# Young-Woo Son @ KIAS
#
# input gnuplot-file for fitting vol(a.u.). vs. energy(a.u.) into
# Birch's equation upto second order. Input & output B0 is given in Gpa
#
# Inclusion of fourth order correction should improve B0,B0P.
# I've found from the example of iron that the fourth order correction
# gives almost equal result as Murnaghan's equation, but third order only
# is slightly wrong. Note that the form of second and third order terms are
# invariant under inclusion of higher order corrections.
E(V) = E0 + 9.0/8.0*B0*V0*((V0/V)**(2.0/3.0)-1)**2 + 9.0/16.0*B0*(B0P-4)* \
V0*((V0/V)**(2.0/3.0)-1.0)**3.0 + R4*((V0/V)**(2.0/3.0)-1.0)**4.0
B0P = 2.60567 ; B0 = 0.111821
V0 = 167.834 ; E0 = 0.1 ; R4 = -73.033
set term postscript eps enhanced 10
set output 'Ca_bondinglength.eps'
set encoding iso_8859_1
set size 0.35,0.4
set xr[150:200]
set yr[0.0:0.24]
set xtics 150, 10, 200
set ytics 0.0,0.04,0.4
set format y "%.2f"
#set mytics 2
set xtics nomirror
set nokey
set title "Ca bulk (Ultra Soft PP, PW91)\n Birch eq. fit: a=5.516{\305}, B_0=17.9 GPa" font 'Helvetica'
set xlabel 'Volume({\305}^3)
set ylabel 'E_{total} (eV)'
fit E(x) 'Etot.dat' u 1:($2*13.6058) via B0P,B0,V0,E0,R4
plot E(x),'Etot.dat' u 1:($2*13.6058) w p pt 7 ps 1
# converting into GPa (energy = eV, length = Ang)
# B0(GPa) = B0(eV/Ang^3) *1.6022*100.0
## In this example, cell volume in Ang^3 and energy is in Ryd (thus *13.6058 is necessary.)
## It is higly recommended to use V0 and E0 at the minimum position.
<fig.log> file
...
Final set of parameters Asymptotic Standard Error
======================= ==========================
B0P = 2.60567 +/- 0.1231 (4.725%)
B0 = 0.111821 +/- 0.0005028 (0.4497%)
V0 = 167.834 +/- 0.02815 (0.01677%)
E0 = 0.0213131 +/- 0.0001264 (0.5932%)
R4 = -73.033 +/- 17.73 (24.28%)
correlation matrix of the fit parameters:
B0P B0 V0 E0 R4
B0P 1.000
B0 -0.489 1.000
V0 -0.922 0.476 1.000
E0 0.108 -0.704 -0.137 1.000
R4 0.829 -0.863 -0.735 0.422 1.000