[molpro-user] UCCSD(T)-F12 warning: poor overlap with frozen basis set

[Benj FitzPatrick]

Hello,
I'm calculating a few bond fission transition state geometries and frequencies
using MOLPRO 2009.1 (patchlevel 4), and I ran into the warning shown below while
calculating the frequencies (it does not appear during the geometry
optimization) of C2H3 -> HCCH + H using the cc-pVQZ-F12 basis sets. This warning
does not occur when using the cc-pVTZ-F12 basis sets. I also haven't seen this
warning in any of the other transition state calculations. The optri basis sets
come straight from Kirk Peterson's website. Is there something I need to change
in my input file (below)? The calculation hasn't finished yet, so I don't know
the status of the results.
Thanks,
Benj FitzPatrick
University of Chicago

-------------warning------------------
 ? Warning
 ? Poor overlap with frozen basis set. May have deleted the wrong vectors!
 ? The problem occurs in SvdConstructOrthoBasis


-----------input file-----------------
  ***,ccsd-f12a opt of c2h3 to h + hcch***
 memory,300,M
 gthresh,oneint=1.d-12,twoint=1.d-12,zero=1.d-12

file,2,c2h3-ts-hcch-freq-uccsdtf12-vqzf12-a.wfu

 CC2  =    1.21727824
 HC3  =    1.06506710
 HCC3 =  162.85515883
 HC4  =    1.88110578
 HCH4 =   88.73011265
 DIH4 =  180.00119779
 HC5  =    1.06390758
 HCC5 =  171.14722870
 DIH5 = -179.98549634

    angstrom;
   nosym;

geometry={
 c;
 c ,  1 ,cc2;
 h ,  1 ,hc3  ,      2, hcc3;
 h ,  1 ,hc4  ,      3, hch4   ,      2, dih4;
 h ,  2 ,hc5  ,      1, hcc5   ,      3, dih5;
}

basis={
 default,cc-pvqz-f12

 set,optri

s,H,4.503623E+00,7.095383E-01,2.728158E-01;
...
h,C,1.228091E+00;
c,1.1,1.000000E+00;
}

 {hf;
   wf,15,1,1;
   accu,14}

{uccsd(t)-f12b,df_basis=av5z/mp2fit,ri_basis=optri,df_basis_exch=v5z/jkfit,scale_trip=1,gem_beta=1.1;
wf,15,1,1}

{frequencies;
save,5300.2}