Incorporating Fluorene (C13H10) into EMC with OPLS-AA/2024 – Missing Typing Rules

EMC
1

Dear EMC Development Team,

I am using EMC (Enhanced Monte Carlo) with the OPLS-AA/2024 force field and am attempting to incorporate fluorene (C13H10, SMILES: c1ccc2c3ccccc3Cc2c1) as a building block.

Here is a summary of what I have done so far:

  1. In opls-aa.define, I added a RESI FLUO entry.

    image
    image318×1302 51.2 KB

  2. In opls-aa.top, I added the following atom typing rules:
    377 c3a C FLUO - 0 ^6c(^6:c)(^6:c)(^6:c)
    378 h1a H FLUO - 0 H(^4c)

  3. In the EMC script, I defined the group and cluster:
    ITEM GROUPS
    fluo c1ccc2c3ccccc3Cc2c1
    ITEM END
    ITEM CLUSTERS
    fluo fluo, 1
    ITEM END

However, I am getting the following warnings and fatal error:

Warning: no rule found for {group, site} = {fluo, 0}.
Warning: no rule found for {group, site} = {fluo, 2}.
Warning: no rule found for {group, site} = {fluo, 4}.
Warning: no rule found for {group, site} = {fluo, 6}.
Error: core/fields.c:433 FieldsApply: Missing rules. Program aborted.

Are there any additional entries required in opls-aa.prm, beyond what is in .define and .top?

Thank you very much for your time and for the excellent EMC software. I would greatly appreciate any guidance.

Best regards,

KP

2

Dear user,

EMC interprets force fields either in a strict or less strict way. Number of ring members in rules are ignored in the latter, while number of ring members are considered in the former case. The number of ring members are indicated by adding a carrot to a rule, as you pointed out correctly. E.g. ^6c indicates an aromatic carbon being part of a 6-membered ring.

Coming back to fluorene:

fluorene
fluorene250×130 4.68 KB

The above structure consists out of two aromatic six membered rings and one mixed five membered ring. EMC’s interpretation of these rings can be determined by using -field_debug=reduced to obtain considered rules, which shows that carbons shared between five and six membered rings are identified as part of the smaller of the two rings. Furthermore, OPLS partial charges are different for SP3 carbons connected to aromatic rings when compared to your suggestion, as can be deduced from entries 148 and 149 in $EMC_ROOT/field/opls/2024/src/oplsaa.par,

148 06 CT   -0.065     3.500     0.066     all-atom C: CH3, toluene
149 06 CT   -0.005     3.500     0.066     all-atom C: CH2, ethyl benzene

where the fourth column represents partial charge. Bearing the above in mind, your entry would need to be altered into

RESI	FLUO	0		! C13H10, fluorene
GROUP					!
ATOM	C1	c3a	-0.115	!          H4          H12
ATOM	H1	h1a	0.115	!          |           |
GROUP					!          C4          C12
ATOM	C2	c3a	-0.115	!         /  \\       /  \\
ATOM	H2	h1a	0.115	!   H3--C3    C5----C13   C11--H11
GROUP					!     	||    |     ||    | 
ATOM	C3	c3a	-0.115	!   H2--C2    C6    C8    C10--H10
ATOM	H3	h1a	0.115	!         \  // \  /  \  //
GROUP					!          C1    C7    C9
ATOM	C4	c3a	-0.115	!          |    /  \   |
ATOM	H4	h1a	0.115	!          H1 H81  H82 H9
GROUP					!
ATOM	C5	c3a	0		!
GROUP					!
ATOM	C6	c3a	-0.115	!
ATOM	C7	c4	0.11	!
ATOM	H71	h1	0.06	!
ATOM	H72	h1	0.06	!
ATOM	C8	c3a	-0.115	!
GROUP					!
ATOM	C9	c3a	-0.115	!
ATOM	H9	h1a	0.115	!
GROUP					!
ATOM	C10	c3a	-0.115	!
ATOM	H10	h1a	0.115	!
GROUP					!
ATOM	C11	c3a	-0.115	!
ATOM	H11	h1a	0.115	!
GROUP					!
ATOM	C12	c3a	-0.115	!
ATOM	H12	h1a	0.115	!
GROUP					!
ATOM	C13	c3a	0		!

CONECT	C1	^6:C2	H1
CONECT	C2	^6:C3	H2
CONECT	C3	^6:C4	H3
CONECT	C4	^5:C5	H4
CONECT	C5	^5:C6
CONECT	C6	^5:C1	^5C7
CONECT	C7	^5C8	H71	H72
CONECT	C8	^6:C9
CONECT	C9	^6:C10	H9
CONECT	C10	^6:C11	H10
CONECT	C11	^6:C12	H11
CONECT	C12	^5:C13	H12
CONECT	C13	^5:C5	^5:C8

The resulting force field files amount to

opls-aa.define (111.4 KB)
opls-aa.prm (66.2 KB)
opls-aa.top (19.8 KB)

Please note that opls-aa.define is provided for completeness sake. It is not used internally to generate the needed opls-aa.prm and opls-aa.top. The latter two can be placed in $EMC_ROOT/field/opls/2024 to be used globally.