In 1982, the Nobel Prize-winning physicist Richard Feynman predicted that digital computers for the development, simulation, and analysis of chemical molecules would soon be ineffective.
He was right.
Despite home PCs having computing power way beyond what even Bill Gates might have foreseen in 1982, digital computers of any size are not powerful enough for some tasks, … and they never will be.
What is needed, specifically for the chemicals industry (as this article will explain), is a new kind of computer, a quantum computer that uses qubits of power.
As Iris Herrmann, a partner at the PwC Network consultancy Strategy&, explains in an insightful Pulse article on LinkedIn, “In contrast to regular bits, qubits can be in the state of 0 and 1 at the same time. Consequently, a qubit can be 0, 1 or any fraction between the two at one time. Therefore, two qubits can represent four numbers, three qubits eight etc. Through this superposition, quantum systems can be in many different states at the same time.”
There are many ways to construct qubits of thinking power; quantum computers such as annealers and gate-type designs, or superconducting, ion trap, or topological devices. At present, the superconducting model has taken an early lead in terms of development, with Google, Intel and IBM, each having constructed quantum processors ranging from 49 to 72 qubits.
They are, as yet, not very big, but are still very powerful. As the computing website The Next Web claims, “At 100 qubits a single quantum computer processor would, theoretically, be more powerful than all the supercomputers on the planet combined.”
While such devices are never likely to make it into your workplace, they will hold a central role in industry, and will be key to chemical company research. As the Royal Chemical Society’s journal Chemistry World explains, “Quantum computers will not supersede the traditional transistor-based computer on your desktop, but for tackling the problems for which they are suited, nothing else will come close.”
This is because QCs have the ability to model the ‘multi-reference states of molecules’. These states occur when molecules interact. They are sometimes described in layman’s terms as ‘excited states’.
Given the hundreds of thousands of different molecules available, each with its own number of electrons, combining molecules to make new chemical structures requires computing the billions of permutations of molecules and electron interactions, including their ‘multi-reference states’.
As the computing journal Singularity Hub explains, “The reason such modelling is significant is that ‘classical’ digital computers find it virtually impossible to tackle multi-reference states; in many cases, classical computing methods fail not only quantitatively but also qualitatively in the description of the electronic structure of the molecules.”
At present, some of the world’s most powerful digital computers are failing at effective molecule design and is why QC modelling of new chemical molecules is so important.
In October 2018, quantum computers finally made the breakthrough the chemical industry had been waiting for, as researchers from Cambridge Quantum Computing in collaboration with a team from JSR Corp announced that they had, “successfully implemented state-of-the-art quantum algorithms to calculate the excited states of molecules that take into account multi-reference characteristics.”
This is where the chemical industry could at last step in.
“Up to now, quantum computing has been dominated by physicists, engineers and computer scientists who have designed and constructed the hardware and architecture. But it is chemists who have the knowledge and expertise needed to describe and define the quantum nature of the chemical problems that these machines are so adept at solving,” observed the journal Chemistry World. “Quantum computer software needs to be partially written in the language of chemistry.”
Foreseeing this breakthrough, the biggest names in the chemical industry have already been investing heavily in QCs. They see it as both a time and money saving strategy. The reason for this is twofold.
1 Current Development of New Chemical Molecules is an Inefficient Process
For example, to create a chemical product, such as a new surfactant or cleaner, requires a highly knowledgeable chemist, a series of often costly and time-consuming experiments, and theoretical models that are part human idea and part computer generated.
Yet, as a recent study on the topic by chemical industry consultants at McKinsey notes, “Quantum computing could help this process with optimization calculations to understand exactly how, for example, detergent molecules interact with a wine stain on a fibre and to identify the best active ingredients and formulations to remove it. A team using a quantum computer and the appropriate algorithm could reduce the required calculation time to seconds.”
It would also remove much of the time and money required on laboratory experiments.
2 The Quantum Computer Processing Limit to Perform the Task in Reason 1 is Low
While other industries are also keen to employ QC, the benefits for the chemical industry are much easier to take, as the computing level required for chemical product R&D is much lower.
For example, a study co-authored by a team from BASF and the Karlsruhe Institute of Technology found the computing capability required to simulate a chemical procedure such as the Haber-Bosch ammonia process to be about 1,000 qubits.
Whereas a 2015 study, by a team from Duke University, found that if QC were to be used for ground-breaking work in the technology industry a much higher level of processing would be needed. For example, the study suggested that QC could be used for developing more secure communications with RSA encryption. Such high-level encryption could require finding two numbers that multiple together to create a 1,024-bit prime number. The Duke team calculated that the computing resources required to generate this would be in the region of 1.5 million qubits.
Even then, the calculation may not be fast enough for practical applications, as the study also states that, “A trapped-ion quantum computer designed with twice as many qubits and one-tenth of the baseline infidelity of the communication channel can factor a 2,048-bit integer in less than five months.”
Clearly there is work to do before quantum computing becomes the saviour of all mankind and renders the human brain redundant. However, the progress made is reaching the level required for practical, real world applications in the chemical industry.
Given the importance of research and the development of new chemical products, it is fair to say that quantum computers will hold a key role for the industry’ biggest players, and will lie at the heart of R&D in the chemical industry of tomorrow.