatalytic reduction of CO2 (from DAC) to make mentioned previously, an extra 2.1 GW is needed for the electrocatalytic
synthesis gas (a mixture of CO and H2), which is then converted to the desired reduction of CO2, which is not 100% efficient.
mixture of hydrocarbons. The FTS process starting from CO requires only two

(all established industrial processes), with Shell and KLM announc- requirements of the processes and for generating useful products
ing successful use of 500 l of such fuel for a flight from Amsterdam by replacing this half-reaction with less energy-demanding selective
to Madrid. oxidation reactions, such as oxidation of methanol or glycerol to formic
It should be noted that synthetic jet fuel cannot be produced from acid45,46, oxidation of hydroxymethylfurfural to furandicarboxylic acid
long-chain paraffins such as FTS products alone, but also requires aro- or the conversion of sorbitol to aldoses, ketoses and aldonic acids46.
matics. We can easily produce these either as a by-product of plastics Electrocatalytic CO2RR can produce CO and also more complex
cracking or lignin deconstruction or also on purpose using methanol molecules such as alcohols, aldehydes and hydrocarbons in the C1–
to hydrocarbons or methane to aromatics processes that are available C3 range42, but we are not aware that it has been used to generate
already and are part of our refinery complex design (vide infra). diesel-range hydrocarbons. Although the CO2-to-CO conversion can
The successful demonstration of jet fuel production from CO2 illus- be performed through both electrocatalytic and thermocatalytic
trates that some of the process technology needed for the anticipated routes, the production of long-chain hydrocarbons is thus probably
refinery of the future is already available, but we need to emphasize best done through FTS, which requires synthesis gas (a CO and H2 gas
that all conversion processes need to be developed with the whole mixture). The CO component is accessible not only through rWGS
refinery concept in mind to ensure maximum carbon-atom and energy and CO2RR but also through methods such as dry methane reform-
efficiency. That is, individual processes should not be developed and ing47 (that is, CO2 + CH4 → 2H2 + 2CO) or CO2 methanation using the
optimized in isolation; instead, process intensification and process inte- Sabatier reaction48 followed by partial oxidation of CH4 (that is,
gration will be paramount, so that no carbon and no energy is wasted. CO2 + 4H2 → CH4 + 2H2O followed by 2CH4 + O2 → 2CO + 4H2). These
In that regard, present-day refineries offer a very high benchmark in chemical routes typically co-produce the H2 needed for FTS. If the CO:H2
overall system optimization. However, this requirement should not ratio needs to be adjusted or when using CO from CO2RR, the required
prevent the development of promising new technologies that do not extra H2 could be generated not only through H2O electrolysis but also
yet offer the desired efficiency or compatibility with existing refinery CH4 pyrolysis49 (that is, CH4 → C + 2H2) that retains the carbon as a solid,
processes. that is, the process avoids CO2 emissions and could therefore serve as
Among the conversion processes we expect for the refinery of the a possible intermediate solution for large-scale H2 production while
future, sketched in Fig. 2, one of the first chemical reactions is the rWGS transitioning from fossil to renewable energies. As a further benefit,
(CO2 + H2 → CO + H2O) that converts CO2 into CO (refs. 39,40) for syngas the required CH4 is not only available from natural gas but can also be
production. This conversion requires a third of the total amount of H2 generated through anaerobic digestion of agricultural and municipal
needed for the overall synthesis of hydrocarbons and, for one refinery, waste and then integrated into chemical H2 production schemes.
the production of this H2 by electrolysis would use about 1.8 GW of Further flexibility for future fuel production could arise from
electrical power. This thermochemical CO production route might, in using methanol (CH3OH) as the starting material for the methanol-
the future, be replaced by direct electrochemical reduction of CO2 to to-hydrocarbons (MTH)50 process, which is very versatile and can
CO (refs. 41–44) (as pictured in Fig. 3) that, if carried out with the same generate paraffins, olefins and aromatics51–53. The methanol could be
efficiency as water electrolysis to produce H2, would consume roughly generated through the conventional methanol synthesis route that
2.1 GW (Fig. 3). CO production directly from CO2 would thus use more uses synthesis gas or possibly through electrocatalytic CO2RR. As well
energy than needed to generate the H2 for CO production through as conversion into hydrocarbon fuel, methanol itself could be used as
the rWGS process (2.1 versus 1.8 GW). However, the thermocatalytic a fuel in fuel cells or even as an additive to gasoline for internal com-
process also requires energy to run the endothermic reaction and to bustion engines.
heat and compress the process gases, which would probably result in Industry has developed gas-to-liquids (GTL) technology that uses
comparable overall energy requirements. It is also worth consider- chemical routes we have just discussed for converting natural gas
ing that the electrolysis of water to produce H2 and the CO2RR share through synthesis gas and H2 into a wide range of liquid products that
the oxygen-producing half-reaction. Because this generated oxygen would otherwise be made from oil. GTL generates lower CO2 emis-
is not needed, there might be scope for lowering the overall energy sions but at present makes only a very minor contribution to global

Nature | Vol 629 | 9 May 2024 | 299

Perspective
liquid hydrocarbon production. Increasing this contribution offers model compounds will often not capture the key challenges of working
an important intermediate step towards more sustainable refineries. with real-world lignins. However, integrated routes for lignin valorization
Ultimately, however, our refinery of the future will generate liquid are emerging and slowly transitioning lignin from being regarded as a
hydrocarbon fuels through routes that only use renewable energy, low-grade fuel to an interesting raw material for various applications59,60.
CO2 and water. This so-called power-to-liquids concept54–57 can tap into
existing industrial processes developed for GTL and needs to combine
these with efficient CO2 conversion strategies to develop technology Plastic-waste-to-polymers conversion
options for making carbon-neutral hydrocarbon fuels. About 368 million tons of polymers are produced each year (ref. 81), of
which almost 80% ends up as plastic waste82. Polymers are produced
by combining monomers that are the direct product of refinery and
Biomass-to-chemicals conversion petrochemical processes. The combination or polymerization of mono-
The carbon coupling processes for converting CO2 and CO described mers is itself not necessarily a refinery process but can be performed
above mainly produce linear hydrocarbons, so other feedstocks are in petrochemical plants associated with refineries.
needed to make products such as bitumen, asphalt and lubes that are The most efficient way to close the polymer–carbon cycle is to recycle
derived from crude oil at present. Plant-derived biomass is an obvious polymer-containing waste using mechanical or (our preferred option)
choice and available in the form of oils and fats (for example, triglyc- chemical processes. In chemical recycling, the waste material is depo-
erides), sugar-containing biopolymers (for example, starch, hemicel- lymerized into its constituent monomers, which can then be repolymer-
lulose and cellulose) and aromatics-containing biopolymers (that is, ized to create a fully circular polymer technology. Concern over plastic
lignin), with each of these three classes requiring specific processes pollution has highlighted the need for recycling, with chemical recycling
for valorization58–60. receiving increasing attention83–86 and some large plastic producers and
The thermal and catalytic conversion of agricultural waste into fuels chemical manufacturers announcing recycling pilot projects87,88. One
and chemicals has been studied extensively61–64. Biomass has been of the challenges in the field is, however, that the percentage of plastics
added as a co-feedstock in conventional refinery processes65–67, some- recycled is still very low in many countries89–91. But recycling and dis-
times together with plastic waste68, and biomass was the basis of the posal schemes are emerging and can be compared using an index that
first generations of biodiesel (for example, the fatty acid methyl esters considers economic, environmental and energy-related parameters92
(or FAME) biodiesel and deoxygenated fatty acids). (with mechanical recycling, in the form of incorporation of plastics into
Oils and fats contain the long linear hydrocarbon chains desired construction materials, scoring highly among available strategies).
for diesel fuels and these can be retrieved by removing the glycerol Polymers differ with respect to their ease of recycling. For example,
and carboxylic acid groups using hydrodeoxygenation (HDO). HDO polyolefins, such as polyethylene and polypropylene, constitute around
of renewable fats and oils followed by isomerization is the basis for 55% of the global production volume of polymers and are challenging
the NExBTL biodiesel production processes developed by the oil and to depolymerize to their monomers (ethylene and propylene). They
refining company Neste and partners. One of their plants in Rotterdam, are most often subjected to a combination of pyrolysis and cracking93,94
the Netherlands, produces approximately 800,000 tons of diesel-type that produces a mixture of naphtha-range molecules, considerable
molecules per year, or about 2.2 kilotons (kt) per day, which matches amounts of aromatics and heavy polyaromatic molecules, that is, a
closely in scale, but not in application, with our target of 2.8 kt per product that is more suited for producing fuels than monomers95.
day of chemicals from biomass in the anticipated refinery concept69. Moreover, recycling that combines pyrolysis and cracking is not 100%
Biomass in the form of sugar-containing biopolymers such as starch carbon-atom-efficient96. By contrast, chemical recycling of polymers
or lignocellulose is typically deconstructed using processes such as such as polyethylene terephthalate, polystyrene and polymethylmeth-
pyrolysis or solvolysis70. Processing sometimes takes into account the acrylate through depolymerization to their monomers seems techni-
range of chemical functionalities present in the feedstock71–74, with a cally possible, although—again—is not 100% carbon-atom-efficient97.
combination of methods such as aqueous-phase reforming, hydro- Solvolysis as the most efficient recycling route for polyethylene tereph-
genation/hydrogenolysis and dehydration used to generate the desired thalate, for example, still generates more than 30% of the CO2 emissions
products74. That said, most of the processes now under development associated with incineration89. This highlights the need for considerable
simply target biomass conversion to hydrocarbon fuels75, even though process developments to limit by-product and CO2 production during
a wide range of platform molecules could, in principle, be made, such the plastic-waste-to-polymers conversion processes to be installed in
as succinic acid, furanics, hydroxypropionic acid, glycerol, sorbitol, the refinery of the future.
xylitol, levulinic acid, isoprene, lactic acid and ethanol (these have
been identified76 as the top 10 biomass-derived platform molecules).
Ethanol, for instance, can serve as a gasoline additive or be converted Biomass and plastic waste conversion
into diesel and jet fuel77. The (partial) removal of oxygen implies that The anticipated 2050 refinery uses 18 kt of CO2 (4.1 × 108 mol C) to
CO2 is a by-product of biomass conversion to chemicals78, which will produce hydrocarbon fuels and another 4.1 × 108 mol C derived from
lower the carbon atom efficiency and thus the maximum yield of these biomass and plastic waste to produce chemicals and materials such as
processes. Economic viability is also affected because H2 is often needed polymers. We therefore have to consider the effectiveness of waste and
to perform the conversion reactions79,80. biomass conversion when designing an integrated refinery process.
Lignin, the third type of biomass, was long treated as a low-value If we assume that the conversions are initially 50% effective for exam-
by-product of biorefineries and the pulp and paper industry and used ple, then we would obtain not only the 4.1 × 108 mol C required for the
as low-grade fuel. It can be converted to valuable aromatics, however, chemicals and polymers production chain but also 4.1 × 108 mol C in
and its lower oxygen and higher aromatics content compared with cel- the form of by-products that we will assume to be (or converted to) CO2.
lulose and hemicellulose make it an interesting starting material for the The 50% effective biomass and plastic waste conversion thus produces
production of chemicals. But lignin is structurally complex and difficult exactly the amount of CO2 needed for the production of hydrocarbon
to process: although the bonds between the building blocks of the cel- fuels, in a form and location in which it can be captured efficiently from
luloses are fairly uniform, those in lignin vary considerably and make high-concentration-process flue gas inside the refinery itself (that is,
it challenging to develop efficient depolymerization strategies. The the CCU mode of operation).
structural and bonding complexity of lignin also mean that exploratory However, we expect that the conversion of agricultural and munici-
research on model compounds can be challenging to extrapolate, as pal waste would reach higher efficiencies in the future so that the CO2

300 | Nature | Vol 629 | 9 May 2024

 Feedstock Fuels Products
CO2
Fuels
CO2 C
12 kt per day 5.6 kt per day
2.72 × 108 mol 4.08 × 108 mol

CO2
1.36 × 108 mol
Biomass
C
8.2 kt per day 2.72 × 108 mol
Chemicals and materials
Chemicals
2.8 kt per day
C C
4.08 × 108 mol 4.08 × 108 mol Materials
C 2.8 kt per day
Plastic waste
2.72 × 108 mol
3.8 kt per day

Fig. 4 | Integration of the carbon streams in the refinery of the future. The and plastic waste conversion, or a lower contribution of biomass and plastic
conversion of biomass and plastic waste is not 100% effective. At an assumed waste conversion to the system, will improve these numbers. No carbon is lost
75% efficiency for both, the conversion of biomass and plastic waste would in this scheme. The differences between input tons and output tons are caused
produce a carbon stream (presumed to be CO2) equivalent to one-third by the different molecular weights on the input molecules: CO2 has a weight
of the carbon required for the fuels train, next to the 4.1 × 108 mol C required of 44 g mol−1, biomass is assumed to be (CH2O)n, at 30 g mol−1 C, and polymers,
for the chemicals and polymers train. As well as this carbon, the refinery as well as all the products, are assumed to be (CH2)n, at 14 g mol−1 C. At 100%
would remove 12 kt per day of CO2 from the atmosphere by means of DAC. efficiency for the conversion of biomass and plastic waste, the process would
Combustion of the hydrocarbon fuels would generate 18 kt of CO2, so this use 18 kt of CO2 (4.08 × 108 mol C), 6.2 kt of biomass (2.05 × 108 mol C) and
particular scenario is not yet carbon-neutral (although it would lower overall 2.8 kt of polymers (2.05 × 108 mol C).
CO2 emissions from fuels by two-thirds). Increased efficiency in the biomass

input stream can be supplemented with DAC. Although DAC will require energy—reactors will have to be heated to process temperature, gases
extra electricity, it moves operation of the overall refinery towards will have to be compressed to increased pressures and the thermo-
being carbon-neutral or carbon-negative (illustrated in Fig. 4 for an dynamic heat of reaction will have to be provided for endothermic
anticipated biomass and plastic waste conversion efficiency of 75%). processes (such as rWGS or biomass/plastic pyrolysis). This energy
If the efficiency for biomass and plastic waste conversion is lower will have to be generated from electricity or by heat integration of
than 50%, more CO2 is produced in the refinery than the hydrocarbon the refinery processes (for instance, FTS and methanol synthesis are
fuels chain can handle and the excess CO2 would have to be captured strongly exothermic). We have carried out thermodynamic heat of
and stored (that is, CCS mode of operation). Coupling the CCU and CCS reaction calculations (provided in the Supplementary information,
modes would be a way to create a carbon-negative refinery scenario. Supplementary Table 2) that indicate that the process energy will be
about 10–15% of the energy required for H2 synthesis and CO2 capture.
But ultimately, the refinery’s energy requirements will depend on the
Process integration and energy needs refinery concept, which should target an integrated design allowing for
The various processes that convert feedstocks and generate fuels, the most efficient flow of energy and molecules through the anticipated
chemicals and materials will need to be integrated into the refinery of refinery, rather than focus on individual process solutions.
the future. They can be described as follows (with processes posing
the main technical challenges in bold):
• Hydrocarbon fuels production: Process developments and scale-up
Hydrogen production We list in Table 1 the main processes required for our refinery of the
Carbon capture → rWGS and/or CO2RR → FTS → HC future. Some of the conversion processes are already available at the
• Biomass conversion: commercial scale (for example, FTS, HC, rWGS, methanol synthesis,
Biomass → Depolymerization → Raw material pool (and excess CO2 MTH and olefin polymerization). All of these processes have been
to CCU or CCS) demonstrated at >1 ton per day capacity, but may need further devel-
• Plastic waste conversion: opment. For example, impurities present in CO2 gas streams and in
Plastic waste → Depolymerization → Raw materials pool (and excess the biomass and plastic waste may deactivate existing catalysts, so
CO2 to CCU or CCS) feedstock pretreatment or the development of more resilient catalysts
• Chemicals and polymers production: may be required to handle the new range of feedstocks. We consider
Raw materials pool → value-added products, such as chemicals and such process development to be relatively straightforward compared
polymers with the substantial scale-up efforts required for some of the other
• Crossovers processes listed in Table 1.
Some products from biomass and plastic waste conversion can be Ensuring that electrolysis and CO2 DAC technologies are available at
used as fuels the scale at which the refinery of the future would need will be particu-
Some routes resulting from CO2-derived synthesis gas can produce larly challenging. However, 2050 is still many years ahead and expo-
platform materials en route to functional chemicals nential technological development, for which capacity doubles every
2 years as with Moore’s law for computing, is not uncommon98–100. If
A large part of the future refinery’s energy demand is for H2 synthe- this holds true, a scale-up factor of 32,768 is possible in 30 years and the
sis and CO2 capture, but running the refinery processes also requires required scale of DAC capacity could be reached in roughly 22 years.

Nature | Vol 629 | 9 May 2024 | 301

Perspective
Table 1 | Scale-up required for the main conversion processes

Process Reactant(s) Product Present scale Desired Estimated scale-up

scale required
DAC of CO2 CO2 in air (400 ppm) Concentrated CO2 11 tons per day (ref. 24) 18 kt per day 1,636×
CO2 capture at stationary CO2 in flue gas Concentrated CO2 5,500 tons per day (ref. 18) 18 kt per day 3.3×
sources
Electrolysis Water Hydrogen 8 tons per day (ref. 153) 2.4 kt per day 300×
Electrocatalysis CO2 CO 25 tons per day in 2030 (ref. 154) 18 kt per day 720×
FTS CO/H2 Long-chain hydrocarbons 20,000 tons per daya 5.6 kt per day Industrially available
MTH CH3OH Olefins, paraffins, aromatics 2,300 tons per dayb 2.8 kt per day Industrially available
HC Hydrocarbons Hydrocarbons of desired range 16,000 tons per dayc 5.6 kt per day Industrially available
rWGS CO2 CO 18 kt per day Industrially available
Depolymerization Plastic waste Monomers 500 tons per day (ref. 82) 2.8 kt per day 5.6×
Polymerization Monomers Polymeric hydrocarbons 2.8 kt per day Industrially available
Depolymerization Biomass, including Platform molecules 400 tons per day (ref. 155) 6.2 kt per day 15.5×
agricultural waste
Synthesis and conversion Biomass, including Platform molecules 280 tons per day (ref. 156) 2.8 kt per day 10×
agricultural waste
Present scale is based on existing technology. Data are for one refinery. Commercial processes in operation at present refineries/complexes: aShell, Ras Laffan, Qatar. bJiangsu Sailboat
Petrochemical Company, Lianyungang, China. cMarathon Petroleum Corporation, Garyville, LA, USA.

This assumes, of course, that all necessary materials and financing the fuel production rate we are targeting, we thus need to produce
are available. (store/transport) 2.4 kt (or 1.2 × 109 mol) of H2 per day. The production
The scale-up factor for DAC of 1,636 may seem daunting, particularly of H2 requires 50–58 kWh per kg H2 (ref. 107) or 1 MW for roughly every
when considering that it accounts for only the CO2 need of one of 615 18 kg H2 produced per hour, which equates to 5.5 GW of electrical power
refineries. But the factor is based on Climeworks’ ORCA unit that was needed to produce 2.4 kt H2 per day. This electricity must be produced
built in 2021 and is capturing 4,000 tons per annum (or 11 tons per day) fossil-free and thus primarily from wind power or solar PV. Consider-
of CO2 (ref. 24). Climeworks recently announced a breakthrough on pro- ing day/night and seasonal cycles for a northwest European location
ject Mammoth with a CO2 capture capacity of 36,000 tons per annum that is at a higher latitude compared with the USA, we assume for PV a
or close to 100 tons per day, making it nine times larger than ORCA101 capacity factor of 12% (on the low side of the 10–21% window reported
and reducing the scale-up challenge to a factor of about 180 in just by the International Energy Agency (IEA)108 and in line with PV potential
2 years. The Haru Oni project in Chile, co-founded by Siemens Energy, studies for countries in Europe109–111). An evaluation of present-day wind
ExxonMobil and others, will produce methanol from captured CO2 and turbine technology in the UK109 and the USA112 has indicated capacity
green hydrogen. It aims to reach by as early as 2027 a production capac- factors between 30% and 35%, although newer turbines claim 60%
ity that is one-fifth the capacity of our refinery (8.6 × 107 mol C per day), capacity factors113. Because of the difference in intermittency of wind
using DAC, power from wind turbines and methanol-to-gasoline tech- and solar energy, we choose to split the energy source used to meet
nology102,103. These developments make us optimistic that, in principle the refinery requirement 50/50 between wind and solar, so we need
and with sufficient support, the technologies needed for the refinery 2.75 GW from each.
of the future can be developed to the desired scale before 2050. This implies that we need 104 million m2 of solar panels (delivering
220 W m−2 peak power and with 12% capacity factor) and about 328
windmills (assuming that the newest wind turbines produce 14 MW
Land and offshore area requirements each at 60% capacity factor) to generate the overall power of 5.5 GW
The main power consumption of the proposed refinery is for water required for making hydrogen; that is, alongside our refinery itself
electrolysis to produce H2. H2 is needed for making synthesis gas used occupying an area of perhaps 1 km by 1 km, we need a solar panel park
in FTS to produce hydrocarbon fuels and in more limited amounts in of 10.2 km by 10.2 km and a windmill park of 28 km by 28 km (which, if
other processes such as HC that establishes the correct carbon length on shore, could also readily house the required DAC units). An artist’s
distribution in the hydrocarbon fuels. We are considering H2 made ‘on impression of the whole refinery complex is depicted in Supplemen-
purpose’ for the anticipated 2050 refinery. An abundance of H2 may also tary Fig. 1. These numbers do not take into account the energy needed
become available as a means of intermediate storage of energy104,105 for DAC (which could reach 1.5 GW if all carbon used in our refinery
from inherently intermittent sustainable power generation technolo- were obtained by means of DAC using current technology) nor the
gies (solar PV and wind power), to address Richard Smalley’s ‘terawatt energy needed to run the processes in the refinery. To put the required
challenge’ that calls for a transition from the approximately 14 TW infrastructure into perspective, the 5.5 GW needed to produce H2 for
power generated in 2003 from fossil fuels to the 30–60 TW power one of our anticipated refineries of the future is comparable with the
that will be needed in 2050 and should be generated using renewa- European Green Deal target of “installing at least 6 GW of renewable H2
bles5,6 (Fig. 1b). Although H2O electrolysis is an established technol- electrolysers in the EU by 2024”114 for primary energy provision, and the
ogy, the scale-up from present-day 10–100 MW units to the future 475 GW needed just to produce H2 for 86 western European refineries
multiple-GW-scale units is still a great challenge, even when taking with a current capacity of around 13.6 million bpd17 is about 12 times
expected cost reductions into account106. more than the European Green Deal target of “40 GW of renewable H2
The H2 and energy demand of the refinery concept can be appreci- electrolysers by 2030”114.
ated when considering that the overall reaction for hydrocarbon fuels CO2 DAC using present technology requires between 0.32 and
production is essentially nCO2 + 3nH2 → (-CH2-)n + 2nH2O, meaning that 4.73 MWh per ton of CO2 (ref. 115). The energy needed to capture CO2
at least three molecules of H2 are required per molecule of CO2. Given from concentrated stationary sources compares at the very low end

302 | Nature | Vol 629 | 9 May 2024

of that range115, whereas the Climeworks technology needs a medium In terms of its input stream, assuming 100% conversion efficiency,
energy input of 2 MWh per ton CO2 (ref. 116). This suggests that DAC the anticipated refinery of the future will use 6,200 tons of agricultural
requires about five times more energy than capture from concen- waste and sustainable-resourced and certified biomass and 2,800 tons
trated stationary sources and that capturing all the CO2 needed in of plastic waste every day. Globally, for 615 refineries, these numbers
one future refinery with present DAC technology would use 1.5 GW translate to 1.4 Gt of biomass and 628 Mt of plastic waste that are
of electrical power. The present Climeworks DAC collector units can needed per year. A report by Imperial College London Consultants133
capture approximately 230 kg per day per unit101,117. To capture the suggests that the available sustainable-resourced biomass in Europe
required 18 kt per day, we would need 79,000 units. If clustered in alone (which would house 86 of the 615 refineries) would be between
blocks of 24 (a stack of four containers with six units each), we would 1 and 1.3 Gt in 2050, of which 539–915 Mt would be available for energy
need 3,285 blocks, or five rows of 8 km each, presumably spaced apart applications. So there should, in principle, be sufficient biomass for the
by at least 500 m. proposed refinery scenario. Polymer production is projected to triple
in 2050 compared with 2014 (ref. 134), reaching a level of 950 Mt. To
meet refinery needs, we would need to recycle two-thirds of that pro-
Cost requirements duction, which is still a challenge given the efficiencies of the current
We calculated installation costs for the electricity generation and elec- chemical recycling technologies.
trolysers to produce the hydrogen required. Given the considerable
disagreement in the literature about cost estimates for PV and wind
power7,118–122, we can only provide a rough estimate for these costs in Challenges towards a carbon-neutral future
2050. With available numbers (see Supplementary Information, Sup- A carbon-neutral future with CO2 emissions balanced by CO2 uptake is
plementary Table 4) and assuming that prices will drop as technologies an ambitious goal that requires not only technological developments
develop and are scaled up, we estimate a range of 5–10 billion euros for affecting sectors such as the chemical industry and transport but also
the manufacturing of the required solar cells, 3–6 billion euros for the targeted and sustained policies and public support.
manufacturing of the required wind turbines and about 1.0–2.5 billion
euros for the electrolysers123,124, or 9–18.5 billion euros in total. As well Technical challenges
as this, we expect the investment costs for the refinery installations The steps that can be taken to decarbonize electricity production and
to be on the order of 5 billion euros (see Supplementary information). a large part of the transport sector are relatively clear. But realizing
These costs do not cover the infrastructure for transporting or storing the proposed 2050 refinery concept will be a more demanding task,
energy or H2 or maintenance of the solar panels and wind turbines. given the complexity of the feedstocks, processes and products that
We assume that the refinery would mainly need to store H2 (used at a are involved and considering the necessary process and technology
rate of 2.4 kt per day). If we assume that we need enough H2 to cover innovations and scale-up of existing technology that are required. Key
several days (say, 4 days), we will need a storage facility for about 10 kt technical challenges need to deliver:
of H2. This is similar in scale to the ConocoPhillips Clemens Terminal • Technologies for CO2 DAC are required on a scale of approximately
storage facility in Texas, USA used for underground H2 storage in salt 20 kt per day for a single refinery.
caverns for more than three decades. Investments for these types of • Electrocatalytic CO2 activation and conversion processes need to
storage facilities seem to be two orders of magnitude lower than the be efficient, selective, use earth-abundant elements as catalysts and
costs required for the manufacturing of solar cells and wind turbines125. ultimately produce molecules more complex than CO (at a scale of
However, the required geological features may not be available at all approximately 20 kt per day for a single refinery).
sites at which future refineries might be constructed. • Large-scale reactors need to be electrified to lower the CO2 footprint
There are of course alternatives to solar and wind power for gen- of endothermic production processes135–137; such process electrifica-
erating the electricity needed in H2 production, with nuclear power a tion might be compatible with existing refineries and could serve as
natural choice. France is producing about 70% of its power from nuclear an intermediate solution for lowering refinery CO2 emission.
reactors that typically have capacities between 0.9 and 1.3 GW (ref. 126). • Electrocatalytic CO2 conversion and H2 production are both accom-
For one refinery, we would thus need about five of these to produce panied by O2 formation at the anode, which should be replaced by
the required 5.5 GW of electricity, at an estimated cost of 20–40 billion reactions (such as oxidation of molecules derived from agricultural
euros (refs. 127,128). and municipal waste) that generate useful products138.
• Carbon-efficient processes are needed for making chemicals from
biomass and agricultural waste and for recycling of plastic and munici-
Resources pal waste; this will require new catalytic technologies that should be
As well as the direct costs, the refinery of the future also has notable based on earth-abundant elements.
resource implications, illustrated by an EU Joint Research Centre evalu- • A considerable infrastructure needs to be developed to ensure suf-
ation of the raw materials requirements for solar PV and wind power ficient solar PV and wind power generation capacity (about 6–7 GW
technologies129. Wind turbines require large amounts of relatively per refinery).
abundant concrete and steel and substantial amounts of critical raw • Effective and reliable energy storage techniques are needed to store
materials130, such as aluminium, chromium, manganese, nickel, zinc and excess energy during peak solar PV and wind power production.
the rare earths dysprosium, neodymium, praseodymium and terbium. • Green H2 generation capacity needs a substantial scale-up to meet
Solar PV cell construction uses concrete, steel and glass and critical future refinery demand and be based on electrolyser technology
elements such as aluminium, indium, gallium and germanium. We that uses stable electrodes composed of earth-abundant elements;
should also consider that developing the electrolysers needed for our in the interim, H2 can be generated through processes such as biom-
refinery may encounter materials availability issues, as the supply of ethane pyrolysis139–141 that yields solid carbon and thereby avoids
the required iridium, scandium, yttrium and platinum may become CO2 emissions.
critical131. The quest to replace fossil fuels may thus trigger undesired
challenges in other critical raw materials cycles132, even for metals as The sheer scale of the transformation needed to realize a carbon-
abundant as iron, copper, nickel and aluminium. More background neutral chemical industry, in terms of the technology and process
information can be found in the Supplementary information (Sup- developments and associated material demands, is staggering. After
plementary Tables 5 and 6). all, the chemical and petrochemical industry transforms 1.6 billion tons

Nature | Vol 629 | 9 May 2024 | 303

Perspective
of raw materials every year (677 Mt of fossil feedstock, 274 Mt of water affect workers and consumers. Although our focus has been on sci-
and 686 Mt of secondary reactants) into 821 Mt of primary products ence and technology, public and political engagement are clearly as
and 815 Mt of secondary products (such as CO2, CH4 and H2O)142. On top important in enabling the transitions that are necessary if we are to
of this, complexity needs to be considered as the industry produces limit the emission-induced warming of Earth’s climate and minimize
hundreds of different products that are used in all parts of modern the adverse physical manifestations and socioeconomic repercussions
society, supplying, for instance, plastics, fibres and fertilizers that of climate change.
are indispensable for our economy. These products are the result of
highly complex, interdependent process flows and we cannot simply 1. National Oceanic and Atmospheric Administration (NOAA). Broken record: atmospheric
drop in a new conversion process but need instead a systems approach carbon dioxide levels jump again. NOAA https://www.noaa.gov/news-release/broken-
to devising solutions137,142–144. As an intermediate solution, renewable record-atmospheric-carbon-dioxide-levels-jump-again (2023).
2. United Nations. Paris Agreement to the United Nations Framework Convention on Climate
electricity could be used to replace combustion-based electricity and, Change (United Nations, 2015).
in some cases, provide process heating, thereby reducing emissions 3. United Nations. Kyoto Protocol to the United Nations Framework Convention on Climate
Change (United Nations, 1997).
of current refineries. This concept has been described in general for
4. European Commission. Proposal for a Regulation of the European Parliament and of the
several endothermic reactions145, such as CO2 reforming of methane Council Establishing the Framework for Achieving Climate Neutrality and Amending
or rWGS146, methane steam reforming147,148 and propane dehydrogena- Regulation (EU) 2018/1999 (European Climate Law) (European Commission, 2020).
5. Smalley, R. E. Future global energy prosperity: the terawatt challenge. MRS Bull. 30,
tion149. A large-scale demonstration plant for electrically heated steam
412–417 (2005).
cracking furnaces is under construction at BASF’s site in Ludwigshafen, This study and the next one discuss the magnitude of the challenges with respect to
Germany150 and another at Coolbrook’s Brightlands Chemelot Cam- the energy transition.
6. Kurtz, S. R. et al. Revisiting the terawatt challenge. MRS Bull. 45, 159–164 (2020).
pus in the Netherlands151. Of course, local circumstances will dictate
7. Breyer, C. et al. On the history and future of 100% renewable energy systems research.
whether green electrons are best used for heating or for electrochemi- IEEE Access 10, 78176–78218 (2022).
cal processes. 8. Yergin, D. The Prize: The Epic Quest for Oil, Money, and Power (Simon & Schuster, 1992).
9. Hagen, J. Industrial Catalysis: A Practical Approach 3rd edn (Wiley, 2015).
10. Rase, H. F. Handbook of Commercial Catalysts: Heterogeneous Catalysts (CRC, 2000).
Economic challenges 11. van Santen, R. A. Modern Heterogeneous Catalysis: An Introduction (Wiley, 2017).
We argue that it should be technically possible to build in 2050 12. Vogt, E. T. C. & Weckhuysen, B. M. Fluid catalytic cracking: recent developments on the
grand old lady of zeolite catalysis. Chem. Soc. Rev. 44, 7342–7370 (2015).
a carbon-neutral refinery that converts CO2 and municipal and agricul-
This paper summarizes some historical developments of the fluid catalytic cracking
tural waste into hydrocarbon fuels, chemicals and the many man-made process, including the co-processing of biomass.
materials that are indispensable to modern life. However, each of the 13. Vogt, E. T. C., Whiting, G. T., Chowdhury, A. D. & Weckhuysen, B. M. Zeolites and zeotypes
for oil and gas conversion. Adv. Catal. 58, 143–314 (2015).
new refineries requires investments on the order of 14–23 billion euros, 14. Reliance Industries Limited. Petroleum Refining & Marketing. Reliance Industries Limited
with 65–80% of this sum covering costs directly related to hydrogen https://www.ril.com/ourbusinesses/petroleumrefiningandmarketing.aspx (2021).
generation. Just replacing all 615 refineries now operating over the next 15. Alabdullah, M. A. et al. A viewpoint on the refinery of the future: catalyst and process
challenges. ACS Catal. 10, 8131–8140 (2020).
25+ years would require investments on the order of 320–520 billion This study provides an excellent perspective on the crude-to-chemicals approaches
euros per year for each year until 2050. To put this into perspective, available and to be developed in the next decade.
our new refineries would use a total of 4.3 TW of electrical power in 16. British Petroleum (BP). Energy Outlook: 2020 Edition (BP, 2020).
17. 2018 Worldwide Refining Survey. Oil Gas J. (2018).
2050. The terawatt challenge5,6 assumes that we will need 30–60 TW of 18. Koytsoumpa, E. I., Bergins, C. & Kakaras, E. The CO2 economy: review of CO2 capture and
non-fossil power in 2050 to supply the world’s energy needs, or roughly reuse technologies. J. Supercrit. Fluids 132, 3–16 (2018).
ten times the requirements for refinery replacement. 19. Osman, A. I., Hefny, M., Abdel Maksoud, M. I. A., Elgarahy, A. M. & Rooney, D. W. Recent
advances in carbon capture storage and utilisation technologies: a review. Environ.
As well as the substantial financial investments, large land or off- Chem. Lett. 19, 797–849 (2021).
shore areas will be needed to site the solar panels and/or wind turbines 20. Leung, D. Y. C., Caramanna, G. & Maroto-Valer, M. M. An overview of current status of
required to supply the refineries with electrical energy. In fact, even carbon dioxide capture and storage technologies. Renew. Sustain. Energy Rev. 39,
426–443 (2014).
nuclear energy may have to be used to meet the immense electric- 21. Realmonte, G. et al. An inter-model assessment of the role of direct air capture in deep
ity demand of the future refineries. Construction of the solar cells mitigation pathways. Nat. Commun. 10, 3277 (2019).
and wind turbines will also require large amounts of materials and 22. Chatterjee, S. & Huang, K. W. Unrealistic energy and materials requirement for direct air
capture in deep mitigation pathways. Nat. Commun. 11, 3287 (2020).
metals, many of which are critical minerals. Solving the fossil fuel 23. Realmonte, G. et al. Reply to “High energy and materials requirement for direct air
challenge may therefore trigger other undesired cascades of raw capture calls for further analysis and R&D”. Nat. Commun. 11, 3286 (2020).
materials issues. 24. Climeworks. Climeworks, ON Power and Carbfix partner together. Climeworks https://
www.climeworks.com/news/climeworks-has-signed-groundbreaking-agreements-with
(2020).
Long-term societal commitment 25. Carbon Engineering. Our Technology. Carbon Engineering https://carbonengineering.
To deliver the large-scale science and technology developments com/our-technology/ (2024).
26. Madhu, K., Pauliuk, S., Dhathri, S. & Creutzig, F. Understanding environmental trade-offs
required for a refinery of the future, it will be essential to have in place and resource demand of direct air capture technologies through comparative life-cycle
adequate policies and a long-term support and financing framework assessment. Nat. Energy 6, 1035–1044 (2021).
that fosters discovery and innovation and enables the necessary dem- 27. Siegelman, R. L., Kim, E. J. & Long, J. R. Porous materials for carbon dioxide separations.
Nat. Mater. 20, 1060–1072 (2021).
onstration projects. Support should also include economic drivers This article summarizes the latest developments and applications of CO2-capturing
that encourage a move to carbon neutrality, such as a form of carbon porous materials.
tax, and a removal of barriers that prevent the adoption of low-carbon 28. Zhu, P. & Wang, H. High-purity and high-concentration liquid fuels through CO2
electroreduction. Nat. Catal. 4, 943–951 (2021).
technologies (such as infrastructure or subsidies on fossil fuels)104,152. 29. Sullivan, I. et al. Coupling electrochemical CO2 conversion with CO2 capture. Nat. Catal.
We have mentioned our optimism that science and technology can 4, 952–958 (2021).
30. Marques Mota, F. & Kim, D. H. From CO2 methanation to ambitious long-chain
ultimately deliver the solutions for moving society towards carbon
hydrocarbons: alternative fuels paving the path to sustainability. Chem. Soc. Rev. 48,
neutrality but recognize that the magnitude of the transition that is 205–259 (2019).
needed, in terms of investments, critical raw materials and required 31. Artz, J. et al. Sustainable conversion of carbon dioxide: an integrated review of catalysis
and life cycle assessment. Chem. Rev. 118, 434–504 (2018).
areas for electricity generation, is unprecedented. After all, we need
32. Kraan, O., Kramer, G. J., Haigh, M. & Laurens, C. An energy transition that relies only on
to replace infrastructure and processes built and improved over more technology leads to a bet on solar fuels. Joule 3, 2286–2290 (2019).
than a century and that today turn cheap fossil fuels into the large num- This paper provides a good primer on the challenges of the energy transition.
33. Pivovar, B. Catalysts for fuel cell transportation and hydrogen related uses. Nat. Catal. 2,
ber of products that satisfy the needs of modern society. The changes
562–565 (2019).
required of the chemical industry, and of other industry sectors, can 34. Maitlis, M. & de Klerk, A. Greener Fischer-Tropsch Processes for Fuels and Feedstocks
thus be expected to fundamentally transform the world economy and (Wiley, 2013).

304 | Nature | Vol 629 | 9 May 2024

35. Khodakov, A. Y., Chu, W. & Fongarland, P. Advances in the development of novel cobalt 71. Xu, C., Paone, E., Rodríguez-Padrón, D., Luque, R. & Mauriello, F. Reductive catalytic
Fischer–Tropsch catalysts for synthesis of long-chain hydrocarbons and clean fuels. routes towards sustainable production of hydrogen, fuels and chemicals from biomass
Chem. Rev. 107, 1692–1744 (2007). derived polyols. Renew. Sustain. Energy Rev. 127, 109852 (2020).
36. Pan, X., Jiao, F., Miao, D. & Bao, X. Oxide–zeolite-based composite catalyst concept that 72. Sudarsanam, P., Peeters, E., Makshina, E. V., Parvulescu, V. I. & Sels, B. F. Advances in
enables syngas chemistry beyond Fischer–Tropsch synthesis. Chem. Rev. 121, porous and nanoscale catalysts for viable biomass conversion. Chem. Soc. Rev. 48,
6588–6609 (2021). 2366–2421 (2019).
37. Shell. Pearl GTL. Shell https://www.shell.com/about-us/major-projects/pearl-gtl.html 73. Corma, A., Iborra, S. & Velty, A. Chemical routes for the transformation of biomass into
(2024). chemicals. Chem. Rev. 107, 2411–2502 (2007).
38. Shell. Synthetic kerosene. Shell https://www.shell.com/business-customers/aviation/ 74. Luterbacher, J. S., Martin Alonso, D. & Dumesic, J. A. Targeted chemical upgrading of
the-future-of-energy/sustainable-aviation-fuel/synthetic-kerosene.html (2021). lignocellulosic biomass to platform molecules. Green Chem. 16, 4816–4838 (2014).
39. Daza, Y. A. & Kuhn, J. N. CO2 conversion by reverse water gas shift catalysis: comparison 75. Gupta, S., Mondal, P., Borugadda, V. B. & Dalai, A. K. Advances in upgradation of pyrolysis
of catalysts, mechanisms and their consequences for CO2 conversion to liquid fuels. bio-oil and biochar towards improvement in bio-refinery economics: a comprehensive
RSC Adv. 6, 49675–49691 (2016). review. Environ. Technol. Innov. 21, 101276 (2021).
40. Bahmanpour, A. M., Signorile, M. & Kröcher, O. Recent progress in syngas production via 76. Fiorentino, G., Ripa, M. & Ulgiati, S. Chemicals from biomass: technological versus
catalytic CO2 hydrogenation reaction. Appl. Catal. B Environ. 295, 120319 (2021). environmental feasibility. A review. Biofuels Bioprod. Biorefin. 11, 195–214 (2017).
41. Lu, Y. et al. Efficient electrocatalytic reduction of CO2 to CO on an electrodeposited Zn 77. Eagan, N. M., Kumbhalkar, M. D., Buchanan, J. S., Dumesic, J. A. & Huber, G. W.
porous network. Electrochem. Commun. 97, 87–90 (2018). Chemistries and processes for the conversion of ethanol into middle-distillate fuels. Nat.
42. Nitopi, S. et al. Progress and perspectives of electrochemical CO2 reduction on copper in Rev. Chem. 3, 223–249 (2019).
aqueous electrolyte. Chem. Rev. 119, 7610–7672 (2019). This article provides an overview of all the catalytic possibilities for making diesel and
This article provides an excellent overview on electrocatalytic conversion of CO2. jet fuel from ethanol.
43. Liu, S., Wang, Y. & Norwood, S. Discovering effective descriptors for CO2 78. Barnés, M. C., de Visser, M. M., van Rossum, G., Kersten, S. R. A. & Lange, J. P. Liquefaction
electro-reduction to predict the catalysts with different selectivity. J. Phys. Chem. C 125, of wood and its model components. J. Anal. Appl. Pyrolysis 125, 136–143 (2017).
4550–4558 (2021). 79. Lange, J. P. Catalysis for biorefineries – performance criteria for industrial operation.
44. Gao, F. Y., Bao, R. C., Gao, M. R. & Yu, S. H. Electrochemical CO2-to-CO conversion: Catal. Sci. Technol. 6, 4759–4767 (2016).
electrocatalysts, electrolytes, and electrolyzers. J. Mater. Chem. A 8, 15458–15478 Key paper highlighting the challenges associated with biomass conversion.
(2020). 80. Lange, J. Performance metrics for sustainable catalysis in industry. Nat. Catal. 4, 186–192
45. Wei, X., Li, Y., Chen, L. & Shi, J. Formic acid electro‐synthesis by concurrent cathodic CO2 (2021).
reduction and anodic CH3OH oxidation. Angew. Chem. 133, 3185–3192 (2021). 81. Plastics Europe. Plastics – the facts 2020. Plastics Europe https://plasticseurope.org/
46. Vass, Á., Kormányos, A., Kószo, Z., Endrődi, B. & Janáky, C. Anode catalysts in CO2 knowledge-hub/plastics-the-facts-2020/ (2020).
electrolysis: challenges and untapped opportunities. ACS Catal. 12, 1037–1051 (2022). 82. Martín, A. J., Mondelli, C., Jaydev, S. D. & Pérez-Ramírez, J. Catalytic processing of plastic
47. Pakhare, D. & Spivey, J. A review of dry (CO2) reforming of methane over noble metal waste on the rise. Chem 7, 1487–1533 (2021).
catalysts. Chem. Soc. Rev. 43, 7813–7837 (2014). 83. Wu, C., Zhang, Y., Zhang, H. & Miskolczi, N. Novel technologies and methods for plastics
48. Vogt, C., Monai, M., Kramer, G. J. & Weckhuysen, B. M. The renaissance of the Sabatier recycling. Process Saf. Environ. Prot. 149, 557–558 (2021).
reaction and its applications on Earth and in space. Nat. Catal. 2, 188–197 (2019). 84. Zhang, F. et al. Current technologies for plastic waste treatment: a review. J. Clean. Prod.
49. Sánchez-Bastardo, N., Schlögl, R. & Ruland, H. Methane pyrolysis for zero-emission 282, 124523 (2021).
hydrogen production: a potential bridge technology from fossil fuels to a renewable and 85. Weckhuysen, B. M. Creating value from plastic waste. Science 370, 400–401 (2020).
sustainable hydrogen economy. Ind. Eng. Chem. Res. 60, 11855–11881 (2021). 86. Lange, J. Managing plastic waste—sorting, recycling, disposal, and product redesign.
50. Hank, C. et al. Economics & carbon dioxide avoidance cost of methanol production ACS Sustain. Chem. Eng. 9, 15722–15738 (2021).
based on renewable hydrogen and recycled carbon dioxide–power-to-methanol. Sustain. 87. Scott, A. Industry plastic-waste initiative advances. Chem. Eng. News 98, 11 (2020).
Energy Fuels 2, 1244–1261 (2018). 88. Tullo, A. H. LyondellBasell opens recycling pilot plant. Chem. Eng. News 98, 12 (2020).
51. Yarulina, I., Chowdhury, A. D., Meirer, F., Weckhuysen, B. M. & Gascon, J. Recent trends 89. Vollmer, I. et al. Beyond mechanical recycling: giving new life to plastic waste. Angew.
and fundamental insights in the methanol-to-hydrocarbons process. Nat. Catal. 1, Chem. Int. Ed. 59, 15402–15423 (2020).
398–411 (2018). This paper provides an overview of the scientific and technological challenges, as well
52. Matieva, Z. M., Snatenkova, Y. M., Kolesnichenko, N. V. & Khadzhiev, S. N. Catalysts for as the perspectives, of chemical recycling of plastic waste.
synthesizing liquid hydrocarbons from methanol and dimethyl ether: a review. Catal. Ind. 90. Ellis, L. D. et al. Chemical and biological catalysis for plastics recycling and upcycling.
11, 101–112 (2019). Nat. Catal. 4, 539–556 (2021).
53. Olsbye, U. et al. The formation and degradation of active species during methanol 91. Di, J., Reck, B. K., Miatto, A. & Graedel, T. E. United States plastics: large flows, short
conversion over protonated zeotype catalysts. Chem. Soc. Rev. 44, 7155–7176 (2015). lifetimes, and negligible recycling. Resour. Conserv. Recycl. 167, 105440 (2021).
54. van Bavel, S., Verma, S., Negro, E. & Bracht, M. Integrating CO2 electrolysis into the 92. Gopinath, K. P., Nagarajan, V. M., Krishnan, A. & Malolan, R. A critical review on the
gas-to-liquids–power-to-liquids process. ACS Energy Lett. 5, 2597–2601 (2020). influence of energy, environmental and economic factors on various processes used to
55. Ouda, M. et al. in Zukünftige Kraftstoffe. ATZ/MTZ-Fachbuch (ed. Maus, W.) 380–409 handle and recycle plastic wastes: development of a comprehensive index. J. Clean.
(Springer, 2019). Prod. 274, 123031 (2020).
56. Nielsen, D. U., Hu, X.-M., Daasbjerg, K. & Skrydstrup, T. Chemically and electrochemically 93. Sivagami, K. et al. Catalytic pyrolysis of polyolefin and multilayer packaging based waste
catalysed conversion of CO2 to CO with follow-up utilization to value-added chemicals. plastics: a pilot scale study. Process Saf. Environ. Prot. 149, 497–506 (2021).
Nat. Catal. 1, 244–254 (2018). 94. Rodríguez-Luna, L., Bustos-Martínez, D. & Valenzuela, E. Two-step pyrolysis for waste
57. Alam, M. I., Cheula, R., Moroni, G., Nardi, L. & Maestri, M. Mechanistic and multiscale HDPE valorization. Process Saf. Environ. Prot. 149, 526–536 (2021).
aspects of thermo-catalytic CO2 conversion to C1 products. Catal. Sci. Technol. 11, 95. Lopez, G., Artetxe, M., Amutio, M., Bilbao, J. & Olazar, M. Thermochemical routes for the
6601–6629 (2021). valorization of waste polyolefinic plastics to produce fuels and chemicals. A review.
58. Maity, S. K. Opportunities, recent trends and challenges of integrated biorefinery: part I. Renew. Sustain. Energy Rev. 73, 346–368 (2017).
Renew. Sustain. Energy Rev. 43, 1427–1445 (2015). 96. Jeswani, H. et al. Life cycle environmental impacts of chemical recycling via pyrolysis
59. Zakzeski, J., Bruijnincx, P. C. A., Jongerius, A. L. & Weckhuysen, B. M. The catalytic of mixed plastic waste in comparison with mechanical recycling and energy recovery.
valorization of lignin for the production of renewable chemicals. Chem. Rev. 110, Sci. Total Environ. 769, 144483 (2021).
3552–3599 (2010). 97. Čolnik, M., Knez, Ž. & Škerget, M. Sub- and supercritical water for chemical recycling of
60. Rinaldi, R. et al. Paving the way for lignin valorisation: recent advances in bioengineering, polyethylene terephthalate waste. Chem. Eng. Sci. 233, 116389 (2021).
biorefining and catalysis. Angew. Chem. Int. Ed. 55, 8164–8215 (2016). 98. Kurzweil, R. The Singularity Is Near (Viking, 2005).
61. Sheldon, R. A. Green and sustainable manufacture of chemicals from biomass: state of 99. Kurzweil, R. The law of accelerating returns. Kurzweil collection https://www.
the art. Green Chem. 16, 950–963 (2014). writingsbyraykurzweil.com/the-law-of-accelerating-returns (2024).
This paper provides a good overview of the field of catalytic biomass conversion. 100. Rosling, H., Rosling, O. & Rosling Rönnlund, A. Factfulness: Ten Reasons We’re Wrong
62. Ennaert, T. et al. Potential and challenges of zeolite chemistry in the catalytic conversion About The World – And Why Things Are Better Than You Think (Hodder & Stoughton,
of biomass. Chem. Soc. Rev. 45, 584–611 (2016). 2019).
63. Melero, J. A., Iglesias, J. & Garcia, A. Biomass as renewable feedstock in standard refinery 101. Climeworks. Climeworks takes another major step on its road to building gigaton
units. Feasibility, opportunities and challenges. Energy Environ. Sci. 5, 7393 (2012). DAC capacity. Climeworks https://climeworks.com/news/climeworks-announces-
64. Huber, G. W., Iborra, S. & Corma, A. Synthesis of transportation fuels from biomass: groundbreaking-on-mammoth (2022).
chemistry, catalysts, and engineering. Chem. Rev. 106, 4044–4098 (2006). 102. International Energy Agency (IEA). Global Hydrogen Review 2021 (IEA, 2021).
65. Kumar, B. & Verma, P. Biomass-based biorefineries: an important architype towards a 103. Haru Oni. Tomorrow’s fuel. Haru Oni https://www.haruoni.com/#/en (2021).
circular economy. Fuel 288, 119622 (2021). 104. Ehteshami, S. M. M. & Chan, S. H. The role of hydrogen and fuel cells to store renewable
66. Han, X., Wang, H., Zeng, Y. & Liu, J. Advancing the application of bio-oils by co-processing energy in the future energy network – potentials and challenges. Energy Policy 73,
with petroleum intermediates: a review. Energy Convers. Manag. X 10, 100069 (2020). 103–109 (2014).
67. Huber, G. W. & Corma, A. Synergies between bio- and oil refineries for the production of 105. Pearson, P. J. G. & Foxon, T. J. A low carbon industrial revolution? Insights and challenges
fuels from biomass. Angew. Chem. Int. Ed. 46, 7184–7201 (2007). from past technological and economic transformations. Energy Policy 50, 117–127
68. Wang, Z., Burra, K. G., Lei, T. & Gupta, A. K. Co-pyrolysis of waste plastic and solid biomass (2012).
for synergistic production of biofuels and chemicals-a review. Prog. Energy Combust. Sci. 106. Nazir, H. et al. Is the H2 economy realizable in the foreseeable future? Part III: H2 usage
84, 100899 (2021). technologies, applications, and challenges and opportunities. Int. J. Hydrog. Energy 45,
69. Ondrey, G. Making renewable diesel and biopropane from vegetable oil. Chem. Eng. 121, 28217–28239 (2020).
11–12 (2014). 107. Yates, J. et al. Techno-economic analysis of hydrogen electrolysis from off-grid
70. Dayton, D. C. et al. Design and operation of a pilot-scale catalytic biomass pyrolysis unit. stand-alone photovoltaics incorporating uncertainty analysis. Cell Rep. Phys. Sci. 1,
Green Chem. 17, 4680–4689 (2015). 100209 (2020).

Nature | Vol 629 | 9 May 2024 | 305

Perspective
108. International Energy Agency (IEA). Average annual capacity factors by technology, 2018. 142. Levi, P. G. & Cullen, J. M. Mapping global flows of chemicals: from fossil fuel feedstocks to
IEA https://www.iea.org/data-and-statistics/charts/average-annual-capacity-factors-by- chemical products. Environ. Sci. Technol. 52, 1725–1734 (2018).
technology-2018 (2019). 143. Garcia‐Martinez, J. Chemistry 2030: a roadmap for a new decade. Angew. Chem. Int. Ed.
109. UK Department for Business, Energy & Industrial Strategy. Digest of UK Energy Statistics 60, 4956–4960 (2021).
(DUKES). gov.uk https://www.gov.uk/government/collections/digest-of-uk-energy- Visionary article on the transition from chemistry of transformation towards chemistry
statistics-dukes (2020). of reuse.
110. Global Solar Atlas. Global Photovoltaic Power Potential by Country; Country factsheet: 144. Talanquer, V., Bucat, R., Tasker, R. & Mahaffy, P. G. Lessons from a pandemic: educating
Netherlands. globalsolaratlas.info https://globalsolaratlas.info/global-pv-potential-study for complexity, change, uncertainty, vulnerability, and resilience. J. Chem. Educ. 97,
(2021). 2696–2700 (2020).
111. Global Solar Atlas. Global Photovoltaic Power Potential by Country; Country factsheet: 145. Zheng, L., Ambrosetti, M. & Tronconi, E. Joule-heated catalytic reactors toward
Denmark. globalsolaratlas.info https://globalsolaratlas.info/global-pv-potential-study decarbonization and process intensification: a review. ACS Eng. Au 4, 4–21 (2024).
(2021). 146. Zheng, L., Ambrosetti, M., Beretta, A., Groppi, G. & Tronconi, E. Electrified CO2
112. U.S. Energy Information Administration. Capacity factors for utility scale generators valorization driven by direct Joule heating of catalytic cellular substrates. Chem. Eng. J.
primarily using non-fossil fuels. U.S. Energy Information Administration https://www.eia. 466, 143154 (2023).
gov/electricity/annual/html/epa_04_08_b.html (2021). 147. Kim, Y. T., Lee, J.-J. & Lee, J. Electricity-driven reactors that promote thermochemical
113. GE Renewable Energy. Haliade-X offshore turbine. GE Renewable Energy https://www. catalytic reactions via joule and induction heating: a review. Chem. Eng. J. 470, 144333
ge.com/renewableenergy/wind-energy/offshore-wind/haliade-x-offshore-turbine (2022). (2023).
114. European Commission. A Hydrogen Strategy for a Climate-Neutral Europe (European 148. Balakotaiah, V. & Ratnakar, R. R. Modular reactors with electrical resistance heating
Commission, 2020). for hydrocarbon cracking and other endothermic reactions. AIChE J. 68, e17542
115. European Commission. Facts Behind the Debate: Direct Air Capture (European Commission, (2022).
2019). 149. Liu, L. et al. Alkane dehydrogenation in scalable and electrifiable carbon membrane
116. Beuttler, C., Charles, L. & Wurzbacher, J. The role of direct air capture in mitigation of reactor. Cell Rep. Phys. Sci. 4, 101692 (2023).
anthropogenic greenhouse gas emissions. Front. Clim. 1, 469555 (2019). 150. BASF. BASF, SABIC and Linde start construction of the world’s first demonstration plant
117. Climeworks. Orca: the first large-scale plant. Climeworks https://climeworks.com/ for large-scale electrically heated steam cracker furnaces. BASF https://www.basf.com/
roadmap/orca (2021). global/en/who-we-are/sustainability/whats-new/sustainability-news/2022/basf-sabic-
118. Tsiropoulos, I., Tarvydas, D. & Zucker, A. Cost Development of Low Carbon Energy and-linde-start-construction-of-the-worlds-first-demonstration-plant-for-large-scale-
Technologies: Scenario-based Cost Trajectories to 2050, 2017 Edition (Publications electrically-heated-steam-cracker-furnaces.html (2022).
Office of the European Union, 2018). 151. Coolbrook. Coolbrook successfully cracks naphtha in its large-scale pilot plant leading
119. Mayer, J. N. et al. Current and Future Cost of Photovoltaics. Long-term Scenarios for the way to zero-carbon steam cracking. Coolbrook https://coolbrook.com/news/
Market Development, System Prices and LCOE of Utility-Scale PV Systems coolbrook-successfully-cracks-naphtha-in-its-large-scale-pilot-plant-leading-the-way-
(Fraunhofer-Institute for Solar Energy Systems, 2015). to-zero-carbon-steam-cracking/ (2023).
120. International Renewable Energy Agency (IRENA). Future of Solar Photovoltaic: 152. Stern, N. The Economics of Climate Change (Cambridge Univ. Press, 2007).
Deployment, Investment, Technology, Grid Integration and Socio-economic Aspects 153. Air Liquide. Air Liquide inaugurates the world’s largest low-carbon hydrogen membrane-
(IRENA, 2019). based production unit in Canada. Air Liquide https://energies.airliquide.com/air-liquide-
121. Sens, L., Neuling, U. & Kaltschmitt, M. Capital expenditure and levelized cost of electricity inaugurates-worlds-largest-low-carbon-hydrogen-membrane-based-production-unit-
of photovoltaic plants and wind turbines – development by 2050. Renew. Energy 185, canada (2021).
525–537 (2022). 154. Küngas, R. Review—Electrochemical CO2 reduction for CO production: comparison of
122. International Renewable Energy Agency (IRENA). Future of Wind: Deployment, low- and high-temperature electrolysis technologies. J. Electrochem. Soc. 167, 044508
Investment, Technology, Grid Integration and Socio-economic Aspects (IRENA, 2019). (2020).
123. Mayyas, A., Ruth, M., Pivovar, B., Bender, G. & Wipke, K. Manufacturing Cost Analysis for 155. Jungmeier, G. The Biorefinery Fact Sheet (IEA Bioenergy, 2014).
Proton Exchange Membrane Water Electrolyzers (National Renewable Energy Laboratory, 156. Cardona-Alzate, C. A., Serna-Loaiza, S. & Ortiz-Sanchez, M. Sustainable biorefineries:
2019). what was learned from the design, analysis and implementation. J. Sustain. Dev. Energy
124. International Renewable Energy Agency (IRENA). Green Hydrogen Cost Reduction: Water Environ. Syst. 8, 88–117 (2020).
Scaling up Electrolysers to Meet the 1.5°C Climate Goal (IRENA, 2020). 157. U.S. Energy Information Administration. Oil and petroleum products explained: refining
125. Lord, A. S., Kobos, P. H. & Borns, D. J. Geologic storage of hydrogen: scaling up to meet crude oil. U.S. Energy Information Administration https://www.eia.gov/energyexplained/
city transportation demands. Int. J. Hydrog. Energy 39, 15570–15582 (2014). oil-and-petroleum-products/refining-crude-oil-inputs-and-outputs.php (2019).
126. World Nuclear Association. Nuclear power in France. World Nuclear Association https:// 158. International Energy Agency (IEA). Net Zero by 2050: A Roadmap for the Global Energy
world-nuclear.org/information-library/country-profiles/countries-a-f/france.aspx (2024). Sector (IEA, 2021).
127. Palacios, G. & Janssen, J. Nuclear Energy Economics: An Update to Fact Finding Nuclear 159. International Energy Agency (IEA). Global Hydrogen Review 2022 (IEA, 2022).
Energy (TNO, 2018).
128. Massachusetts Institute of Technology (MIT). The Future of Nuclear Energy in a
Carbon-Constrained World (MIT, 2018).
129. Carrara, S., Alves Dias, P., Plazzotta, B. & Pavel, C. Raw Materials Demand for Wind and Acknowledgements B.M.W. acknowledges financial support from the Netherlands
Solar PV Technologies in the Transition Towards a Decarbonised Energy System Organization for Scientific Research (NWO) in the frame of a Gravitation Programme MCEC
(Publications Office of the European Union, 2020). (Netherlands Center for Multiscale Catalytic Energy Conversion; www.mcec-researchcenter.nl),
130. U.S. Geological Survey. U.S. Geological Survey releases 2022 list of critical minerals. as well as from the Advanced Research Center (ARC) Chemical Building Blocks Consortium
U.S. Geological Survey https://www.usgs.gov/news/national-news-release/us-geological- (CBBC), a public–private research consortium in the Netherlands (www.arc-cbbc.nl). We
survey-releases-2022-list-critical-minerals (2022). acknowledge the help of G.-J. Kramer at Utrecht University (UU), the Netherlands, and C. Vogt
131. Kiemel, S. et al. Critical materials for water electrolysers at the example of the energy at Technion, Israel Institute of Technology in Haifa, Israel, for valuable discussions. T. Hartman
transition in Germany. Int. J. Energy Res. 45, 9914–9935 (2021). (UU) is acknowledged for help with the figures.
132. Vidal, O., Goffé, B. & Arndt, N. Metals for a low-carbon society. Nat. Geosci. 6, 894–896
(2013). Author contributions E.T.C.V. and B.M.W. contributed to the conceptualization, analysis,
Paper articulating that a shift to renewable energy will replace one non-renewable literature review, writing and responses to reviewer comments for this manuscript.
resource (fossil fuel) by another non-renewable resource (metals and minerals).
133. Panoutsou, C. & Maniatis, K. Sustainable Biomass Availability in the EU, to 2050 (Imperial Competing interests The authors declare no competing interests.
College London Consultants, 2021).
134. Ellen MacArthur Foundation. The New Plastics Economy: Rethinking the Future of Plastics
Additional information
(World Economic Forum, Ellen MacArthur Foundation and McKinsey & Company, 2016).
Supplementary information The online version contains supplementary material available at
135. Mallapragada, D. S. et al. Decarbonization of the chemical industry through
https://doi.org/10.1038/s41586-024-07322-2.
electrification: barriers and opportunities. Joule 7, 23–41 (2023).
Correspondence and requests for materials should be addressed to Eelco T. C. Vogt or
136. Meloni, E. Electrification of chemical engineering: a new way to intensify chemical
Bert M. Weckhuysen.
processes. Energies 15, 5469 (2022).
Peer review information Nature thanks Peter Levi, Shelie Miller and the other, anonymous,
137. van Geem, K. M. & Weckhuysen, B. M. Toward an e-chemistree: materials for electrification
reviewer(s) for their contribution to the peer review of this work.
of the chemical industry. MRS Bull. 46, 1187–1196 (2021).
Reprints and permissions information is available at http://www.nature.com/reprints.
138. Winiwarter, A. et al. Towards an atomistic understanding of electrocatalytic partial
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
hydrocarbon oxidation: propene on palladium. Energy Environ. Sci. 12, 1055–1067 (2019).
published maps and institutional affiliations.
139. Parkinson, B. et al. Hydrogen production using methane: techno-economics of
decarbonizing fuels and chemicals. Int. J. Hydrog. Energy 43, 2540–2555 (2018).
140. Pavičić, J., Mavar, K. N., Brkić, V. & Simon, K. Biogas and biomethane production and Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
usage: technology development, advantages and challenges in Europe. Energies 15, article under a publishing agreement with the author(s) or other rightsholder(s); author
2940 (2022). self-archiving of the accepted manuscript version of this article is solely governed by the
141. Scheiblehner, D., Neuschitzer, D., Wibner, S., Sprung, A. & Antrekowitsch, H. Hydrogen terms of such publishing agreement and applicable law.
production by methane pyrolysis in molten binary copper alloys. Int. J. Hydrog. Energy
48, 6233–6243 (2023). © Springer Nature Limited 2024

306 | Nature | Vol 629 | 9 May 2024