JP7302621B2 - Organic material transport property evaluation device and organic material transport property evaluation method - Google Patents

Description

TECHNICAL FIELD The present invention relates to an organic material transport property evaluation apparatus and an organic material transport property evaluation method.

Equilibrium molecular dynamics (EMD) calculations using the Green-Kubo (GK) formula and non-equilibrium molecular dynamics for transport properties that indicate temporal changes in physical quantities such as thermal conductivity, viscosity, and diffusion coefficient for organic materials (NEMD) calculations have been performed using direct calculation methods.

A technique for performing high-speed equilibrium molecular dynamics (EMD) calculations using the Green-Kubo (GK) formula has been proposed. For example, a method has been proposed in which the amount of calculation required to obtain the required accuracy is determined by error analysis (Non-Patent Document 1). In addition, in the molecular dynamics (EMD) calculation using the Green-Kubo (GK) formula, the heat flow autocorrelation function (HFACF) is modeled with a random walk, and harmonic analysis is performed to separate the frequency components that cause errors. proposed a method of obtaining the standard deviation of calculation errors (Non-Patent Document 2). A method based on model accuracy evaluation and harmonic analysis using the Akaike information criterion has also been proposed (Non-Patent Document 3). Also, in the Green-Kubo (GK) formula, a method of analyzing an inhomogeneous system by spatially dividing the calculation is proposed (Patent Document 1). Also, load reduction in molecular dynamics (EMD) calculations using adversarial generation networks has been proposed (Patent Document 2).

JP 2019-148446 A JP 2020-87456 A

R. E. Jones, J. Chem. Phys. 136, 154102 (2012) L. Ercole, Scientific Reports, 7, 15835 (2017) L. S. Oliveira, Phys. Rev. E 95, 023308 (2017)

By the way, in the conventional techniques in general, a parameter that serves as a threshold value is required in error separation by harmonic analysis. Therefore, pretreatment such as prior examination and parameter setting is required, and a simpler apparatus and method are desired for automating the material search.

Although the technique described in Non-Patent Document 1 can obtain the minimum amount of calculation for obtaining the desired accuracy, it does not lead to a substantial reduction in error. Also, analysis for harmonic analysis and separation of errors is necessary. In addition, the technique described in Non-Patent Document 2 requires analysis such as error separation by harmonic analysis, and is not suitable for automatic calculation search. Further, the technique described in Patent Document 1 does not describe a process for speeding up the evaluation of the transport properties of organic materials. In addition, the technique described in Patent Document 2 requires prior learning because it uses machine learning, and in many cases the calculation speed is low, making it unsuitable for automatic search of various materials.

One aspect of the present invention is an organic material transport property evaluation apparatus for evaluating the transport property of an organic material, wherein the transport property is calculated by removing the correlation of the intermolecular flux of the organic material with respect to the transport property. It is an organic material transport property evaluation device characterized by

Here, when calculating the transport properties using the Green-Kubo formula, it is preferable to calculate the transport properties by removing the correlation of the intermolecular flux of the organic material.

It is also preferable to use the Green-Kubo formula to calculate the thermal conductivity of a linear polymer chain material having an amorphous structure, excluding intermolecular heat flow correlation.

It is also preferable to use the Green-Kubo formula to calculate the viscosity of organic materials, excluding intermolecular stress correlations.

It is also preferred to use the Green-Kubo formula to calculate the diffusion coefficients of organic materials, excluding intermolecular velocity correlations.

Another aspect of the present invention is an organic material transport property evaluation method for evaluating the transport property of an organic material, wherein the transport property is calculated by removing the correlation of the intermolecular flux of the organic material with respect to the transport property. It is an organic material transport property evaluation method characterized by

ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to evaluate the transport property of an organic material at high speed.

BRIEF DESCRIPTION OF THE DRAWINGS It is a figure which shows the structure of the organic material transport characteristic evaluation apparatus in embodiment of this invention. FIG. 4 is a diagram showing evaluation results of a heat flow correlation function for polyethylene. It is a figure which shows the evaluation result of the thermal conductivity with respect to polyethylene. FIG. 4 is a diagram showing evaluation results of thermal conductivity with respect to polystyrene; It is a figure which shows the evaluation result of thermal conductivity. FIG. 2 is a diagram showing the molecular structure of a monomer of a typical linear polymer material; FIG. 5 is a diagram showing evaluation results of time correlation of stress fluctuations in a single-component system; It is a figure which shows the evaluation result of the time correlation of the stress fluctuation in a binary system. FIG. 4 is a diagram showing evaluation results of diffusion coefficients for each component according to the mixing ratio in a mixed system;

An organic material transport property evaluation apparatus 100 according to an embodiment of the present invention includes a processing unit 10, a storage unit 12, an input unit 14, an output unit 16 and a communication unit 18, as shown in FIG. The processing unit 10 includes means for performing arithmetic processing such as a CPU. The processing unit 10 executes the organic material transport property evaluation program stored in the storage unit 12 to implement the functions of the organic material transport property evaluation method according to the present embodiment. The storage unit 12 includes storage means such as a semiconductor memory and a memory card. The storage unit 12 is connected to the processing unit 10 so as to be accessible, and stores an organic material transport property evaluation program and information necessary for the processing. The input unit 14 includes means for inputting information. The input unit 14 includes, for example, a keyboard, a touch panel, buttons, and the like for receiving input from the user. The output unit 16 includes means for outputting the processing results of the organic material transport property evaluation apparatus 100, such as a user interface screen (UI) for accepting input information from the user. The output unit 16 includes, for example, a display that presents images to the user. The communication unit 18 includes an interface for communicating information with an external information processing device via the information communication network 102 . Communication by the communication unit 18 may be wired or wireless.

Various devices can be applied to the organic material transport property evaluation device 100 as long as they are information processing devices capable of executing an organic material transport property evaluation program. For example, as the organic material transport property evaluation device 100, a stationary or portable personal computer (PC), a tablet computer, or the like can be used.

A method for evaluating transport characteristics of an organic material according to the present embodiment will be described below. The organic material transport property evaluation method is realized by executing an organic material transport property evaluation program for executing the following processes in the organic material transport property evaluation apparatus 100. FIG.

The transport characteristic K in a fluid is represented by the corresponding physical quantity Q(t) and its time derivative flux ∂Q(t)/∂t (Q(t) is indicated by a dot in the formula) as follows: be done.

[Evaluation of thermal conductivity]
When the thermal conductivity is evaluated as the transport property of the organic material, the thermal conductivity κ is obtained as shown in Equation (3) by applying the flux represented by Equation (2) to Equation (1). where kB is the Boltzmann constant, _ei is the energy of the i-th atom, v _i is the velocity of the i-th atom and _Si is the stress of the i-th atom.

The heat flow J(t) of the system is expressed as the sum of the per-atom contributions Ji. By replacing this sum with Jmol, which is the sum of atoms belonging to the same molecule, and ignoring the correlation between atoms belonging to different molecules, the thermal conductivity κ in equation (3) is thermally It can be expressed as an approximation of the conductivity κ. Here, the dashed Σ symbol in Equation (4) means an operator that takes the sum of each molecule.

By ignoring the intermolecular heat flow correlations, which contribute little to the thermal conductivity κ of the polymer, the random errors involved are removed. This reduces the statistical error in the organic material transport property evaluation, eliminates the need to calculate the average for a plurality of initial configurations, and can greatly reduce the computational load in the evaluation.

[Example 1]
The contribution of the correlation function of heat flow within the same molecule and between different molecules to the thermal conductivity calculated using the Green-Kubo formula was examined.

Organic material transport characteristics were evaluated by equilibrium molecular dynamics (EMD) calculations using the Green-Kubo (GK) formula for the thermal conductivity κ of linear polymer materials with an amorphous structure.

Molecular dynamics (MD) calculations were performed by creating an amorphous structure consisting of 50,000 atoms, where the length of one polymer molecule is about 120 Å and the number of atoms is N. The programs EMC (“Enhanced Monte Carlo (EMC),” http://montecarlo.sourceforge.net) and LAMMPS (S. Plimpton, “ Fast parallel algorithms for short-range molecular dynamics," Journal of Computational Physics 117, 1 (1995).) was used. OPLS-UA, which is a United Atom force field omitting hydrogen atoms, was used as the force field. A time step of 2.0 fs was used. For structural equilibration, Nose-Hoover heat bath (W. G. Hoover, "Canonical dynamics: Equilibrium phase-space distributions," Physical Review A 31, 1695 (1985).) and Parrinello-Rahman pressure bath (M Parrinello and A. Rahman, "Polymorphic transitions in single crystals: A new molecular dynamics method," Journal of Applied Physics 52, 7182 (1981). Hold and equilibrate. The Green-Kubo (GK) formula was used to evaluate the thermal conductivity κ. At this time, a heat flow of 500 ps was calculated at 300 K using an NVE ensemble with constant volume and energy. The integration time for HFACF expected value calculation was 460 ps, and the integration time in the GK formula was 40 ps.

FIG. 2 shows the calculated results of the heat flow (HF) correlation function for polyethylene (PE). FIG. 2(a) shows the autocorrelation function (ACF) of the heat flow within the same molecule. FIG. 2(b) shows the cross-correlation function (CCF) of heat flow between different molecules. In FIGS. 2(a) and 2(b), the calculation results in the period from 10 ² fs to 10 ⁴ fs are enlarged and shown as interpolated diagrams.

The thermal conductivity κ is obtained by integrating these sums over time. From FIG. 2(a), it was found that the autocorrelation function (ACF) of the intramolecular heat flow takes a relatively large value in a period of 10 ³ fs or less from the start of the evaluation, which greatly contributes to the thermal conductivity. On the other hand, from FIG. 2(b), the cross-correlation function (CCF) of the heat flow between different molecules is sufficiently higher than the auto-correlation function (ACF) of the heat flow within the molecule in the period of 10 ³ fs or less from the start of the evaluation. , and contains random errors in periods longer than 10 ³ fs.

From FIGS. 2(a) and 2(b), dividing the heat flow correlation function into the intramolecular autocorrelation function (ACF) and the intermolecular cross-correlation function (CCF) contributes mainly to the thermal conductivity Components and error components can be separated. In this embodiment, the thermal conductivity is obtained only from the intramolecular autocorrelation function (ACF) as the correlation function of the heat flow. This avoids the effects of intermolecular cross-correlation function (CCF) errors. On the other hand, in the conventional method, it is necessary to perform calculations for multiple initial structures and take the average.

FIG. 3 shows the results of calculating the thermal conductivity of polyethylene (PE) using the Green-Kubo (GK) formula. In FIG. 3, the horizontal axis indicates the cutoff time of time integration used for calculation of the Green-Kubo (GK) formula. FIG. 3(a) shows the average value of the thermal conductivity calculated for 20 initial configurations of polyethylene (PE). In FIG. 3(a), the thick solid line indicates the contribution of the autocorrelation function (ACF) within the same molecule to the thermal conductivity, and the dashed line indicates the contribution of the cross-correlation function (CCF) of the heat flow between different molecules to the thermal conductivity. show. In addition, in FIG. 3(a), the thin solid line indicates the sum of the contribution of the autocorrelation function (ACF) within the same molecule and the contribution of the cross-correlation function (CCF) of the heat flow between different molecules, which is the same as in the conventional method. Calculated thermal conductivity will be shown.

FIG. 3(b) shows the calculated results of the intramolecular autocorrelation function (ACF) contribution of thermal conductivity for each of the 20 initial configurations of polyethylene (PE). FIG. 3(c) shows the calculated cross-correlation function (CCF) contribution of heat flow between molecules with different thermal conductivities for each of the 20 initial configurations of polyethylene (PE).

In FIG. 3( a ), the calculated thermal conductivity (thin line) of polyethylene (PE) increased over a period of 2×10 ⁴ fs or more. This is due to statistical errors due to random errors in the cross-correlation function (CCF) of heat flow between different molecules to thermal conductivity. As indicated by the thick solid line in FIG. The thermal conductivity converged to a constant value for a period of 10 ⁴ fs or longer.

FIG. 4 shows the results of calculating the thermal conductivity of polystyrene (PS) using the Green-Kubo (GK) formula. In FIG. 4, the horizontal axis indicates the cutoff time of the time integration used for calculating the Green-Kubo (GK) formula. FIG. 4(a) shows the average value of the thermal conductivity calculated for 20 initial arrangements of polystyrene (PS). In FIG. 4(a), the thick solid line indicates the contribution of the autocorrelation function (ACF) within the same molecule to the thermal conductivity, and the dashed line indicates the contribution of the cross-correlation function (CCF) of the heat flow between different molecules to the thermal conductivity. show. In addition, in FIG. 4(a), the thin solid line indicates the sum of the contribution of the autocorrelation function (ACF) within the same molecule and the contribution of the cross-correlation function (CCF) of the heat flow between different molecules. Calculated thermal conductivity will be shown.

FIG. 4(b) shows the results of calculating the contribution of the intramolecular autocorrelation function (ACF) of the thermal conductivity for each of the 20 initial configurations of polystyrene (PS). FIG. 4(c) shows the calculated cross-correlation function (CCF) contribution of heat flow between molecules with different thermal conductivities for each of the 20 initial configurations of polystyrene (PS).

As shown in FIG. 4(a), in the calculation results of the thermal conductivity of polystyrene ( ^PS ), the autocorrelation function (ACF) within the same molecule with respect to the thermal conductivity indicated by the thick solid line is and the sum of the contribution of the autocorrelation function (ACF) within the same molecule and the contribution of the cross-correlation function (CCF) of heat flow between different molecules, indicated by the thin solid line, converged to a constant value. That is, in the calculation of the thermal conductivity of polystyrene (PS), the contribution of the cross-correlation function (CCF) of heat flow between different molecules was sufficiently small to be negligible.

Thus, by ignoring the cross-correlation function (CCF) of the heat flow between different molecules, the random errors contained therein are removed, and the heat conduction It was found that the degree can be evaluated with high accuracy. It was also found that the computational load can be significantly reduced.

FIG. 5 shows the molecular length dependence of thermal conductivity when connecting monomers of a typical linear polymer material. FIG. 5(a) shows the molecular length dependency of thermal conductivity calculated by the conventional method as a comparative example. FIG. 5(b) shows the molecular length dependence of thermal conductivity calculated ignoring the contribution of the cross-correlation function (CCF) of the heat flow between different molecules in the example. In FIG. 5, the horizontal axis indicates the molecular length (Å) used in the calculation. Also, in FIG. 5, circles indicate average values of thermal conductivities calculated with respect to initial configurations of 20 molecules. Also, in FIG. 5, the error bars indicate the standard deviation of the calculation results.

In addition, FIG. 6 shows the molecular structure of a monomer of a typical linear polymer material used for the calculation. In FIG. 6, asterisks represent the binding positions of the monomers.

As shown in FIG. 5(a), in the conventional method, the standard deviation (error bar) of the individual calculated values for the 20 initial arrangements increased. Therefore, the trend of thermal conductivity was obtained by calculating the average value after performing calculations for a plurality of initial configurations (circles). However, since the average value did not lie on the smooth curve, it is presumed that an error remains in the results shown in FIG. 5(a).

On the other hand, as shown in FIG. 5(b), there was little variation in individual calculated values for 20 initial arrangements to which the organic material transport property evaluation method of the example was applied. Therefore, it was possible to evaluate the thermal conductivity with sufficient accuracy even with individual calculated values for one initial configuration. In the example, the contribution of the cross-correlation function (CCF) of the heat flow between different molecules was ignored, so the thermal conductivity was slightly underestimated, but the order of the thermal conductivity of each linear polymer material was the same as that of the conventional method. The results were in good agreement. Therefore, the method for evaluating transport characteristics of organic materials in the example also had sufficient accuracy for automatic computational search. Moreover, in the example, the standard deviation was reduced to about 1/10 of the conventional method, that is, the statistical error was reduced to about 1/100.

[Time-stress fluctuation correlation function]
Viscosity η can be evaluated using the Green-Kubo formula. The off-diagonal component p ^ab (ab={xy, xz, yx}) of the system stress at a certain point in time can be expressed by Equation (5). where V is the volume of the system, N is the total number of atoms in the system, m _i is the mass of the i-th atom, v _i is the velocity of the i-th atom, r _ij is the The distance between, f _ij is the interaction between the i-th atom and the j-th atom.

Equation (6) is defined using the time correlation function for this stress. where kB is the Boltzmann constant and T is the temperature. Also, <...> means an ensemble averaging operation sampled at different initial times t0.

The time integral with respect to φ(t) can be represented by Equation (7), and the viscosity η can be obtained by Equation (8).

Numerical calculations cannot handle time t→∞ in formula (8), but by taking a long time t, Φ(t) converges to a finite value. can be evaluated.

However, since the convergence is slow and the numerical calculation results vary widely, a method of improving accuracy by performing statistical averaging by multiple sampling or fitting to a model formula has been conventionally adopted.

Organic material transport property evaluation apparatus 100 in the present embodiment utilizes the fact that shear elastic constant _G∞ is associated with formula (9) using formula (7).

The viscosity η is calculated using the Arrhenius equation of Equation (10), taking advantage of the fact that the statistical error of the quantity represented by Equation (9) and the calculation load are significantly small. where T _b is the boiling point of the system, α, α' are the parameters, and η ₀ is the viscosity at the boiling point.

Equation (5) can be rewritten as Equation (12) using Equation (11), which indicates the stress per atom. Here, Nmol is the number of molecules contained in the system, and p bar (a value with a bar above p) is the stress per molecule. The per-molecule stress p bar means the sum of the stresses of the atoms composing a certain molecule k.

When determining the shear elastic constant G∞ of Equation (9), the correlation part of Equation (6) can be expressed as Equation (13). In the approximation formula (13), the equal sign holds for a pure solvent of a single molecular species.

In this way, by converting the time correlation of the stress of the entire system into the time correlation of the stress of each molecule, it is possible to take a statistical average of not only the difference in the initial time _t0 but also the number of molecules. . Therefore, variations in numerical calculation can be suppressed as compared with the conventional method.

[Evaluation of Viscosity and Diffusion Coefficient]
Based on Equation (13), a method for evaluating viscosity and diffusion coefficient in multi-component systems is examined. In the following, a liquid two-component system consisting of component A and component B will be considered for simplicity of explanation.

Assuming that the molar ratio of component A and component B is r _A :r _B (where r _A +r _B =1), equation (13) is expressed by equations (14) and (15). Note that _t0 indicating the initial time is omitted below. In addition, the notation is omitted as in Expression (16).

The shear elastic constant _G∞ is expressed by Equations (17) and (18). As shown in Equation (18), the shear elastic constant _G∞ can be represented by the linear sum of the molar fraction of the component x and the shear elastic constant G∞ _,x .

Here, using the simplest Arrhenius equation (equation (19)) for viscosity prediction, the viscosity η is represented by equations (20) and (21).

This indicates that the viscosity η of the entire system can be evaluated using the ratio and viscosity of each component constituting the system. Furthermore, it agrees with the empirical formula for predicting mixed solvent viscosity (e.g., formula (5.28) in Viswanath, D. S. et al., “Viscosity of Liquids: Theory, Estimation, Experiment, and Data” (Springer, 2007). .

The relationship between the viscosity η and the diffusion coefficient D can be expressed as the Stokes-Einstein relationship of Equation (22). Here, R is the diffusion particle diameter, and λ is theoretically 6 (Stick condition) or 4 (Slip condition). The value of λ is determined by the relationship between the diffusing particles and the solvent, and may not be a constant experimentally.

As shown by Equation (22), the diffusion coefficient D and the viscosity η are in an inversely proportional relationship. The diffusion coefficient D is further represented by Equation (23). where c is the constant of proportionality. Also, Equation (21) was used.

Equation (23) means that the diffusion coefficient D of the entire system is expressed using the viscosity ηx of the component x constituting the system and the component ratio. Therefore, when generalized to the M component system, the diffusion coefficient Dx of the x component is expressed by Equation (24), and the diffusion coefficient D of the entire system is expressed by Equation (25).

[Example 2]
In this example, it was confirmed whether the sum expression of the molecular stress correlation (equation (13)) was the same as the total stress correlation (equation (6)) in a single-component liquid.

Molecular dynamics (MD) calculations were performed using the LAMMPS program (version 5 Jun 2019) (Plimpton, S., "Fast Parallel Algorithms for Short-Range Molecular Dynamics", J. Comput. Phys. 117, 1-19 (1995), http://lammps.sandia.gov.). Molecular force field parameters are obtained from GAFF (Wang, J. et al., “Development and testing of a general amber force field”, J. Comput. Chem. 25, 1157-1174 (2004). doi:10.1002/jcc.20035 ), and the charge is determined by the RESP method (Bayly, C. I. et al., “A well-behaved electrostatic potential based method using charge restraints for deriving atomic charges: the RESP model”, J. Phys. Chem. 97, 10269- 10280) using the Antechamber program (Case, D. A. et al., “AMBER 2018”, University of California, San Francisco, 2018.). HF/6-31G//B3LYP/6-31++G** according to the Gaussian09 program (Frisch, M.J. et al., "Gaussian 09, Revision E.01", Gaussian Inc. Wallingford CT.) was used to calculate the RESP charge. I used the result of the level.

For the initial structure, the Packmol program (Martinez, L. et al., "PACKMOL: A package for building initial configurations for molecular dynamics simulations", J. Comput. Chem. 30, 2157-2164 (2009). doi:10.1002/ jcc.21224.), cubic cells filled with constituent molecules in random arrangement were used. The cell size was set to about 1 g/cm ³ . The number of molecules in the cell was adjusted to 600 in accordance with the molar ratio.

In the molecular dynamics (MD) calculation, first, after energy minimization, initial structural relaxation (NVE and NVT ensemble) is performed at a time step of 0.25 fs, and then with a time step of 1.0 fs and an NPT ensemble under 1 atm. The volume was also relaxed to create an equilibrium structure. Based on the obtained equilibrium structure, this calculation was performed for 1 ns (1,000,000 steps) with an NVT ensemble with a time step of 1.0 fs, and used for analysis. The temperature was room temperature. The PPPM method (accuracy parameter of 10 ⁻⁴ ) was used for calculation of long-range interactions, and the interaction cut-off value was 10 Å.

A compute stress/atom command and a compute reduce/chunk command of the LAMMPS program were used to calculate the stress for each atom/molecule. The former is a function for calculating the stress for each atom, and the latter is a function for summing and outputting a set group (in this case, on a per-molecule basis).

FIG. 7 shows the results of plotting stress versus time for polycarbonate (PC). FIG. 7(a) shows the calculation results in the short-time domain, and FIG. 7(b) shows the calculation results in the long-time domain. In FIGS. 7(a) and 7(b), the thick solid line indicates the total stress correlation, and the thin solid line indicates the stress correlation for each molecule (value obtained by subtracting -0.05 from the calculation result in the figure). FIG. 7(c) shows the difference between the total stress correlation and the stress correlation for each molecule in FIG. 7(b).

As shown in FIGS. 7( a ) and 7 ( b ), the oscillation behaviors of the total stress correlation and the sum of molecular stress correlations generally agreed well in both the short-term region and the long-term region. However, as shown in FIG. 7(c), there was a slight difference between the total stress correlation and the molecular stress correlation sum. The difference between the total stress correlation and the stress correlation for each molecule at time t=0 was -0.0009 [mPas/fs]. As the correlation time progressed, the error gradually increased, reaching a maximum around time t=30 and then decreasing, but the oscillation continued with a constant width around t=0. The error is presumed to be due to the approximation used in the examples, that is, due to ignoring the cross-correlation between molecules. However, in the organic material transport property evaluation method in this example, information in a very short correlation time (10 fs or less) is important, and the error in this time is relatively small, so it is thought that it can be used to evaluate the diffusion coefficient. be done.

[Example 3]
In this example, the stress in a two-component solution system was evaluated. FIG. 8 shows the time correlation of stress fluctuations for a 50%:50% mixed system of polycarbonate (PC) and 1,2-dimethoxyethane (DME). FIG. 8(a) shows the calculation results in the short-time domain, and FIG. 8(b) shows the calculation results in the long-time domain. In FIGS. 8(a) and 8(b), the thick solid line indicates the total stress correlation, and the thin solid line indicates the stress correlation for each molecule (value obtained by subtracting -0.05 from the calculation result in the figure). Furthermore, in FIG. 8(a), the thin dashed line indicates the sum of the components of polycarbonate (PC) only, and the thin dashed line indicates the sum of the components of 1,2-dimethoxyethane (DME) only. Also, FIG. 8(c) shows the difference between the total stress correlation and the stress correlation for each molecule in FIG. 8(b).

Similar to the discussion of the single-component system shown in FIG. 7, the summation of the molecular stress correlations in the two-component system also agreed well with the total stress correlation in both the short-term region and the long-term region. In the difference between the two shown in FIG. 8(c), the difference between the total stress correlation and the stress correlation for each molecule at time t=0 was −0.0012 [mPas/fs]. The difference between the two showed a behavior similar to that of the single-component system, although there was a slight difference in magnitude and amplitude.

In the calculation results for each component, it was found that there is a difference between the components of polycarbonate (PC) and 1,2-dimethoxyethane (DME) (the coincidence at time t=0 is a coincidence). The behavior of the polycarbonate (PC) component corresponds to just half the amplitude of the stress fluctuation in a single component due to the 50% molar ratio. That is, even if the types of solvent molecules are different, not only can the stress correlation of the entire system be shown by summing the contributions of individual molecules, but also the component contribution can be evaluated by summing up for each solvent type.

[Example 4]
In this example, the evaluation of the diffusion coefficient D in a multi-component system was verified. The diffusion coefficient D of each component in a multi-component system can be tested by the NMR measurement method, and there are many reported examples. The systems to be verified in this embodiment are a two-component system of polycarbonate (PC) and 1,2-dimethoxyethane (DME) and a two-component system of polycarbonate (PC) and diethyl carbonate (DEC).

When calculating the diffusion coefficient D for each component, the conventional method used the total stress correlation considering the intermolecular stress correlation, and the embodiment used the stress correlation for each molecule while ignoring the intermolecular stress correlation. In the conventional method, a general Einstein method (slope of mean square displacement) was used as MD analysis. The calculation conditions were the same as in the example except that the number of sampling steps used in the analysis was significantly longer (20 times) than in the example.

FIG. 9(a) shows the diffusion coefficient D of each component for a binary system of polycarbonate (PC) and 1,2-dimethoxyethane (DME). FIG. 9(b) shows the diffusion coefficient for each component for a binary system of polycarbonate (PC) and diethyl carbonate (DEC). 9(a) and 9(b) show the diffusion coefficient D for each component with respect to the composition ratio of 1,2-dimethoxyethane (DME) or diethyl carbonate (DEC) to polycarbonate (PC). Black squares and thick solid lines are experimental values (Kikuko Hayami, “3. NMR Study of Structure and Transport Phenomenon of Electrolytes for Lithium Batteries”, Electrochemistry 81, 995-1000 (2013). doi:10.5796/electrochemistry.81.995 NMR measurement values according to the present invention), white triangles and dashed lines indicate calculation results by the conventional method, and white circles and dashed lines indicate calculation results by this embodiment.

In pure solvents, the diffusion coefficients D are known to be polycarbonate (PC)<1,2-dimethoxyethane (DME) and polycarbonate (PC)<diethyl carbonate (DEC). However, as shown in FIG. 9, when the mixed solvent is used, the diffusion coefficient D of polycarbonate (PC), which was low when used alone, increased, and conversely, 1,2-dimethoxyethane (DME) and diethyl carbonate (DEC), which were high when used alone, increased. ) has been reported to be lower and correlated with concentration. In addition, the tendency differs depending on the type of solvent, and in the case of polycarbonate (PC)-1,2-dimethoxyethane (DME), there is a difference in the diffusion coefficient D of each component, but polycarbonate (PC)-diethyl carbonate (DEC ) system, the diffusion coefficients of each component show close values.

As shown in FIGS. 9(a) and 9(b), the diffusion coefficient D of each component calculated by the conventional method shows a trend of change according to the solvent concentration of polycarbonate (PC)-1,2-dimethoxy Both the ethane (DME) system and the polycarbonate (PC)-diethyl carbonate (DEC) system were reproduced. However, the diffusion coefficient D of each component calculated by the conventional method was quantitatively significantly underestimated relative to the experimental value.

On the other hand, in this example, when polycarbonate (PC) is 0% and 1,2-dimethoxyethane (DME) or diethyl carbonate (DEC) is used as a pure solvent, the diffusion coefficient D of each component agrees with the experimental value. . On the other hand, the evaluation accuracy of the diffusion coefficient D of polycarbonate (PC) slightly decreased. Therefore, in the polycarbonate (PC)-1,2-dimethoxyethane (DME) system, the diffusion coefficient D for the component of 1,2-dimethoxyethane (DME) showed good agreement with the concentration change, but the polycarbonate (PC) ) was generally underestimated. As mentioned above, in the polycarbonate (PC)-diethyl carbonate (DEC) system, the diffusion coefficients D of the two solvents tend to match, and both the polycarbonate (PC) component and the diethyl carbonate (DEC) component agree with the experimental values. showed a slightly lower trend. However, compared with the conventional method, quantification with respect to experimental values was greatly improved. Furthermore, in the example, although the calculation cost was about 1/20 that of the conventional method, the variation in the calculated values could be calculated with an accuracy of 1/5 to 1/10.

10 processing unit, 12 storage unit, 14 input unit, 16 output unit, 18 communication unit, 100 organic material transport property evaluation device.

Claims (5)

An organic material transport property evaluation device for evaluating the transport property of an organic material,
the computer,
The transport property K of the fluid in the organic material is

is represented by
The flux [Q(t) dot], which is the time derivative of the physical quantity Q(t) corresponding to the transport characteristic K, is used as a means for calculating the transport characteristic K by removing the correlation of the flux between the molecules of the organic material. An organic material transport property evaluation device characterized by: The organic material transport property evaluation device according to claim 1 ,
An organic material transport characteristic evaluation apparatus for calculating the thermal conductivity of a linear polymer chain material having an amorphous structure by removing the heat flow correlation between molecules of the organic material. The organic material transport property evaluation device according to claim 1 ,
An organic material transport characteristic evaluation apparatus, wherein the viscosity of the organic material is calculated by removing the stress correlation between the molecules of the organic material. The organic material transport property evaluation device according to claim 1 ,
An organic material transport characteristic evaluation apparatus, wherein the diffusion coefficient of the organic material is calculated by removing the velocity correlation between the molecules of the organic material. An organic material transport property evaluation method for evaluating the transport properties of an organic material, comprising:
to the computer,
The transport property K of the fluid in the organic material is

is represented by
The transport characteristic K is calculated by removing the correlation of the intermolecular flux of the organic material in the flux [Q(t) dot] that is the time derivative of the physical quantity Q(t) corresponding to the transport characteristic K. Organic material transport property evaluation method.

Images