second force constants from ALAMODE (from XML reconstruction) do not satisfy ASR in Ti–O system (diagonal error ≈ 0.2 eV/Ų)

[Li-KaiHua]

Dear Dr. Tadano,
I’m currently studying the anharmonic effects in a Ti–O system using ALAMODE. When reconstructing the second-order force constants (FC2) from the XML output, I found that they do not satisfy the acoustic sum rule (ASR) — the diagonal terms deviate by more than 0.2 eV/Ų.

To clarify, my system is based on the experimentalTi-O structure (128 atoms), and I have not performed geometry optimization before the force-constant fitting. Below I attach my opt.in input for your reference. Thank you for taking the time to answer my question

&general
PREFIX = TiO_222
MODE = opt
NAT = 128; NKD = 2
KD = Ti O
/

&interaction
NORDER = 5 # 1: harmonic, 2: cubic, ..
NBODY = 2 3 3 2 2
/
&optimize
LMODEL= enet
DFSET=force_set

CV = 0
L1_RATIO = 1.0
#CV_MINALPHA = 1.0e-8
#CV_MAXALPHA = 0.02
CV_NALPHA = 100
L1_ALPHA = 5.12204e-06
STANDARDIZE = 1
#CONV_TOL = 1.0e-8
/
&cutoff
- None None 12.0 12.0 12.0