Studies On The Density Functional Theory And Computation For The Fe-containing Multiply Spin Systems

The theory. and computation of Density functional theory (DFT)have been discussed by studying several Fe-containing systems onGaussian03 package. The spin of translate metal is usually multiplybecause of the active valence electrons of them, but the computationperformance of DFT for the unrestricted system is not perfect. Therefore,we took the Fe-containing systems as our objects to study thecomputational and theoretical behavior of different kinds of DFT, such asLDA, GGA/meta-GGA, Hybrid forms. The main jobs are listed asfollows:(1) Effective simulation of heine-containing biological systemsThe theoretical simulations of the biology system are challenging forcomputational chemistry because of its complexity. QM/MM is widelyused in simulating the large biology system of interest, but for the smallbiology system, such as heme, which plays a vital role in biology systems,quantum calculations are proposed to ensure the accurate, which requirescorrect calculation method and basis set. In order to select correctly thetwo prerequisites of simulating the heme: computation method and thebasis set, we studied the behaviors of a series of DFT forms forforecasting the energy ordering of heme, and contrasted the calculationresults of heme's properties from different basis sets, from which wepropound a suggestion to simulate heme effectively.The results showed that GGA and meta-GGA formulas bias thelower spin state by producing levitated frontier orbitals and often fail togive a correct description of the ground state. A few composite basis setshave been proposed. The economic basis sets of double and triple zeta quality with polarization functions added only to transition metal andelectronegative atoms are able to predict accurate geometrical structures,electronic properties, optical spectra, and DFT reactivity indices,providing significant reduction in the total number of basis functions andtremendous gain in computational efficiency.Our current proposal of selecting density functionals and basis setscan be regarded as a general guideline in modeling heme-containingsystems and is likely to be valid for other transition metal containingsystems of biological interest as well.(2) Towards understanding performance differences betweenapproximate DFT for spin states of iron complexes.Iron ion is multiply spin and easy for assorting ligands. Fe(â…¡) andFe(â…¢) with 3 ligands, ion ligand Cl^-, molecular ligand NH₃, and organicligand pyridine (py) were chosen as our research objects. In this study, wetook not only the multiply spin of Fe(â…¡) and Fe(â…¢), but also thesymmetry of ligand compounds and the influence of ligands selves intoaccount. The behaviors of theory and computation of various DFT werestudied by comparing the results of total energies and the front orbitalenergies, for which an appropriate offer to choose calculation method andbasis set was also proposed.The proper way of dealing with multiplets in DFT should be througheither an ensemble based approach or the yet-unknown symmetrydependent exchange-correlation energy density functional. If approximateground-state exchange-correlation energy density formulas under theLDA, GGA, and hybrid schemes are employed, unreasonable results canbe obtained. Confirming that LDA/GGA formulas often, but not always,bias the lower spin state energetically, the present work demonstrates that it is the outermost singly occupied molecular orbital that distinguishes theperformance difference of different categories of approximate functionalsfrom each other. It is found that LDA and GGA functionals unanimouslyproduce a significantly energetically higher and spatially more compactSOMO for the lower spin state.The implication of the present work to the DFT study ofiron-containing biological and other spinrelated systems is that great careis needed when a functional form is picked and that hybrid functionals arerecommended for dealing with multiple spin states. If the lower spin stateof the system is the ground state, an LDA or GGA functional will likelyproduce unreliable results, leading to unreasonable or even incorrectconclusions.