Page 40 - 23_EETE_03

40 EE|Times EUROPE

EVs
Ford Enlists Quantum Computing in EV Battery

Materials Hunt

By Stefano Lovati
C an quantum computers be used to chemically simulate new That is due to the large size of the problem space, which exponentially

materials and make future electric-vehicle batteries safer,
increases with the system size. Quantum computers can overcome
more energy-dense and easier to recycle?
this issue because they offer massive computational power that can be
Ford Motor Co.’s quantum research group recently released
the results of a study, conducted with quantum computing company scaled exponentially.
The use of quantum computers to explore the properties of new
Quantinuum, in which the team investigated the use of quantum com- materials promises obvious advantages. Compared with performing
puters to model materials destined for next-generation EV batteries. physical experiments, tools like AI and machine learning have already
sped up the research process for developing novel materials, but quan-
FROM SIMULATION TO QUANTUM COMPUTING tum computing has the potential to save substantially more time. And
Lithium-ion batteries are chemical energy storage devices that are compared with traditional computers, quantum computers can manip-
currently the primary source of energy for EVs. Given the rapid growth ulate data on a far larger scale, solving problems that simply can’t be
in EV adoption worldwide (see Figure 1), the hunt is on for the energy solved on a conventional computer.
sources that will power the next generation of vehicles. Further For the Ford team, the hope is that using quantum computers to find
advancements in battery chemistry are required not only to increase improved materials will speed the development of EV batteries with
battery range but to do so while delivering all the performance, com- increased power, quicker charging times and longer life.
fort and user-experience features available in gas-powered vehicles. In their work, Farag and Ghosh researched Li-ion battery chemistry
Additionally, different battery chemistry types will be required to strike using quantum computers. More specifically, the two scientists used
a balance between the anticipated demand and supply of essential the variational quantum eigensolver (VQE), an algorithm for finding
battery components. the ground-state energy (or the normal atomic energy state) of LiCoO 2 ,
EV manufacturers have recognized the need to improve battery a candidate transition metal oxide used for battery cathodes. The VQE
materials’ density, power density, life cycle, safety, cost and, most hybrid quantum-classical method is used on current-generation quan-
crucially, recyclability if the industry is to advance the state of battery tum computers to solve only those portions of a molecular system that
technology. Insights on charge/discharge mechanisms, electrochemical benefit the most from the quantum computation, with the remaining
and thermal stability, structural phase transition and surface behavior calculations performed on a classical computer.
can be gained using computational chemistry, which is therefore essen- The scientists simulated the Li 2 Co 2 O 4 and Co 2 O 4 gas-phase models
tial for identifying suitable materials that can improve the performance (see Figure 2), which reflect the charge and discharge of the battery,
and robustness of batteries. using the VQE technique. Consistent with the VQE hybrid
Ford’s quantum researchers are seeking new ways to simulate the quantum-classical approach, the quantum computer was used to solve
chemistry of Li-ion batteries. The group of scientists, known as the only those parts of the molecule simulation that would benefit the
Core AI-ML-QC team, is led by Devesh Upadhyay and includes most from its unique characteristics. Everything else was handled by
Marwa H. Farag, a quantum computer scientist, theoretical chemist and computers based on traditional architecture.
computational modeling expert, and Joydip Ghosh, a physicist. Farag The team tried three methods with VQE, as this was a proof of con-
and Ghosh are the authors of a new scientific paper describing a quan- cept for quantum computing:
tum computing (QC)-based approach to complex chemical modeling. 1 • Unitary coupled-cluster singles and doubles (UCCSD)
Today, most sophisticated conventional computers are unable to per- • Unitary coupled-cluster generalized singles and doubles (UCCGSD)
form extremely accurate simulations of complex, real-world molecules. • k-unitary pair coupled-cluster generalized singles and doubles
(k-UpCCGSD)
The researchers contrasted the quantitative outcomes as well as
the quantum resources required for accurate calculation with classical
wavefunction-based methods. The team discovered that the results
from the VQE approaches agreed with those obtained using tradi-
tional methods, such as coupled-cluster singles and doubles (CCSD)
and complete active space configuration interaction (CASCI), and that
k-UpCCGSD yielded similar results to UCCSD, at a reduced cost.
Even though all calculations were carried out on a state vector
simulator with 20 qubits, the researchers suggested that a 400-qubit
quantum computer (a near-term target for many companies) will be
required to simulate strongly correlated systems of large size, providing
more insights as the quantum hardware matures.

QUANTINUUM’S INQUANTO
Quantinuum’s InQuanto quantum chemistry platform and its
H-series ion-trap quantum hardware were used to implement the
Figure 1: Future battery-EV and plug-in hybrid-vehicle research team’s hybrid approach (VQE algorithm executed on both
projections, stated policy and sustainable development scenarios quantum and conventional computers), applied to the molecules that
(Source: Quantinuum) are directly relevant to battery research.

MARCH 2023 | www.eetimes.eu