Barborini Matteo, Guidoni Leonardo
J Chem Theory Comput. 2015 Sep 8;11(9):4109-18. doi: 10.1021/acs.jctc.5b00427.
Due to the crucial role played by electron correlation, the accurate determination of ground state geometries of π-conjugated molecules is still a challenge for many quantum chemistry methods. Because of the high parallelism of the algorithms and their explicit treatment of electron correlation effects, Quantum Monte Carlo calculations can offer an accurate and reliable description of the electronic states and of the geometries of such systems, competing with traditional quantum chemistry approaches. Here, we report the structural properties of polyacetylene chains H-(C₂H₂)(N)-H up to N = 12 acetylene units, by means of Variational Monte Carlo (VMC) calculations based on the multi-determinant Jastrow Antisymmetrized Geminal Power (JAGP) wave function. This compact ansatz can provide for such systems an accurate description of the dynamical electronic correlation as recently detailed for the 1,3-butadiene molecule [J. Chem. Theory Comput. 2015 11 (2), 508-517]. The calculated Bond Length Alternation (BLA), namely the difference between the single and double carbon bonds, extrapolates, for N → ∞, to a value of 0.0910(7) Å, compatible with the experimental data. An accurate analysis was able to distinguish between the influence of the multi-determinantal AGP expansion and of the Jastrow factor on the geometrical properties of the fragments. Our size-extensive and self-interaction-free results provide new and accurate ab initio references for the structures of the ground state of polyenes.
由于电子关联所起的关键作用,对于许多量子化学方法而言,准确确定π共轭分子的基态几何结构仍是一项挑战。由于算法具有高度并行性且能明确处理电子关联效应,量子蒙特卡罗计算能够提供对此类系统电子态和几何结构的准确且可靠的描述,可与传统量子化学方法相媲美。在此,我们通过基于多行列式贾斯特罗反对称双电子幂(JAGP)波函数的变分蒙特卡罗(VMC)计算,报告了多达N = 12个乙炔单元的聚乙炔链H-(C₂H₂)(N)-H的结构性质。这种简洁的近似方法能够为这类系统提供动态电子关联的准确描述,正如最近针对1,3 - 丁二烯分子所详细阐述的那样[《化学理论与计算杂志》2015年第11卷第2期,508 - 517页]。计算得到的键长交替(BLA),即单碳键和双碳键之间的差值,对于N → ∞时外推至0.0910(7) Å的值,与实验数据相符。精确分析能够区分多行列式AGP展开和贾斯特罗因子对片段几何性质的影响。我们的尺寸扩展性和无自相互作用结果为多烯基态结构提供了新的精确从头算参考。