In this blog post, we’ll explore the research methods and applications of computational chemistry, a field that has grown significantly alongside advances in computer technology.
Computational chemistry is a branch of chemistry that studies the composition, properties, structure, and changes in substances. Generally, when people think of chemical research, they tend to imagine synthesis, analysis, and measurement experiments conducted in laboratories. However, as theoretical research methods have systematically advanced and computer performance has improved dramatically, the field of computational chemistry—which can complement or partially replace traditional experiments—has developed significantly. The greatest advantage of computational chemistry is its ability to predict results before actual experiments are performed. This allows researchers to review necessary experimental conditions in advance and even design molecular structures beforehand to obtain the desired substances. For these reasons, computational chemistry is utilized as a key research method in various fields, including new drug development, new material design, and catalyst research. In this article, we will briefly introduce representative research methods in computational chemistry used for molecular structure design.
Research methods in computational chemistry can be broadly categorized into two types: quantum chemistry methods and classical mechanics methods. Among these, quantum chemistry methods—which are theoretically significant—are further subdivided into purely theoretical quantum chemistry methods, semi-empirical quantum chemistry methods, and empirical quantum chemistry methods.
Purely theoretical quantum chemistry methods calculate all chemical properties of a target substance using only basic information such as fundamental physical laws, physical constants, and particle masses. Although they offer the highest accuracy, they have the limitation of being difficult to apply to large molecules because they require a vast amount of memory and storage space, as well as long computation times. The Hartree-Fock method is the representative foundation of this approach, and further developed methods include Configuration Interaction (CI) and Møller-Plesset Perturbation Theory (MP).
The Hartree-Fock method is a technique for approximately calculating the wavefunction and energy of the most stable state in a multi-electron system. It treats the influence each electron receives from other electrons as an average field and, through iterative calculations, gradually converges to a value close to the actual solution. Due to this characteristic, it is also called the Self-Consistent Field (SCF) method.
Semi-empirical quantum chemistry methods are computational approaches that replace certain terms—which are computationally very intensive in purely theoretical methods—with experimental values. These methods began to develop in earnest in the late 1960s and are still used today for calculating molecules composed of a relatively large number of atoms. Most of these methods are based on the Hartree-Fock method, with certain terms approximated using experimental values; representative examples include Extended Hückel Theory (EHT).
Empirical quantum chemistry methods derive only the basic equations from quantum chemistry theory and rely primarily on experimental values for the remaining terms. Although widely used in the early days when computer performance was limited, they are rarely used in research today as more accurate computational methods have become widespread. A representative example is Hückel Theory (HT).
Hückel Theory is a method that simplifies the Hamiltonian required for quantum chemical calculations, solves the resulting matrix equations, and derives the desired values. Extended Hückel Theory (EHT) is an extension of this method.
Classical mechanics methods are based on Newton’s classical mechanics. Calculations are performed by assuming that atoms are spheres and chemical bonds behave like springs. The parameters representing the characteristics of chemical bonds are determined based on experimental results, and it is assumed that these parameters can be applied identically regardless of the specific molecule. Thanks to these characteristics, the computational load is much lower than that of quantum chemistry methods, allowing for efficient calculations even for proteins or polymers composed of thousands or more atoms. Representative methods include Molecular Mechanics and AMBER (Assisted Model Building with Energy Refinement).
The Molecular Mechanics method calculates the stability of an atomic arrangement and the energy differences within a molecule as the sum of the potential energies resulting from interactions between individual atoms. While it is highly effective for calculating the structures of organic compounds, it has the limitation that its accuracy may be lower for substances whose bonding patterns differ significantly from those of organic compounds.
The AMBER method shares the same basic principles as molecular mechanics but employs a technique that groups certain atoms into a single unit to efficiently calculate macromolecules such as proteins and nucleic acids. Consequently, while it is widely used in structural studies of biomolecules, it may not always be suitable for precise calculations of small molecules.
We have reviewed the representative research methods in computational chemistry thus far. While computational chemistry was initially used primarily to analyze and predict molecular structures, its scope of application has now greatly expanded to include spectroscopic analysis, catalyst design, new drug development, new material development, battery material research, and semiconductor material research. Furthermore, with advances in artificial intelligence (AI), high-performance computing (HPC), and quantum computing technologies, the accuracy and scope of these calculations continue to improve, and computational chemistry is expected to play an even more important role in chemical research in the future.
