计算化学公社

标题: 势能面扫描出错 [打印本页]

作者
Author: whtu 时间: 2016-4-4 19:49
标题: 势能面扫描出错
输入文件如下：
%nprocshared=12
%mem=5GB
%chk=TPESscan1.chk
# scan BVP86/TZVP int=ultrafine scf=qc

Title Card Required

-1 1
C
C                1          B1
C                2          B2 1          A1
O                3          B3 2          A2 1          D1 0
S                4          B4 3          A3 2          D2 0
O                5          B5 4          A4 3          D3 0
O                5          B6 4          A5 6          D4 0
O                5          B7 6          A6 7          D5 0
C                1          B8 2          A7 3          D6 0
F                3          B9 4          A8 5          D7 0
F                3          B10 4          A9 5          D8 0
F                2          B11 3          A10 4          D9 0
F                1          B12 2          A11 3          D10 0
F                1          B13 2          A12 3          D11 0
F                9          B14 1          A13 2          D12 0
F                9          B15 1          A14 2          D13 0
F                9          B16 1          A15 2          D14 0
H                2          B17 3          A16 4          D15 0

B1          1.54601098
B2          1.56551790
B3          1.30677298
B4          1.98378086
B5          1.48769442
B6          1.48846811
B7          1.48761747
B8          1.57490705
B9          1.41950968
B10          1.39534334
B11          1.39448529
B12          1.37323334
B13          1.36439740
B14          1.35776384
B15          1.34635492
B16          1.35856179
B17          1.09932470
A1          116.78764446
A2          108.63682184
A3          123.33609584
A4          100.48493748
A5          102.74618498
A6          117.08134671
A7          113.61877769
A8          113.95254089
A9          115.00579851
A10       109.02509231
A11       109.07161429
A12       113.50826318
A13       110.52220879
A14       112.43176207
A15       109.88688991
A16       106.99733951
D1          13.9674 18 10.0
D2       -166.38976079
D3          84.46603020
D4       -121.44268730
D5          147.74881731
D6       -166.88298206
D7          75.77337052
D8          -44.30208448
D9          71.25342752
D10          75.54359333
D11       -45.27497734
D12       -47.62530288
D13          73.96354296
D14       -166.25225803
D15       -45.88623546

输出文件的后面几行如下：
Minimum is close to point  2 DX=  1.39D-02 DF= -8.93D-08 DXR=  6.51D-02 DFR=  4.25D-03 which will be used.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
Restarting incremental Fock formation.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
   Minimum is close to point  2 DX= -1.22D-02 DF= -8.75D-08 DXR=  6.47D-02 DFR=  4.21D-03 which will be used.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
   Minimum is close to point  2 DX=  1.19D-03 DF= -7.21D-10 DXR=  5.93D-03 DFR=  3.52D-05 which will be used.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
   Minimum is close to point  2 DX=  1.67D-02 DF= -1.20D-07 DXR=  7.69D-02 DFR=  5.95D-03 which will be used.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
   Minimum is close to point  2 DX= -1.42D-02 DF= -1.18D-07 DXR=  7.65D-02 DFR=  5.88D-03 which will be used.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
   Minimum is close to point  2 DX= -1.18D-03 DF= -6.91D-10 DXR=  5.93D-03 DFR=  3.51D-05 which will be used.
Gradient too large for Newton-Raphson -- use scaled steepest descent instead.
   Minimum is close to point  2 DX=  1.39D-02 DF= -8.23D-08 DXR=  6.51D-02 DFR=  4.25D-03 which will be used.
Convergence failure.
Error termination via Lnk1e in /home/whtu/gaussian/g09/l508.exe at Mon Apr  4 13:40:37 2016.
Job cpu time:    1 days 14 hours 11 minutes 59.9 seconds.

File lengths (MBytes):  RWF= 157 Int=    0 D2E=    0 Chk=    9 Scr=    2

请问为什么出现这样的错误？该怎么解决？

作者
Author: sobereva 时间: 2016-4-4 20:31
这是因为用了SCF=QC，但是QC方法又没法收敛所致。通常建议考虑用其它方式解决SCF不收敛：
解决SCF不收敛问题的方法
http://sobereva.com/61

但你这个体系不收敛很大程度原因是扫描到那个点之后，结构明显不合理。结构越是不合理，通常SCF收敛越困难。

欢迎光临计算化学公社 (http://bbs.keinsci.com/)