Get my PhD Thesis

More documents

Recommendations

Info

Part 3 Benchmarking for Radicals As expected, the cc-pVDZ basis set does not provide accurate geometries and frequencies, and the cc-pVTZ numbers are clearly more in the range of the experimental data than the cc-pVDZ numbers. CCSD displays its insufficiency for prediction of equilibrium properties by differing from the FCI values by 0.01Å in the geometry, 90 cm -1 in the frequency, and 35 kJ/mol in the atomization energy. The errors in R eq and ω e are reduced by a factor of four going to the CCSDT level and a factor of five going from the CCSDT to the CCSDTQ level. The error in the atomization energy is reduced by a factor of nine going to the CCSDT level and a factor of eight going from the CCSDT to the CCSDTQ level, but while the equilibrium geometry on the CCSDTQ level is only 0.0005Å from the FCI value, the harmonic frequency is still about 5 cm -1 too high. Both the equilibrium geometry and the harmonic frequency are apparently better approximated by the CCSDT method than the CCSDTQ. This is due to a favorable cancellation in errors for CCSDT calculations in small basis sets. By extrapolation to the larger aug-cc-pVQZ basis, 67,75 we get an equilibrium distance of 1.1759Å and a harmonic frequency of 2060cm -1 at the CCSDTQ level. 3.3.4 The Vertical Electron Affinity of CN Calculations on CN - and CN were carried out in the aug´-cc-pVDZ basis at the experimental equilibrium geometry for CN. The FCI calculation on CN - is one of the largest FCI calculations carried out so far containing about 20 billion Slater determinants. The vertical electron affinity (EA) was found and is displayed in Table 3-2. Again only CC(spin-orb) calculations have been carried out because of the rather small difference in performance of CC(spin-orb) and CC(orb). Table 3-2 The vertical electron affinity of CN. EA / E h EA - EA FCI CCSD(spin-orb) aug’-cc-pVDZ 0.13025 0.00063 CCSDT(spin-orb) aug’-cc-pVDZ 0.12977 0.00014 CCSDTQ(spin-orb) aug’-cc-pVDZ 0.12966 0.00003 FCI aug’-cc-pVDZ 0.12962 --- The convergence is remarkable; already at the CCSD level we are down to an error of 0.5% of the FCI value, on the CCSDT level it is 0.1% and on the CCSDTQ level 0.02%. The reason for the excellent convergence is found in a cancellation of errors that influence the result. The deviations of the individual energies are always roughly an order of magnitude larger than the deviation of the affinity, 75 but the errors cancel when the CN and CN - energies are subtracted. That the convergence is from above is also noteworthy. This is because the CC hierarchy converges faster for CN - than for 82
Numerical Results CN. This seems surprising since CN - contains one more electron than CN, but it could be explained by CN - being more one-configurational than CN. 3.3.5 The Equilibrium Geometry of CCH The equilibrium geometry of CCH found from FCI/cc-pVDZ calculations is used in ref. 76 to calibrate coupled cluster calculations in larger basis sets. The FCI correction is assumed to be independent of basis set. To optimize for the two variables R(CC) and R(CH), the CCH radical is assumed linear and the CC and CH bonds are then distorted in step-lengths of δ = 0.01Å from an initial geometry making a grid of single-point calculations around the equilibrium geometry with R(CC) on the one axis and R(CH) on the other. The initial geometry is taken from a CCSDT cc-pVDZ study 76 , the geometry being R CCSDT (CC) = 1.23448Å and R CCSDT (CH) = 1.07924Å. The resulting potential energy surface is seen in Fig. 3.5. -76.4020 -76.4024 E FCI / E h -76.4028 -76.4032 -76.4036 1.09924 1.08924 1.07924 1.06924 R (C-H)/Å 1.21448 1.22448 1.23448 R (C-C)/Å 1.24448 1.25448 1.05924 Fig. 3.5 The potential energy surface of CCH. From finite-difference expressions with the error being of the order δ 4 , the gradient and Hessian are found for the initial geometry and a Newton step is taken giving an improved guess for the equilibrium geometry. The FCI equilibrium geometry is thus found as FCI CCSDT −1 R = R −H G, (3.1) where G is the gradient, H the Hessian, and R CCSDT the CCSDT geometry. The equilibrium geometry at the FCI level is found to be 83
Page 1:
PhD Thesis Optimization of Densitie
Page 5 and 6:
Contents Preface ..................
Page 7:
Summary............................
Page 11:
List of Publications This thesis in
Page 14 and 15:
Part 1 Improving Self-consistent Fi
Page 16 and 17:
Page 18 and 19:
Page 20 and 21:
Page 22 and 23:
Page 24 and 25:
Page 26 and 27:
Page 28 and 29:
Page 30 and 31:
Page 32 and 33:
Page 34 and 35:
Page 36 and 37:
Page 38 and 39:
Page 40 and 41:
Page 42 and 43:
Page 44 and 45: Part 1 Improving Self-consistent Fi
Page 72 and 73: Part 2 Atomic Orbital Based Respons
Page 90 and 91: Part 3 Benchmarking for Radicals ar
Page 92 and 93: Part 3 Benchmarking for Radicals le
Page 96 and 97: Part 3 Benchmarking for Radicals R
Page 99: Summary The developments in compute
Page 103: Appendix A The Derivatives of the D
Page 106 and 107: Thus, the density element D µν is
Page 109 and 110: References 1 2 3 4 5 6 7 8 9 C. C.
Page 111: 65 T. Helgaker and P. Jørgensen, T
Page 115 and 116: JOURNAL OF CHEMICAL PHYSICS VOLUME
Page 117 and 118: 18 J. Chem. Phys., Vol. 121, No. 1,
Page 127: Part 1 The Trust-region Self-consis
Page 130 and 131: 074103-2 Thøgersen et al. J. Chem.
Page 138 and 139: 074103-10 Thøgersen et al. J. Chem
Page 144 and 145:
074103-16 Thøgersen et al. J. Chem
Page 147:
Part 3 A Coupled Cluster and Full C
Page 150 and 151:
L. Thøgersen, J. Olsen / Chemical
Page 152 and 153:
Page 154 and 155:
Page 156 and 157:
Page 159 and 160:
3030 J. Phys. Chem. A 2004, 108, 30
Page 161 and 162:
3032 J. Phys. Chem. A, Vol. 108, No
Page 163:
3034 J. Phys. Chem. A, Vol. 108, No
show all

Get my PhD Thesis

Create successful ePaper yourself

Delete template?

Save as template?