The generation of isotopically-substituted molecular crystal spectra has become a point of interest, which means blog post. To be clear, this is for cases where isotopic substitution does not affect the crystal geometry – the crystal cell does not change significantly upon deuteration (and for those who believe isotopic substitution never leads to significant changes in the solid, I refer you Zhou, Kye, and Harbison’s article on Isotopomeric Polymprphism and their work on 4-methylpyridine pentachlorophenol, which changes dramatically upon deuteration. I beat on this point because blindly assuming of the crystal cell geometry in such cases will produce spectra noticeably different than measured. It’s NOT the calculation’s fault!).

The generation of isotopically-substituted spectra and intensities in Crystal09 is trivial provided that you KEEP THE FREQINFO.DAT FILE. In fact, you need keep ONLY the FREQINFO.DAT to generate these spectra, which greatly reduces file transfer loads and allows for the scripted calculation of new vibrational spectra and thermodynamic data post-frequency calculation.

As my example system, I’m using the dispersion-corrected crystal cell of alpha-HMX (I have it handy, it’s a small system, and having anything about HMX on your website is proven to increase traffic) at the B3LYP/6-31G(d,p) level of theory. Original input file (the one where the original normal mode analysis is performed) is below:

Test - alpha-HMX 6-31Gdp set DFT/B3LYP FREQ
CRYSTAL
0 0 0
43
15.14 23.89 5.913 124.3
14
6      1.016493675797E-01 -4.109909899348E-02 -3.351438244488E-03
6     -6.539109813231E-02 -6.180633576707E-02 -1.110575784790E-02
1      9.149797846691E-02 -4.382919469310E-02 -1.860042940246E-01
1      1.558888705857E-01 -6.829708099502E-02  4.595161229829E-02
1     -5.138242817334E-02 -5.844587273099E-02 -1.920922064181E-01
1     -9.781600273101E-02 -1.015710562102E-01  2.063738273292E-02
7      1.992579327285E-02 -5.951921578598E-02  1.040704228546E-01
7      1.232154652110E-01  1.634305404407E-02  5.951841980010E-02
7      2.220759010770E-02 -7.142100857312E-02  3.299259852838E-01
7      2.054067942916E-01  2.817244373261E-02  1.473285310628E-01
8     -4.761487685316E-02 -8.656669456613E-02  4.192568497756E-01
8      9.327421157186E-02 -6.479426971916E-02  4.286363161888E-01
8      2.563441491059E-01 -1.128705054032E-02  1.760581823035E-01
8      2.225071782791E-01  7.736574474011E-02  1.903699942346E-01
FREQCALC
INTENS
END
END
8 4
0 0 6 2.0 1.0
 5484.671700         0.1831100000E-02
 825.2349500         0.1395010000E-01
 188.0469600         0.6844510000E-01
 52.96450000         0.2327143000    
 16.89757000         0.4701930000    
 5.799635300         0.3585209000  
0 1 3 6.0 1.0
 15.53961600        -0.1107775000         0.7087430000E-01
 3.599933600        -0.1480263000         0.3397528000    
 1.013761800          1.130767000         0.7271586000    
0 1 1 0.0 1.0
 0.2700058000          1.000000000          1.000000000
0 3 1 0.0 1.0
 0.800000000          1.00000000    
7 4
0 0 6 2.0 1.0
       4173.51100         0.183480000E-02
       627.457900         0.139950000E-01
       142.902100         0.685870000E-01
       40.2343300         0.232241000    
       12.8202100         0.469070000    
       4.39043700         0.360455000    
0 1 3 5.0 1.0
       11.6263580        -0.114961000         0.675800000E-01
       2.71628000        -0.169118000         0.323907000    
      0.772218000          1.14585200         0.740895000    
0 1 1 0.0 1.0
      0.212031300          1.00000000          1.00000000    
0 3 1 0.0 1.0
 0.800000000          1.00000000    
6 4
0 0 6 2.0 1.0
    .3047524880D+04   .1834737130D-02
    .4573695180D+03   .1403732280D-01
    .1039486850D+03   .6884262220D-01
    .2921015530D+02   .2321844430D+00
    .9286662960D+01   .4679413480D+00
    .3163926960D+01   .3623119850D+00
0 1 3 4.0 1.0
    .7868272350D+01  -.1193324200D+00   .6899906660D-01
    .1881288540D+01  -.1608541520D+00   .3164239610D+00
    .5442492580D+00   .1143456440D+01   .7443082910D+00
0 1 1 0.0 1.0
    .1687144782D+00   .1000000000D+01   .1000000000D+01
0 3 1 0.0 1.0
    .8000000000D+00   .1000000000D+01
1 3
0 0 3 1.0 1.0
    .1873113696D+02   .3349460434D-01
    .2825394365D+01   .2347269535D+00
    .6401216923D+00   .8137573262D+00
0 0 1 0.0 1.0
    .1612777588D+00   .1000000000D+01
0 2 1 0.0 1.0
    .1100000000D+01   .1000000000D+01
99 0
END
DFT
B3LYP
XLGRID
END
EXCHSIZE
10654700
BIPOSIZE
10654700
TOLINTEG
8 8 8 8 16
SCFDIR
MAXCYCLE
100
TOLDEE
11
GRIMME
1.05 20. 25.
4
1 0.14 1.001
6 1.75 1.452 
7 1.23 1.397
8 0.70 1.342
SHRINK
8 8
LEVSHIFT
5 0
FMIXING
50
END
END

Upon completion of this run, you need only the FREQINFO.DAT file, the last set of coordinates from the .OUT file (for atom counting purposes) and an input file which is modified from the original only in the specification of the ISOTOPES section and which includes a RESTART.

Question - how does one deal with isotopically-labeling atoms when it breaks the space group symmetry? If I isotopically label Atom 1 in the asymmetric unit, what happens to the other N symmetry-related atoms?

Answer - Crystal09, in its infinite wisdom, does not consider the asymmetric unit in the isotopic substitution scheme. If you’ve 14 atoms in the asymmetric unit (the symmetry-unique atoms you provide in the input file)…

14
6      1.016493675797E-01 -4.109909899348E-02 -3.351438244488E-03
6     -6.539109813231E-02 -6.180633576707E-02 -1.110575784790E-02
...
8      2.563441491059E-01 -1.128705054032E-02  1.760581823035E-01
8      2.225071782791E-01  7.736574474011E-02  1.903699942346E-01

and 56 atoms in the full unit cell…

ATOMS IN THE ASYMMETRIC UNIT   14 - ATOMS IN THE UNIT CELL:   56
     ATOM              X/A                 Y/B                 Z/C    
 *******************************************************************************
   1 T   6 C    -1.460999048177E-01  1.393970283287E-01  6.390170683069E-02
   2 F   6 C     1.393970283287E-01 -1.460999048177E-01 -5.719883034171E-02
   3 F   6 C     3.071988303417E-01  1.860982931693E-01  1.106029716713E-01
   4 F   6 C     1.860982931693E-01  3.071988303417E-01  3.960999048177E-01
...
  53 T   8 O     4.522856069554E-02  3.355114277736E-01  1.095029287847E-01
  54 F   8 O     3.355114277736E-01  4.522856069554E-02 -4.902429172538E-01
  55 F   8 O    -2.597570827462E-01  1.404970712153E-01 -8.551142777356E-02
  56 F   8 O     1.404970712153E-01 -2.597570827462E-01  2.047714393045E-01

your ISOTOPES section relies on the numbering of the atoms in the “56 atom” list.

The input file below will calculate an isotopically-labeled vibrational spectrum for 8 of the hydrogen atoms that ends up breaking the unit cell symmetry (which will be more obvious from the produced mode energies). Again, the atom numbers come from the “ATOMS IN THE ASYMMETRIC UNIT” part of the original optimization by which you performed the original normal mode analysis (hopefully).

Test - alpha-HMX 6-31Gdp set DFT/B3LYP FREQ - Isotopic Substitution
CRYSTAL
0 0 0
43
15.14 23.89 5.913 124.3
14
6      1.016493675797E-01 -4.109909899348E-02 -3.351438244488E-03
6     -6.539109813231E-02 -6.180633576707E-02 -1.110575784790E-02
1      9.149797846691E-02 -4.382919469310E-02 -1.860042940246E-01
1      1.558888705857E-01 -6.829708099502E-02  4.595161229829E-02
1     -5.138242817334E-02 -5.844587273099E-02 -1.920922064181E-01
1     -9.781600273101E-02 -1.015710562102E-01  2.063738273292E-02
7      1.992579327285E-02 -5.951921578598E-02  1.040704228546E-01
7      1.232154652110E-01  1.634305404407E-02  5.951841980010E-02
7      2.220759010770E-02 -7.142100857312E-02  3.299259852838E-01
7      2.054067942916E-01  2.817244373261E-02  1.473285310628E-01
8     -4.761487685316E-02 -8.656669456613E-02  4.192568497756E-01
8      9.327421157186E-02 -6.479426971916E-02  4.286363161888E-01
8      2.563441491059E-01 -1.128705054032E-02  1.760581823035E-01
8      2.225071782791E-01  7.736574474011E-02  1.903699942346E-01
FREQCALC
RESTART
ISOTOPES
8
9  2
10 2
11 2
13 2
14 2
15 2
16 2
18 2
INTENS
END
END
8 4
0 0 6 2.0 1.0
 5484.671700         0.1831100000E-02
 825.2349500         0.1395010000E-01
 188.0469600         0.6844510000E-01
 52.96450000         0.2327143000    
 16.89757000         0.4701930000    
 5.799635300         0.3585209000  
0 1 3 6.0 1.0
 15.53961600        -0.1107775000         0.7087430000E-01
 3.599933600        -0.1480263000         0.3397528000    
 1.013761800          1.130767000         0.7271586000    
0 1 1 0.0 1.0
 0.2700058000          1.000000000          1.000000000
0 3 1 0.0 1.0
 0.800000000          1.00000000    
7 4
0 0 6 2.0 1.0
       4173.51100         0.183480000E-02
       627.457900         0.139950000E-01
       142.902100         0.685870000E-01
       40.2343300         0.232241000    
       12.8202100         0.469070000    
       4.39043700         0.360455000    
0 1 3 5.0 1.0
       11.6263580        -0.114961000         0.675800000E-01
       2.71628000        -0.169118000         0.323907000    
      0.772218000          1.14585200         0.740895000    
0 1 1 0.0 1.0
      0.212031300          1.00000000          1.00000000    
0 3 1 0.0 1.0
 0.800000000          1.00000000    
6 4
0 0 6 2.0 1.0
    .3047524880D+04   .1834737130D-02
    .4573695180D+03   .1403732280D-01
    .1039486850D+03   .6884262220D-01
    .2921015530D+02   .2321844430D+00
    .9286662960D+01   .4679413480D+00
    .3163926960D+01   .3623119850D+00
0 1 3 4.0 1.0
    .7868272350D+01  -.1193324200D+00   .6899906660D-01
    .1881288540D+01  -.1608541520D+00   .3164239610D+00
    .5442492580D+00   .1143456440D+01   .7443082910D+00
0 1 1 0.0 1.0
    .1687144782D+00   .1000000000D+01   .1000000000D+01
0 3 1 0.0 1.0
    .8000000000D+00   .1000000000D+01
1 3
0 0 3 1.0 1.0
    .1873113696D+02   .3349460434D-01
    .2825394365D+01   .2347269535D+00
    .6401216923D+00   .8137573262D+00
0 0 1 0.0 1.0
    .1612777588D+00   .1000000000D+01
0 2 1 0.0 1.0
    .1100000000D+01   .1000000000D+01
99 0
END
DFT
B3LYP
XLGRID
END
EXCHSIZE
10654700
BIPOSIZE
10654700
TOLINTEG
8 8 8 8 16
SCFDIR
MAXCYCLE
100
TOLDEE
11
GRIMME
1.05 20. 25.
4
1 0.14 1.001
6 1.75 1.452 
7 1.23 1.397
8 0.70 1.342
SHRINK
8 8
LEVSHIFT
5 0
FMIXING
50
END
END

The difference is in the FREQCALC section, which calls RESTART (to use the FREQINFO.DAT file), ISOTOPES (obvious), the total number of atoms that are having their isotopes changed (8), then the list, containing the atom number and the new mass (here, 2 for deuterium).

The proof is in the high-frequency region, where the last 16 modes (H-atom motion) in the non-deuterated form…

 HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

    MODES         EIGV          FREQUENCIES     IRREP  IR   INTENS    RAMAN
             (HARTREE**2)   (CM**-1)     (THZ)             (KM/MOL)
...
  153- 153    0.2003E-03   3106.1384   93.1197  (A2 )   I (     0.00)   A
  154- 154    0.2003E-03   3106.5054   93.1307  (B1 )   A (     0.02)   A
  155- 155    0.2004E-03   3106.5586   93.1323  (A1 )   A (     0.23)   A
  156- 156    0.2004E-03   3106.8420   93.1408  (B2 )   A (     0.48)   A
  157- 157    0.2017E-03   3117.1664   93.4503  (B2 )   A (     1.13)   A
  158- 158    0.2018E-03   3117.4901   93.4600  (B1 )   A (     2.33)   A
  159- 159    0.2021E-03   3120.2876   93.5439  (A1 )   A (   115.24)   A
  160- 160    0.2022E-03   3120.7805   93.5586  (A2 )   I (     0.00)   A
  161- 161    0.2131E-03   3203.6552   96.0432  (A1 )   A (    44.59)   A
  162- 162    0.2131E-03   3203.6581   96.0433  (B2 )   A (   115.98)   A
  163- 163    0.2132E-03   3204.6505   96.0730  (B1 )   A (    15.30)   A
  164- 164    0.2132E-03   3204.8874   96.0801  (A2 )   I (     0.00)   A
  165- 165    0.2157E-03   3223.4669   96.6371  (A1 )   A (    44.98)   A
  166- 166    0.2157E-03   3223.5803   96.6405  (B2 )   A (    27.02)   A
  167- 167    0.2158E-03   3223.8536   96.6487  (B1 )   A (    35.26)   A
  168- 168    0.2158E-03   3224.3355   96.6631  (A2 )   I (     0.00)   A

change to the following last 16 modes (H/D-atom motion) upon deuteration. Note the mode energies split and the mode symmetries go from (A1,A2,B1,B2) to (A). Also note your IR mode intensities change, giving you the complete picture upon isotopic substitution.

 HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

    MODES         EIGV          FREQUENCIES     IRREP  IR   INTENS    RAMAN
             (HARTREE**2)   (CM**-1)     (THZ)             (KM/MOL)
...
  153- 153    0.1074E-03   2274.8942   68.1996  (A  )   A (     1.07)   A
  154- 154    0.1075E-03   2275.5949   68.2206  (A  )   A (     3.75)   A
  155- 155    0.1075E-03   2275.7008   68.2238  (A  )   A (     2.93)   A
  156- 156    0.1099E-03   2300.7446   68.9746  (A  )   A (     4.68)   A
  157- 157    0.1148E-03   2351.7846   70.5047  (A  )   A (    11.32)   A
  158- 158    0.1183E-03   2387.0269   71.5613  (A  )   A (    36.17)   A
  159- 159    0.1183E-03   2387.2610   71.5683  (A  )   A (    16.04)   A
  160- 160    0.1184E-03   2387.6687   71.5805  (A  )   A (     3.73)   A
  161- 161    0.2006E-03   3108.6223   93.1942  (A  )   A (     0.93)   A
  162- 162    0.2009E-03   3110.5061   93.2506  (A  )   A (    12.43)   A
  163- 163    0.2009E-03   3110.7567   93.2581  (A  )   A (    13.67)   A
  164- 164    0.2039E-03   3134.0133   93.9554  (A  )   A (    40.48)   A
  165- 165    0.2147E-03   3215.5160   96.3987  (A  )   A (    19.38)   A
  166- 166    0.2157E-03   3223.4291   96.6360  (A  )   A (    35.29)   A
  167- 167    0.2157E-03   3223.5925   96.6409  (A  )   A (    29.50)   A
  168- 168    0.2158E-03   3223.8729   96.6493  (A  )   A (     8.37)   A

In press, in CrystEngComm (DOI:10.1039/C2CE26523). This is my first full paper completely internet-powered, in that I’ve not physically met any of the other co-authors (also in the internet-powered context, the recent paper on [18]-annulene was written and submitted without sharing a room with Dr. Bruce Hudson, but we’re in the same building, so it doesn’t quite count). Also, one of the few papers for which I had no image generation duties (a rare treat).

The discussion of the very interesting possibilities of molecular redox materials in lithium-ion batteries aside, this paper presents a very thorough example of the power of computational approaches to greatly improve the understanding of solid-state molecular materials by (specifically) 1: overcoming the hydrogen position identification problems inherent in X-ray diffraction methods, 2: reproducing the changes that come with temperature variations in molecular crystals and explaining the origins of those (possibly subtle) changes by way of dispersion-corrected density functional theory, and 3: demonstrating that the nature of intermolecular interactions (specifically hydrogen bonding) can be rigorously cataloged across varied materials using post-optimization tools (in this case, using Carlo Gatti’s excellent TOPOND program).

Gaëtan Bonnard, Anne-Lise Barrès, Olivier Mentré, Damian G. Allis, Carlo Gatti, Philippe Poizot and Christine Frayret^*

Abstract

Following our first experimental and computational study of the room temperature (RT) form of the tetrahydrated 3,6-dihydroxy-2,5-dimethoxy-p-benzoquinone (LiM₂DHDMQ⋅4H₂O) compound, we have researched the occurrence of hydrogen ordering in a new polymorph at lower temperature. The study of polymorphism for the Li₂DHDMQ⋅4H₂O phase employs both experimental (single crystal X-ray diffraction) and theoretical approaches. While clues for disorder over one bridging water molecule were observed at RT (beta-form),a fully ordered model within a supercell has been evidenced at 100K (alpha-form) and is discussed in conjunction with the features characterizing the first polymorphic form reported previously. Density functional theory (DFT) calculations augmented with an empirical dispersion correction (DFT-D) were applied for the prediction of the structural and chemical bonding properties of the alpha and beta polymorphs of Li₂DHDMQ·4H₂O. The relative stability of the two polymorphic systems is evidenced. An insight into the interplay of hydrogen bonding, electrostatic and van der Waals (vdW) interactions in affecting the properties of the two polymorphs is gained. This study also shows how information from DFT-D calculations can be used to augment the information from the experimental crystal diffraction pattern and can so play an active role in crystal structure determination, especially by increasing the reliability and accuracy of H-positioning. These more accurate hydrogen coordinates allowed for a quantification of H-bonding strength through a topological analysis of the electron density (Atoms-in-molecules theory).

For those itching for polarizability derivative orientation information and wondering where it is when you ask for it… what’s included below is a combination of a few points in one, specifically pointing out that the IOp options are not just “another part” of the Gaussian input file (with the IOp Overlays currently linked HERE).

The problem I realized after an email from Gaussian HQ was that, as was the case for the KMLYP density functional call discussed in previous posts about [18]-annulene, “opt” and “freq” keyword combinations are seen as two distinct runs in Gaussian that don’t pass the IOp information along (and, admittedly, I should have remembered that). Specifically, the additional print-out for the polarizability info is called by IOp(7/33=3).

What I provide below is a two-in-one input file that saves you from having to run double-duty input files in the checkpoint file. This also serves as a template for those looking for examples of combining multi-step input files that include mixed basis sets (as many of the problems I’ve been emailed stem from carriage return issues more than anything else). Note that the input file is set to run Raman intensities and produce higher-precision (hpmodes) eigenvectors (so, if you just want to test this, remove the “raman”).

%chk=C4H5Cl_B3LYP_631Gdp_LanL2DZ_IR_Raman.chk
#p scf=tight opt=tight b3lyp/GEN pseudo=read

C4H5Cl_B3LYP_631Gdp_LanL2DZ_IR_Raman Opt

0 1
 C                 -1.74671095   -0.64168298    0.00000000
 H                 -1.53944096   -1.69141587    0.00000000
 C                 -0.73010315    0.25446188    0.00000000
 H                 -0.93737314    1.30419477    0.00000000
 C                  0.73010315   -0.25446188    0.00000000
 H                  0.93737314   -1.30419477    0.00000000
 C                  1.74671095    0.64168298    0.00000000
 H                  1.53944096    1.69141587    0.00000000
 H                 -3.73526840    0.03531673    0.00000000
 Cl                 3.73526840   -0.03531673    0.00000000

C H 0
6-31G(d,p)
****
Cl
Lanl2DZ
****

Cl
Lanl2DZ

--Link1--
%chk=C4H5Cl_B3LYP_631Gdp_LanL2DZ_IR_Raman.chk
#p Geom=Check Guess=Read freq(raman,hpmodes) iop(7/33=3)
 
C4H5Cl_B3LYP_631Gdp_LanL2DZ_IR_Raman Freq
     
0 1

Note the carriage return after the second “0 1″.

For the demo molecule above, additional print-out below.

 Dipole derivatives wrt mode   1:  3.96988D-14 -1.15747D-14 -1.96904D-01
 Polarizability derivatives wrt mode          1
                 1             2             3 
      1   0.000000D+00  0.000000D+00  0.206435D+00
      2   0.000000D+00  0.000000D+00  0.143916D-01
      3   0.206435D+00  0.143916D-01  0.000000D+00
 Vibrational polarizability contributions from mode   1       0.0000000       0.0000000       0.0257731
 IFr=  0 A012= 0.23D-23 0.77D+00 0.13D+00 Act= 0.90D+00 DepolP= 0.75D+00 DepolU= 0.86D+00

Alternately, keep track of the checkpoint file.

In press (DOI:10.1016/j.molstruc.2012.07.051) in the Journal Of Molecular Structure. May go down in history as a hardest-fought paper acceptance. In a similar line of research as the [18]-annulene study, but exploring the infinite limit of geometry and bond length alternation energy barrier for this infinite case. If the numbers are correct, the infinite polyene chains (polyacetylene) do not exhibit bond length alternation because the Peierls’ barrier between the single-double and double-single bond alternate minima is below the vibrational zero-point level. Plenty of ramifications.

Bruce S. Hudson and Damian G. Allis

Abstract. The potential energy of the infinite periodic chain model of polyacetylene (pPA) is symmetric with two equivalent minima separated by the Peierls’ stabilization barrier. In this work it is shown how an energy scale and vibrational energy levels for this highly anharmonic Peierls’ degree of freedom can be estimated. Attention is given to the potential energy increase for large deformations. The Born-Kármán treatment of translational symmetry is applied. Two empirical methods and a direct periodic boundary condition (PBC) density functional theory (DFT) calculations are in semi-quantitative agreement, each leading to the conclusion that pPA has a zero-point level that is above the Peierls’ barrier. The argument does not depend critically on the barrier height or the other parameters of the model or the computation method. It is concluded that pPA will not exhibit bond alternation and that the zero-point average geometry does not preclude possible conductivity.

In press (DOI:10.1016/j.molstruc.2012.05.016) in the Journal Of Molecular Structure (Volume 1023, 12 September 2012, Pages 212–215) in the special issue: MOLECULAR VIBRATIONS AND STRUCTURES: THEORY AND EXPERIMENT — A collection of papers dedicated to Professor Jaan Laane on the occasion of his 70th birthday.

This paper on the “actual” geometry of [18]-annulene is part of several larger stories addressing a larger polyene (or larger-polyene) issue. First among these is the meaning of experimental results obtained by various spectroscopic methods (in this case, using previous X-ray, Raman (with the C2 (blue) and D6h (red) simulated spectra shown in the image above), IR, and NMR data that produce different results within the limitations of the methods to study the single molecule). Second is the quality of the theoretical method for reproducing certain types of spectroscopic data. In the case of the [N]-annulene series, the ever-present B3LYP density functional is found to produce the time-average geometry of [18]-annulene found in X-ray data, but another density functional (in this case, KMLYP), finds that bond-alternate minima exist. Third is the importance of the zero-point level in the treatment of systems for which bond-alternate geometries exist with transition-state barriers calculated to be below the zero-point level in the classical approximation of nuclear positions (the Born-Oppenheimer Approximation).

NOTE 1: The KMYLP density functional is called in Gaussian with the following keyword set:

BLYP iop(3/76=1000005570) iop(3/77=0000004430) iop(3/78=0448010000)

NOTE 2: Optimization and Frequency calculations must be performed as TWO SEPARATE CALCULATIONS. The iop-called density functional does not carry itself over between opt + freq (or other properties) in the same input file. If you opt + freq in the same input file, you will Opt with KMLYP but freq with BLYP. This will be obvious by the number of imaginary modes.

Bruce S. Hudson and Damian G. Allis

Abstract. [18]-annulene has been of great interest from the structural point of view of its bond alternation. High-level calculations based on structures selected for agreement with NMR spectra lead to a bond-alternate C2 form over a non-alternating planar D6h structure deduced from diffraction, infrared (IR) and electronic spectral studies. Here it is shown that computed Raman spectra for the D6h and C2 forms are expected to be very different. However, two equivalent non-D6h bond-alternate minima of D3h or C2 geometries are separated by only a small barrier along a motion that involves CC stretching and compression. It is shown here that the zero-point level is above the barrier for this species. In light of that fact, the NMR calculations are reconsidered with inclusion of zero-point level averaging.