ADC(2) excitation energies do not match value reported in literature.

Hi,

I'm trying to replicate some ADC(2) excitation energies from this paper: pubs.rsc.org/en/content/articlelanding/2013/cp/c2cp42694c .

However, my calculated ADC(2) values do not match theirs, even though I'm using the same basis set and geometries (included in the SI of the above paper). I believe the paper uses Turbomole for their calculations. I'm using a binary download of MRCC on a linux cluster.

For example, the paper reports a salicylic acid excitation energy of 3.414 eV (Exp is 3.7 eV) with ADC(2)/aug-cc-pVTZ with an MP2 ground state geometry. I'm getting 5.13 eV with MRCC.
Am I doing something wrong? Here's the input file I'm using.

basis=aug-cc-pVTZ
calc=ADC(2)
mem=10GB
# ntrip=6
nsing=3

unit=bohr
geom=xyz
16

C 2.082333027969 3.013890416754 0.000000000000
C 2.143516457278 0.372773693943 0.000000000000
C -0.155200829985 -0.971594395968 0.000000000000
C -2.457384397055 0.342083096372 0.000000000000
C -2.489672002148 2.957678572813 0.000000000000
C -0.206145531489 4.287264056233 0.000000000000
O 4.417066783742 -0.764360697936 0.000000000000
C -0.075900043263 -3.740146154136 0.000000000000
O -2.363380067596 -4.849597714063 0.000000000000
O 1.868263397661 -4.998221122013 0.000000000000
H -4.197456909867 -0.723979855371 0.000000000000
H 3.860408481051 4.018891214680 0.000000000000
H -4.267797641530 3.960539023082 0.000000000000
H -0.213590129513 6.331154231723 0.000000000000
H 4.089257930714 -2.586837626709 0.000000000000
H -2.034318525969 -6.649536739403 0.000000000000
 

Hi,

With the MRCC program, currently, only vertical excitation energies can be calculated, while the above-mentioned article deals with 0-0 transitions. The difference between these two transitions can be summarized in the following figure:
d3i71xaburhd42.cloudfront.net/660735c89a...3a31/2-Figure1-1.png

Soon, we plan to implement gradients for the ADC(2) method, which is essential for calculating 0-0 transitions.

Best wishes,
Dávid