Nobel Prize in Chemistry 2013: Multiscale Models for Complex Chemical Systems

This note covers the Nobel Prize in Chemistry 2013: who won it, what "multiscale models" mean, how classical physics and quantum physics were combined in one computer method, how the idea developed from the 1970s onward, why the method matters for modern chemistry and biology, and quick facts for exams.

What was the Nobel Prize in Chemistry 2013 awarded for?

The Royal Swedish Academy of Sciences gave the prize jointly to three scientists "for the development of multiscale models for complex chemical systems". This is the exact wording of the official citation.

In plain words, the three laureates built computer methods that let chemists study reactions happening inside huge molecules, such as proteins or enzymes, without needing impossible amounts of computing power.

Before their work, scientists had to choose between two kinds of physics to model a molecule: classical mechanics (the simple physics of balls and springs, good for large molecules but unable to show a reaction actually happening) or quantum mechanics (the physics of electrons, accurate for reactions but only usable on tiny molecules because it needs enormous computer time).

The laureates showed how to use both together in a single calculation, applying the detailed, expensive quantum method only to the small part of the molecule where the chemistry actually happens, and the cheaper classical method to everything around it.

The official name of this award is the Nobel Prize in Chemistry, presented each year by the Royal Swedish Academy of Sciences and announced in October.

Who are the laureates?

All three laureates shared the prize equally, each receiving one third, for jointly developing the same family of methods through sustained collaboration in the 1970s.

Martin Karplus

Martin Karplus was born on 15 March 1930 in Vienna, Austria, and died on 28 December 2024 in Cambridge, Massachusetts, USA.

At the time of the award he was affiliated with both the Université de Strasbourg in France and Harvard University in Cambridge, USA. He received one third of the prize.

Karplus had a strong background in quantum chemistry, and his Harvard laboratory was where the first attempts to combine quantum and classical calculations began in the early 1970s.

He is also known among chemists for the "Karplus equation", used in nuclear magnetic resonance (NMR) spectroscopy.

Michael Levitt

Michael Levitt was born on 9 May 1947 in Pretoria, South Africa. At the time of the award he worked at Stanford University School of Medicine in Stanford, California, USA, and received one third of the prize.

Levitt trained at the University of Cambridge studying large molecules of life, including DNA, RNA and proteins, and he later worked again with Arieh Warshel to build a general quantum-classical model that could handle enzymes and proteins of any size.

Arieh Warshel

Arieh Warshel was born on 20 November 1940 in Kibbutz Sde-Nahum, in what was then the British Mandate of Palestine and is now Israel.

At the time of the award he worked at the University of Southern California in Los Angeles, USA, and received one third of the prize.

Warshel trained at the Weizmann Institute of Science in Rehovot, Israel, where he helped build an early classical computer program for modelling large molecules, before joining Karplus at Harvard and later working again with Levitt to build a general method for simulating enzyme reactions.

What problem were chemists trying to solve?

Chemical reactions happen extremely fast. According to the prize committee's press release, electrons can jump from one atom to another in a fraction of a millisecond, a speed that makes it "virtually impossible to experimentally map every little step in a chemical process".

Traditional laboratory experiments can show what goes into a reaction and what comes out, but they struggle to reveal the in-between stages, called the transition state, where bonds are actually breaking and forming.

Before computers became powerful, chemists built plastic ball-and-stick models to represent atoms and bonds.

These models were useful for showing the resting shape of a molecule, but they could say nothing about how a reaction actually proceeds, because a real reaction involves electrons rearranging themselves, which is a quantum effect, not something that balls and sticks can show.

Scientists who wanted to study very large biological molecules, such as proteins, DNA and enzymes, faced a particular difficulty. These molecules can contain tens of thousands of atoms.

A full quantum mechanical calculation on a system that size would need far more computing power than existed, even with today's computers, because quantum calculations have to track every electron and every atomic nucleus.

Something had to give: either accuracy or size. The three laureates, working mostly in the 1970s, set out to find a way to keep the accuracy of quantum physics for the important atoms while still being able to model enormous molecules such as entire enzymes.

How do classical and quantum mechanics combine in multiscale models?

A multiscale model is a computer method that treats different parts of the same molecule with different levels of detail, so that computing power is spent only where it is really needed.

The two underlying kinds of physics have opposite strengths and weaknesses, as the table below summarises, based on the prize committee's press release.

The laureates' key insight was to let the two methods work side by side within one calculation rather than forcing chemists to pick one or the other.

The quantum part of the calculation is applied only to the atoms directly involved in breaking or forming bonds, which the academy's popular-science account calls the "heart of the action", while the classical part handles the rest of the molecule, which mostly just sits nearby without changing chemically.

The academy's popular account described this collaboration vividly, noting that in these models "Newton and his apple collaborate with Schrödinger and his cat", a phrase used by the Royal Swedish Academy of Sciences to describe the joining of two historically rival branches of physics.

How did the laureates build a universal multiscale method?

The first working version of this idea came from Martin Karplus's laboratory at Harvard in the early 1970s, when Arieh Warshel arrived after finishing his doctorate at the Weizmann Institute of Science.

Warshel brought with him a classical modelling program he had built there with Michael Levitt using a powerful computer called the Golem.

Working together, Karplus and Warshel built a new program that applied quantum calculations only to certain mobile "free electrons", found for example in the light-sensing molecule retinal, while treating the rest of the molecule classically.

They published this method in 1972, but it only worked for molecules with mirror symmetry, which was a serious limitation.

The breakthrough that removed this limitation came a few years later, when Levitt reunited with Warshel. Their goal was to model enzymes, the proteins that control almost all chemical reactions inside living cells. In general, the steps that a modern multiscale (combined quantum/classical) calculation follows are:

1. Identify which small group of atoms is directly involved in the chemical reaction, such as the atoms where bonds are breaking or forming.
2. Apply the detailed, computationally expensive quantum mechanical calculation only to that small core region.
3. Apply the simpler, faster classical mechanical calculation to the rest of the molecule surrounding the core.
4. Build coupling terms that let the quantum region and the classical region interact with each other in a physically sensible way.
5. For the parts of the system far from the reaction, group many atoms together or treat them as a single uniform background, called a dielectric medium, to save even more computing effort.

In 1976, Warshel and Levitt published a general version of this scheme, studying how the enzyme lysozyme breaks a chemical bond. Unlike the 1972 method, this one placed no restriction on the shape or size of the molecule, so it could be applied to any enzyme or protein.

Draw a large blob representing a protein or enzyme. In the very centre, mark a small cluster of atoms and label it "quantum region: bonds break and form here".

Draw a ring around it labelled "classical region: atoms modelled as balls and springs". Draw an outer shaded ring labelled "dielectric medium: distant atoms lumped together as background".

What can multiscale models simulate today?

Once the core idea was established, Levitt and Warshel refined it further to make calculations even more efficient.

They showed that groups of several atoms in the less important, outer parts of a molecule could be merged and treated as single larger units rather than tracking every atom separately, which saved further computing effort without losing useful accuracy in the regions that mattered.

The press release gave concrete examples of what such simulations are used for today, including how a catalyst helps purify exhaust fumes from a vehicle, how photosynthesis occurs inside the green parts of plants, and how a drug molecule binds to its target protein inside the body.

In the drug-binding example, the computer performs the demanding quantum calculation only on the atoms of the target protein that actually touch the drug, while the rest of the large protein is handled with the cheaper classical approach.

Because the method does not depend on any particular molecule's shape or size, it has been described by the academy as universal: the same underlying approach can study the molecules of life, such as enzymes, DNA and proteins, as well as industrial chemical processes such as catalysis.

The press release summed up the overall shift in chemistry this way: "The computer – your Virgil in the world of atoms", a phrase from the press release, pointing to how computer simulation now guides chemists through processes that cannot be watched directly in a test tube.

How did the discovery unfold?

The method did not appear overnight. It grew out of separate lines of work on classical molecular models and quantum chemistry that gradually merged over about two decades, as the laureates' careers brought them together at the same institutions at the right moments.

Why does it matter?

The multiscale method turned the computer into a routine tool for chemistry, alongside the test tube.

The academy's materials state plainly that today chemists spend as much time at their computers as in their laboratories, because realistic simulations can now predict the outcome of real experiments before they are ever carried out on a bench.

Practical uses mentioned in the official sources include designing better catalysts that clean exhaust gases from vehicles, understanding how photosynthesis splits water molecules inside green leaves, and predicting how drug molecules bind to their target proteins in the human body, which helps in designing medicines.

The scientific background document also notes that the approach has spread well beyond the laureates' own work, with other scientists extending it to organic chemistry, biochemistry, and the chemistry of molecules dissolved in liquids.

The method also has open horizons. The popular science text records that Michael Levitt wrote about dreaming of simulating an entire living organism at the molecular level, and the academy's own account leaves the ultimate reach of the method as a question for the future: "exactly how far they can advance our knowledge is for the future to decide."

How does this connect to what you study?

This prize sits exactly at the meeting point of physics and chemistry that many school courses introduce separately.

Classical mechanics, the physics of forces, motion and springs that describes how everyday objects move, is the same physics used to model the "background" atoms in these simulations.

Quantum mechanics, which describes how tiny particles such as electrons behave, including ideas like electrons jumping between energy levels, is the physics needed wherever bonds are actually breaking or forming.

Topics such as atomic structure, chemical bonding and reaction mechanisms in chemistry classes are the everyday versions of exactly what this prize's computer methods try to capture mathematically.

Where a textbook diagram shows electrons moving during a reaction with an arrow, a multiscale simulation tries to calculate that same movement numerically, atom by atom, inside a far larger molecule.

The prize is also a useful example of how two very different theories in physics, built up over nearly a century, were eventually joined for a practical purpose.

It shows that scientific progress often comes not from a single new discovery but from finding a clever way to combine existing ideas that previously seemed incompatible.

Quick facts for exams

The Nobel Prize in Chemistry 2013 was awarded to Martin Karplus, Michael Levitt and Arieh Warshel, each receiving one third of the prize, "for the development of multiscale models for complex chemical systems".

The prize was announced on 9 October 2013 by the Royal Swedish Academy of Sciences. Karplus was born in Vienna, Austria, and was affiliated with the Université de Strasbourg, France, and Harvard University, USA.

Levitt was born in Pretoria, South Africa, and was affiliated with Stanford University School of Medicine, USA. Warshel was born in Kibbutz Sde-Nahum, then British Mandate of Palestine, now Israel, and was affiliated with the University of Southern California, USA.

The total prize amount that year was 8,000,000 Swedish kronor, shared equally.

Note: Source. The prize facts in this note are from the Nobel Prize's official site, nobelprize.org.

Glossary

• Multiscale model — a computer model that treats different parts of the same molecule with different levels of physical detail.
• Classical mechanics — the physics of forces and motion, such as Newton's laws, used here to model large groups of atoms simply and quickly.
• Quantum mechanics — the physics describing tiny particles such as electrons, needed to show bonds breaking or forming during a reaction.
• Transition state — the fleeting, highest-energy arrangement of atoms that links the starting materials of a reaction to its products.
• Enzyme — a protein that speeds up and controls a chemical reaction inside a living cell.
• Catalyst — a substance that speeds up a chemical reaction, such as the catalyst that cleans exhaust gases from a vehicle.
• Dielectric medium — a simplified, uniform background used in a simulation to represent distant atoms and molecules that are not individually tracked.
• Retinal — a light-sensitive molecule in the retina of the eye whose quantum behaviour was an early test case for the laureates' methods.
• Lysozyme — an enzyme that breaks a bond in a sugar chain, used by Warshel and Levitt in 1976 to test their general model.
• Protein folding — the process by which a chain of amino acids arranges itself into its working three-dimensional shape.
• Nuclear magnetic resonance (NMR) — a laboratory method used to study molecular structure, linked to the "Karplus equation" that Martin Karplus developed.
• Prize share — the fraction of a Nobel Prize, and of its money, given to each laureate when several share one award.

Common errors and misconceptions

• Misconception: The laureates invented quantum mechanics or classical mechanics. Correct: Both branches of physics already existed; the laureates developed a way to combine them in one computer calculation.
• Misconception: The method only works on small, simple molecules. Correct: Its main strength is that it can handle very large molecules, such as whole enzymes and proteins, by treating most atoms classically.
• Misconception: Each laureate worked entirely alone on a separate idea. Correct: The method grew through direct collaboration, with Warshel working first with Karplus and later again with Levitt.
• Misconception: The 1972 method already solved the whole problem. Correct: It only worked for molecules with mirror symmetry; the general method for any molecule came in 1976.
• Misconception: Quantum calculations are used for the whole molecule in these models. Correct: Quantum calculations are applied only to the small core region where the reaction occurs.
• Misconception: This prize is mainly about physics, not chemistry. Correct: The citation specifically credits the laureates for chemistry, namely modelling complex chemical systems using physics as a tool.

Exam-style questions with model answers

Q1. In which year was the Nobel Prize in Chemistry 2013 announced? [1 mark]

1. It was announced on 9 October 2013 by the Royal Swedish Academy of Sciences.

Q2. Name the three laureates of the Nobel Prize in Chemistry 2013. [2 marks]

1. The laureates were Martin Karplus, Michael Levitt and Arieh Warshel, who shared the prize equally for developing multiscale models for complex chemical systems.

Q3. State the official citation for the Nobel Prize in Chemistry 2013. [2 marks]

1. The citation reads "for the development of multiscale models for complex chemical systems".

Q4. Explain why chemists could not use classical mechanics alone to study a chemical reaction. [4 marks]

1. Classical mechanics treats atoms as simple objects obeying Newtonian physics, which allows fast calculations on large molecules and shows a resting, stable shape clearly.
2. However, a chemical reaction involves electrons rearranging as bonds break and form, which is a quantum effect that classical physics has no way to represent.
3. So classical models could only show a molecule at rest, not the fast-moving process of a reaction actually taking place.
4. Chemists therefore needed a way to add quantum physics into the calculation without losing the speed that classical physics gave for large molecules.

Q5. Describe how a multiscale model splits a large molecule for calculation. [4 marks]

1. The small group of atoms where bonds actually break or form is identified as the reaction core.
2. This core is treated with detailed, computationally expensive quantum mechanical calculations.
3. The rest of the molecule, which does not change chemically, is treated with simpler, faster classical mechanics.
4. Coupling terms link the two regions, and very distant atoms may be lumped together as a uniform dielectric background to save further computing effort.

Q6. Give three practical applications of multiscale modelling mentioned by the Nobel committee. [3 marks]

1. One application is designing catalysts that purify exhaust fumes from vehicles more effectively.
2. Another is studying how drug molecules bind to their target proteins in the body, which helps in designing medicines.
3. The committee also mentioned using the method to understand how photosynthesis splits water molecules inside green leaves.

Q7. Trace the development of multiscale modelling from 1970 to 1976, naming the key publications and limitations at each stage. [6 marks]

1. In 1970, Arieh Warshel joined Martin Karplus's laboratory at Harvard, bringing a classical computer program he had built earlier with Michael Levitt using the Golem computer at the Weizmann Institute.
2. In 1972, Karplus and Warshel published the first method combining quantum and classical calculations, applying quantum physics only to certain mobile electrons and classical physics to the rest, but it only worked for molecules with mirror symmetry.
3. In 1975, Levitt and Warshel studied the folding of a protein called bovine pancreatic trypsin inhibitor, showing that groups of atoms could be simplified into larger units to speed up calculations.
4. In 1976, Warshel and Levitt published a general model, applied to the enzyme lysozyme, that removed the symmetry restriction, introduced coupling terms between regions, and could be used on molecules of any size.
5. This progression turned a limited, special-case method into a universal tool usable across organic chemistry and biochemistry.
6. Later refinements by Levitt and Warshel also allowed distant atoms to be merged into single units, saving further computing effort.

Q8. Why did the Nobel committee describe the laureates' work as bringing together Newton and Schrödinger? [5 marks]

1. Classical mechanics, associated with Newton's laws, is the physics historically used to model the motion and position of everyday-sized objects and large groups of atoms.
2. Quantum mechanics, associated with ideas such as Schrödinger's description of particles, is needed to capture the behaviour of electrons during a chemical reaction.
3. Before the laureates' work, these two theories were used separately, because combining them in one calculation on a real chemical system had not been achieved.
4. The laureates built computer methods that let quantum calculations handle the reactive core of a molecule while classical calculations handled everything else, allowing the two physics theories to work together in a single simulation.
5. The academy's popular account summed this up by saying that in these models, Newton and Schrödinger's cat effectively collaborate.

Q9. What is the role of a dielectric medium in a modern multiscale calculation? [3 marks]

1. A dielectric medium is a simplified, uniform background used for the parts of a molecule that are very far from the chemical reaction.
2. Instead of tracking every distant atom individually, the computer treats this outer region as a single homogeneous mass.
3. This saves further computing effort while still letting the quantum core and the surrounding classical region interact realistically.

Q10. Where did each laureate work at the time the Nobel Prize in Chemistry 2013 was announced? [2 marks]

1. Martin Karplus was affiliated with the Université de Strasbourg and Harvard University, Michael Levitt with Stanford University School of Medicine, and Arieh Warshel with the University of Southern California.

Key takeaways

• The Nobel Prize in Chemistry 2013 went to Martin Karplus, Michael Levitt and Arieh Warshel for multiscale models of complex chemical systems.
• Each laureate received one third of the prize, and the total prize amount was 8,000,000 Swedish kronor.
• Classical mechanics is fast and suits large molecules, but cannot show a reaction happening; quantum mechanics shows reactions but needs huge computing power.
• The laureates' method applies quantum calculations only to the reactive core of a molecule and classical calculations to the rest.
• The first hybrid model, in 1972, only worked for mirror-symmetric molecules; the general model, in 1976, worked for any molecule.
• Distant atoms in a simulation can be lumped together as a uniform background, called a dielectric medium, to save computing effort.
• Applications mentioned by the Nobel committee include catalysts for exhaust purification, photosynthesis, and drug-protein binding.
• The method is described as universal because it applies across organic chemistry, biochemistry and industrial chemical processes alike.

Test yourself

Who announced the Nobel Prize in Chemistry 2013, and on what date?

The Royal Swedish Academy of Sciences announced the prize on 9 October 2013.

What share of the prize did each laureate receive?

Martin Karplus, Michael Levitt and Arieh Warshel each received one third of the Nobel Prize in Chemistry 2013.

Where was Michael Levitt affiliated at the time of the award?

Michael Levitt was affiliated with Stanford University School of Medicine in Stanford, California, USA, at the time of the award.

Why could classical mechanics alone not simulate a chemical reaction?

Classical mechanics cannot represent electrons rearranging as bonds break and form, which is a quantum effect essential to any reaction.

What limitation did the 1972 Karplus-Warshel method have?

The 1972 method could only be applied to molecules with mirror symmetry, so it was not a general solution.

Which enzyme did Warshel and Levitt study in their 1976 general model?

They studied lysozyme, an enzyme that breaks a bond in a sugar chain, to test their general quantum-classical model.

Give one real-world use of multiscale modelling mentioned by the Nobel committee.

One use is designing catalysts that help purify exhaust fumes from vehicles more effectively.

What is a dielectric medium in this context?

It is a simplified, uniform background used in a simulation to represent distant atoms not individually tracked, saving computing effort.