temperature control in reactive system

Dear lammps usres
i use reax/c pair_style to simulate dissociation of oxygen molecule on nickel surface at 300 K. i check output temp for oxygen and nickel group in initial steps and i paste here these output:

all
oxygen
nickel
558.371
4965.63
194.934
507.158
4409.14
185.399
305.965
1671.86
193.440
491.799
3906.12
210.292
500.338
3607.36
244.213
404.971
2364.31
243.528
408.230
2079.86
270.545
445.102
2602.96
267.301
442.292
2442.62
277.496
395.243
1886.58
272.435
472.325
2433.50
310.787
449.507
1996.88
322.117
417.280
1551.67
323.963
397.574
1248.73
327.627
384.673
1079.91
327.589
374.752
1201.68
306.786
369.889
1134.60
307.057
377.123
1218.38
307.976
373.436
1276.46
299.185
345.520
983.659
293.118
323.711
815.297
283.391
326.392
870.932
281.702
318.260
824.014
276.766
319.760
951.530
267.862
334.220
1101.1
271.178
319.072
927.373
269.112
316.088
827.240
274.146
296.53
721.847
261.659
298.474
717.161
264.153
300.387
688.507
268.591
293.527
704.807
259.813
297.035
769.953
258.235
296.047
766.777
257.426
296.900
719.217
262.278
301.432
798.412
260.648
290.015
746.880
252.534
305.997
863.194
260.245
292.654
730.857
256.717
314.462
765.422
277.489
309.284
846.880
265.153
323.021
885.192
276.872
318.432
848.415
274.938
315.162
995.139
259.280
304.206
818.294
262.012
313.052
822.538
271.245
298.417
702.584
265.294
314.604
703.472
282.759
298.915
725.674
263.928
304.668
756.014
267.655
299.651
680.282
268.473
300.365
660.303
270.896
293.049
628.446
265.6
305.241
705.316
272.461
303.750
688.861
272.205
289.438
759.795
250.842
299.943
709.615
266.367
305.236
647.466
277.233
301.097
693.691
268.932
294.633
644.537
265.987
298.807
661.510
269.109
296.476
584.632
272.931
301.138
573.225
278.923
302.073
594.123
278.211
288.875
560.118
266.720
290.065
615.749
263.416
295.441
588.220
271.513
299.620
655.922
270.451
309.273
627.710
283.238
299.901
639.725
272.092
299.768
654.721
270.710
293.115
629.826
265.557
304.892
608.278
280.096
302.126
723.432
267.591
291.577
652.077
262.054
304.237
695.262
272.204
293.455
587.336
269.435
308.609
594.049
285.298
284.073
531.277
263.899
295.199
621.325
268.517
294.245
573.902
271.399
287.544
608.701
261.266


and here is my input file:
dimension 3
units real
boundary p p f
atom_style charge

Dear lammps usres
i use reax/c pair_style to simulate dissociation of oxygen molecule on
nickel surface at 300 K. i check output temp for oxygen and nickel group in
initial steps and i paste here these output:

​[...]

and here is my input file:
dimension 3
units real
boundary p p f
atom_style charge
###################
read_data data.mix
group nickel type 1
group oxygen type 2

#Potentiel ReaxFF
pair_style reax/c NULL
pair_coeff * * ffield.reax Ni O

fix 2 all qeq/reax 100 0.0 10.0 1.0e-6 reax/c
#############
neighbor 2 bin
neigh_modify delay 0 every 1 check no
#############
compute new3d all temp
compute tempo oxygen temp
compute tempni nickel temp

variable T equal c_new3d
variable T1 equal c_tempo
variable T2 equal c_tempni

fix qq all print 100 "T {T1} ${T2}" screen no file temp.out
fix walls all wall/reflect zhi 21.000000 units box
region lower block -0.79863834 35.098638 -0.74863834 35.148638 -1
2 units box
group lower region lower
group rest subtract all lower

velocity rest create 300 458127641 mom yes rot yes dist gaussian
fix 33 lower setforce 0.0 0.0 0.0
fix qw all nvt temp 300 300 100 drag 2
run 1000000

in fact dissociation of oxygen on surface occur at initial steps and in
these steps temp for oxygen group is high,and the high temp affect reaction
rate.how can i maintain the whole system temp and oxygen and nickel group
temp at 300 K?

sorry, but how can this be a realistic description of a system where an
exothermic reaction is happening? and where there is nothing your atoms in
reality would couple to that they would exchange kinetic energy with? in
fact, even thermostatting the oxygen atoms *at all* is in my personal
opinion a questionable approach (as is the temperature control of the metal
atoms at the surface).

​axel.​



Dear Axel
Thanks for your suggestion. Do you mean I should just thermostat nickel slab(no surface atoms or oxygen atom)?




Sent from Yahoo Mail on Android

|

Dear Axel
Thanks for your suggestion. Do you mean I should just thermostat nickel
slab(no surface atoms or oxygen atom)?

Sent from Yahoo Mail on Android
<https://overview.mail.yahoo.com/mobile/?.src=Android>

​what i mean is that you have to understand the chemistry and
thermodynamics of the process you are simulating and not just make it look
like any regular equilibrium bulk simulation. what is adequate, you need to
decide by yourself and that is best done by looking at the published
literature of similar systems.

my usual recommendation for such slab systems is a bit more elaborate:
bottom layer of metal atoms is immobilized.
the rest of the system is integrated with fix nve.
the lower section (but not the topmost N layers) are thermalized with fix
langevin, or temp/csvr or whatever else tickles your fancy to model the
thermal coupling to the bulk.

the only problem with this setup is that many times, people deposit atoms
faster than what is realistic in an experiment or at a higher density, so
that the simulation doesn't take a very long time. that is where the person
doing the simulation has to decide what is the best compromise for your
specific needs. that is impossible to say from remote, since only you know
what is important in your research.

axel.



Dear Axel
I am trying to use your recommendation setup for my system, as you said the simulation doesnt take a very long time, but there is one thing i cant understand, at the beginning of run as expected temperature is rising from 300 to 700 k, and then is decreased and becomes below 100k and because of this decreasing the reaction does not go on, what causes this temperature decreasing?



Best Regard


|

Dear Axel
I am trying to use your recommendation setup for my system, as you said
the simulation doesnt take a very long time, but there is one thing i cant
understand, at the beginning of run as expected temperature is rising from
300 to 700 k, and then is decreased and becomes below 100k and because of
this decreasing the reaction does not go on, what causes this temperature
decreasing?

​sorry, but i am not a psychic. most likely you either set up your
simulation wrong or are looking at the wrong temperature or both.​

​axel.​

thanks