This action might not be possible to undo. Are you sure you want to continue?

From: http://www.gaussian.com/iops.htm

Nov. 15 2005

Index

Overlay 1 IOp(1/5) IOp(1/6) IOp(1/7) IOp(1/8) IOp(1/9) IOp(1/10) IOp(1/11) IOp(1/12) IOp(1/13) IOp(1/14) IOp(1/15) IOp(1/16) IOp(1/17) IOp(1/18) IOp(1/19) IOp(1/20) IOp(1/21) IOp(1/22) IOp(1/23) IOp(1/24) IOp(1/25) IOp(1/26) IOp(1/27) IOp(1/28) IOp(1/29) IOp(1/30) IOp(1/32) IOp(1/33) IOp(1/34) IOp(1/35) 1 1 1 1 1 2 3 3 4 4 4 4 5 5 5 5 6 6 6 7 7 7 7 7 8 8 9 9 9 9 9

IOp(1/36) IOp(1/37) IOp(1/38) IOp(1/39) IOp(1/40) IOp(1/41) IOp(1/42) IOp(1/43) IOp(1/44) IOp(1/45) IOp(1/46) IOp(1/47) IOp(1/48) IOp(1/49) IOp(1/52) IOp(1/53) IOp(1/54) IOp(1/55) IOp(1/56) IOp(1/57) IOp(1/58) IOp(1/59) IOp(1/60) IOp(1/61) IOp(1/62) IOp(1/63) IOp(1/64) IOp(1/65) IOp(1/66) IOp(1/67) IOp(1/70) IOp(1/71) IOp(1/72)

10 10 10 10 11 11 11 11 12 12 12 13 13 13 13 13 14 14 14 14 15 15 15 15 15 16 16 16 17 17 17 17 17

i

IOp(1/73) IOp(1/74) IOp(1/75) IOp(1/76) IOp(1/77) IOp(1/78) IOp(1/79) IOp(1/80) IOp(1/81) IOp(1/82) IOp(1/83) IOp(1/84) IOp(1/85) IOp(1/86) IOp(1/87) IOp(1/88) IOp(1/89) IOp(1/90) IOp(1/91) IOp(1/92) IOp(1/94. 102. 104) IOp(1/105) IOp(1/106) IOp(1/107) IOp(1/108) IOp(1/109) IOp(1/110) IOp(1/111) IOp(1/112) IOp(1/113) IOp(1/114) Overlay 2 ii 18 18 18 19 19 19 19 19 20 20 20 20 20 20 21 21 21 21 22 22 22 22 22 23 23 23 23 24 24 24 24 24 25 IOp(2/9) IOp(2/10) IOp(2/11) IOp(2/12) IOp(2/13) IOp(2/14) IOp(2/15) IOp(2/16) IOp(2/17) IOp(2/18) IOp(2/19) IOp(2/20) IOp(2/29) IOp(2/30) IOp(2/40) IOp(2/41) Overlay 3 IOp(3/5) IOp(3/6) IOp(3/7) IOp(3/8) IOp(3/9) IOp(3/10) IOp(3/11) IOp(3/12) IOp(3/13) IOp(3/14) IOp(3/15) IOp(3/16) IOp(3/17) IOp(3/18) IOp(3/19) IOp(3/20) 25 25 25 25 26 26 26 27 27 27 27 27 28 28 28 28 29 29 30 31 32 32 32 34 34 34 34 35 35 35 36 36 36 . 96. 103. 95. 97. 98) IOp(1/101.

IOp(3/21) IOp(3/22) IOp(3/23) IOp(3/24) IOp(3/25) IOp(3/26) IOp(3/27) IOp(3/28) IOp(3/29) IOp(3/30) IOp(3/31) IOp(3/32) IOp(3/33) IOp(3/34) IOp(3/35) IOp(3/36) IOp(3/37) IOp(3/38) IOp(3/39) IOp(3/40) IOp(3/41) IOp(3/42) IOp(3/43) IOp(3/44) IOp(3/45) IOp(3/46) IOp(3/47) IOp(3/48) IOp(3/49) IOp(3/51) IOp(3/52) IOp(3/53) IOp(3/54) 36 37 37 37 37 37 38 38 38 38 38 38 39 39 39 39 40 40 40 40 40 41 41 42 42 42 42 43 44 44 44 44 45 IOp(3/55) IOp(3/56) IOp(3/57) IOp(3/58) IOp(3/59) IOp(3/60) IOp(3/61) IOp(3/62) IOp(3/63) IOp(3/64) IOp(3/65) IOp(3/66) IOp(3/67) IOp(3/70) IOp(3/71) IOp(3/72) IOp(3/73) IOp(3/74) IOp(3/75) IOp(3/76) IOp(3/77) IOp(3/78) IOp(3/79) IOp(3/80) IOp(3/81) IOp(3/82) IOp(3/83) IOp(3/84) IOp(3/85) IOp(3/86) IOp(3/87) IOp(3/88) IOp(3/89) 45 45 45 45 45 46 46 46 46 46 46 47 47 47 47 48 48 48 49 49 50 50 50 50 51 51 51 51 51 52 52 52 52 iii .

IOp(3/90) IOp(3/91) IOp(3/92) IOp(3/93) IOp(3/94) IOp(3/95) IOp(3/96) IOp(3/97) IOp(3/98) IOp(3/99) IOp(3/100) IOp(3/101) IOp(3/102) IOp(3/103) IOp(3/104) IOp(3/105) IOp(3/106) Overlay 4 IOp(4/5) IOp(4/6) IOp(4/7) IOp(4/8) IOp(4/9) IOp(4/10) IOp(4/11) IOp(4/13) IOp(4/14) IOp(4/15) IOp(4/16) IOp(4/17) IOp(4/18) IOp(4/19) IOp(4/20) iv 53 53 53 53 54 54 54 54 54 54 55 55 55 55 55 55 56 57 57 57 58 58 58 59 59 60 60 60 60 60 61 61 61 IOp(4/21) IOp(4/22) IOp(4/23) IOp(4/24) IOp(4/25) IOp(4/26) IOp(4/28) IOp(4/29) IOp(4/31) IOp(4/33) IOp(4/34) IOp(4/35) IOp(4/36) IOp(4/37) IOp(4/38) IOp(4/43) IOp(4/44) IOp(4/45) IOp(4/46) IOp(4/47) IOp(4/48) IOp(60-62) IOp(4/63) IOp(4/64) IOp(4/65) IOp(4/66) IOp(4/67) IOp(4/68) IOp(4/69) IOp(4/71) IOp(4/72) IOp(4/80) IOp(4/81) 61 62 62 62 62 63 63 63 63 63 63 64 64 64 64 64 64 65 65 65 65 65 66 66 66 66 66 67 67 67 67 67 68 .

IOp(4/82) IOp(4/110) Overlay 5 IOp(5/5) IOp(5/6) IOp(5/7) IOp(5/8) IOp(5/9) IOp(5/10) IOp(5/11) IOp(5/12) IOp(5/13) IOp(5/14) IOp(5/15) IOp(5/16) IOp(5/17) IOp(5/18) IOp(5/19) IOp(5/20) IOp(5/21) IOp(5/22) IOp(5/23) IOp(5/24) IOp(5/25) IOp(5/26) IOp(5/27) IOp(5/28) IOp(5/29) IOp(5/30) IOp(5/31) IOp(5/32) IOp(5/33) IOp(5/34) 68 68 69 69 70 70 71 71 72 72 72 72 73 74 74 75 76 76 77 77 77 78 78 78 78 79 79 79 79 80 80 80 80 IOp(5/35) IOp(5/36) IOp(5/37) IOp(5/38) IOp(5/39) IOp(5/40) IOp(5/41) IOp(5/42) IOp(5/43) IOp(5/45) IOp(5/47) IOp(5/48) IOp(5/49) IOp(5/50) IOp(5/51) IOp(5/52) IOp(5/53) IOp(5/55) IOp(5/56) IOp(5/57) IOp(5/55-58) IOp(5/59) IOp(5/60-62) IOp(5/63) IOp(5/64) IOp(5/65) IOp(5/70) IOp(5/71) IOp(5/73) IOp(5/74) IOp(5/75-78) IOp(5/79) IOp(5/80) 81 81 81 81 81 82 82 83 83 83 83 83 84 84 84 84 85 85 85 85 85 86 86 86 86 87 87 87 87 87 88 88 88 v .

IOp(5/81) IOp(5/82) IOp(5/83) IOp(5/84) IOp(5/85) IOp(5/86) IOp(5/87) IOp(5/88) IOp(5/89) IOp(5/90) IOp(5/91) IOp(5/92) IOp(5/93) IOp(5/94) IOp(5/95) IOp(5/96) IOp(5/97) IOp(5/98) IOp(5/99) IOp(5/100) IOp(5/101) IOp(5/102) Overlay 6 IOp(6/15) IOp(6/16) IOp(6/17) IOp(6/18) IOp(6/19) IOp(6/20) IOp(6/21) IOp(6/22) IOp(6/23) IOp(6/24) vi 88 88 89 89 89 89 90 90 90 90 90 90 91 92 92 92 93 93 93 93 94 94 95 95 96 96 96 96 96 97 97 97 97 IOp(6/25) IOp(6/26) IOp(6/27) IOp(6/28) IOp(6/29) IOp(6/30) IOp(6/31) IOp(6/32) IOp(6/35) IOp(6/36) IOp(6/37) IOp(6/38) IOp(6/39) IOp(6/40) IOp(6/41) IOp(6/42) IOp(6/43) IOp(6/44) IOp(6/45) IOp(6/46) IOp(6/47) IOp(6/48) IOp(6/49) IOp(6/51) IOp(6/52) IOp(6/53) IOp(6/54) IOp(6/55) IOp(6/56) IOp(6/57) IOp(6/58) IOp(6/59) IOp(6/60-62) 98 98 98 98 98 99 99 99 99 100 100 100 101 101 101 102 102 102 102 102 103 103 103 103 103 104 104 104 104 104 104 105 105 .

IOp(6/63) IOp(6/64) IOp(6/65) IOp(6/70) IOp(6/71) IOp(6/72) IOp(6/73) IOp(6/74) IOp(6/75) IOp(6/76) IOp(6/77) IOp(6/78) IOp(6/79) IOp(6/80) IOp(6/81) IOp(6/82) Overlay 7 IOp(7/6) IOp(7/7) IOp(7/8) IOp(7/9) IOp(7/10) IOp(7/11) IOp(7/12) IOp(7/13) IOp(7/14) IOp(7/15) IOp(7/16) IOp(7/18) IOp(7/25) IOp(7/28) IOp(7/29) IOp(7/30) 105 105 105 105 106 106 106 106 106 106 107 107 107 107 107 107 109 109 109 109 109 110 110 110 110 111 111 111 111 112 112 112 112 IOp(7/31) IOp(7/32) IOp(7/40) IOp(7/41) IOp(7/42) IOp(7/43) IOp(7/44) IOp(7/45) IOp(7/52) IOp(7/53) IOp(7/60-62) IOp(7/63) IOp(7/64) IOp(7/65) IOp(7/70) IOp(7/71) IOp(7/72) IOp(7/74) IOp(7/75) IOp(7/76) IOp(7/77) IOp(7/87) Overlay 8 IOp(8/5) IOp(8/6) IOp(8/7) IOp(8/8) IOp(8/9) IOp(8/10) IOp(8/11) IOp(8/12) IOp(8/13) IOp(8/14) 112 113 113 113 113 113 114 114 114 114 114 115 115 115 115 115 115 116 116 116 116 116 117 117 117 117 117 117 118 119 119 119 119 vii .

IOp(8/15) IOp(8/16) IOp(8/17) IOp(8/18) IOp(8/19) IOp(8/20) IOp(8/22) IOp(8/27) IOp(8/28) IOp(8/29) IOp(8/30) IOp(8/31) IOp(8/32) IOp(8/35) IOp(8/36) IOp(8/38) IOp(8/39) IOp(8/40) IOp(8/41) IOp(8/42) IOp(8/43) IOp(8/44) IOp(8/45) IOp(8/46) IOp(8/47) Overlay 9 IOp(9/5) IOp(9/6) IOp(9/7) IOp(9/8) IOp(9/9) IOp(9/10) IOp(9/11) viii 119 120 120 120 120 120 121 121 121 121 122 122 122 122 122 123 123 123 123 123 124 124 124 124 125 126 126 126 126 127 127 127 128 IOp(9/12) IOp(9/13) IOp(9/14) IOp(9/15) IOp(9/16) IOp(9/17) IOp(9/18) IOp(9/19) IOp(9/20) IOp(9/21) IOp(9/22) IOp(9/23) IOp(9/25) IOp(9/26) IOp(9/27) IOp(9/28) IOp(9/30) IOp(9/31) IOp(9/36) IOp(9/37) IOp(9/38) IOp(9/40) IOp(9/41) IOp(9/42) IOp(9/43) IOp(9/44) IOp(9/45) IOp(9/46) IOp(9/47) IOp(9/48) IOp(9/49) IOp(9/60-62) IOp(9/70) 128 128 128 129 129 129 129 130 130 130 130 130 131 131 132 132 132 132 132 132 133 133 133 134 134 134 135 135 135 135 135 136 136 .

IOp(9/71) IOp(9/72) IOp(9/73) IOp(9/74) IOp(9/75) IOp(9/81) IOp(9/82) IOp(9/83) IOp(9/84) IOp(9/85) IOp(9/86) Overlay 10 IOp(10/5) IOp(10/6) IOp(10/7) IOp(10/8) IOp(10/9) IOp(10/10) IOp(10/11) IOp(10/13) IOp(10/14) IOp(10/15) IOp(10/16) IOp(10/17) IOp(10/18) IOp(10/19) IOp(10/20) IOp(10/21) IOp(10/22) IOp(10/28) IOp(10/29) IOp(10/30) IOp(10/31) 136 136 136 136 136 137 137 137 137 137 138 139 139 139 140 140 140 140 140 140 141 141 141 141 142 142 142 142 143 143 143 143 143 IOp(10/32) IOp(10/45) IOp(10/46) IOp(10/47) IOp(10/48) IOp(10/60-62) IOp(10/63) IOp(10/72) IOp(10/73) IOp(10/74) IOp(10/75) IOp(10/76) IOp(10/77) IOp(10/78) IOp(10/79) Overlay 11 IOp(11/5) IOp(11/6) IOp(11/7) IOp(11/8) IOp(11/9) IOp(11/10) IOp(11/11) IOp(11/12) IOp(11/13) IOp(11/14) IOp(11/15) IOp(11/16) IOp(11/17) IOp(11/18) IOp(11/19) IOp(11/20) IOp(11/21) 144 144 144 144 145 145 145 145 145 146 146 146 146 146 147 148 148 148 148 148 148 149 149 149 150 150 150 150 151 151 151 151 151 ix .

IOp(11/22) IOp(11/23) IOp(11/24) IOp(11/26) IOp(11/27) IOp(11/28) IOp(11/29) IOp(11/30) IOp(11/31) IOp(11/32) IOp(11/33) IOp(11/39) IOp(11/40) IOp(11/41) IOp(11/42) IOp(11/43) IOp(11/45) IOp(11/46) IOp(11/53) IOp(11/60-62) IOp(11/63) IOp(11/70) IOp(11/71) IOp(11/75) Overlay 9999 IOp(9999/5) IOp(9999/6) IOp(9999/7) IOp(9999/8) IOp(9999/9) IOp(9999/10) IOp(9999/11) IOp(9999/12) x 152 152 152 152 152 152 153 153 153 153 154 154 154 154 154 155 155 155 155 155 155 156 156 156 157 157 157 157 158 158 158 158 159 IOp(9999/13) IOp(9999/14) IOp(9999/15) IOp(9999/16) IOp(9999/17) IOp(9999/18) IOp(9999/33) 159 159 159 159 159 159 160 .

L113.Overlay 1 IOp(1/5) L103 MODE OF OPTIMIZATION 0 1 N FIND LOCAL MINIMUM FIND A SADDLE POINT FIND A STATIONARY POINT ON THE ENERGY SURFACE WITH N NEGATIVE EIGENVALUES OF THE 2ND DERIVATIVE MATRIX L107: MODE OF SEARCH 0 1 LOCATE THE MAXIMUM IN THE LST PATH. L109) = Min(40. L112: MAXIMUM STEP SIZE ALLOWED DURING OPT.LT. SCAN THE LST PATH.0003 HARTREE/BOHR OR RADIAN CONVF = N*10**-6 L116. L112. L113. 10**-7 in L117. L112) = Min(20. 10**-N.LT.NVAR+10) (L102. L109. 0 NSTEP = Max(20. IOp(1/6) L102. 1 . L114: CONVERGENCE ON THE FIRST DERIVATIVE AND ESTIMATED DISPLACEMENT FOR THE OPTIMIZATION RMS FIRST DERIVATIVE . L114: MAXIMUM NUMBER OF STEPS (OR NUMBER OF STEPS FOR AN LST SCAN). CONFV. L109.NVAR+10) (L103. L109. Default value for single-points: 10**-5 in L116. L114) N NSTEP = N IOp(1/7) L103. L103.NVar+20) (L113. IOp(1/8) L103. L107. L117: Convergence on electric field/charges -1 0 N Default value for optimizations: 10**-7. DISPLACEMENT . L105. L112. CONVX=4*CONVF -1 0 N ConvF = 1/600 HARTREE/BOHR OR RADIAN CONVF = 0. L105. RMS EST. L105.

0 1 2 00 10 20 Whether to update trust radius (DXMaxT. max is 9. Use MOs.Do-Self Polarization. Number of small steps per ABM step to be used in starting ABM and when "slow down" is needed Which approximation to make.01 * N L117: General control. L107: WHETHER TO MAINTAIN SYMMETRY ALONG THE SEARCH PATH. Default is to use density. N DXMAXT = 0.Do III and Allow Surface To "Relax" in Solution if no spheres Whether to evaluate densities using orbitals or density matrix. Apprx. Default is 5. Yes. Default is Yes for minima.3 Bohr or Radian (L103. Apprx.Do Self-Polarization and Compensation. Which type of basin to use to partition the density isosurface. Whether to scale or search the sphere when reducing the step size to the trust radius (Default Scale. L117: Whether to delete points which are too close together: 0 2 No .0 DXMAXT = 0. Default is 4 GradVne GradRho Don't Use Basins.1 IOp(1/9) L103: Use of Trust radius. IV . = 0. = 0. III . Use only the Center of NuclearCharge Use Interlocking Spheres Order of Adam's-Bashforth-Moulton (ABM) predictor-corrector method to use in solving diff. in ABM.). I .2 Bohr or Radian (L105). NO.Don't Do Self-Polarization or "Compensation" Apprx. = 0. No. 1000 2000 3000 4000 N0000 10000 20000 Apprx. L121: Time step. II .0001 fs. N*0. Default is 4. default Yes). Use density.3 Bohr or Radian (L113. Estm or UnitFC). L114). default 0. Default is III for Tomasi (interlocking spheres) and IV for general for the grad RHO or Vne trajectories. no for TS. But No Compensation. Read or CalcFC). search for minima. eqns. scale for transition states. 0 1 YES. Search.1 BOHR OR RADIAN (L103. 0 1 2 3 4 N0 N00 N000 surface.

IOp(1/11) L103: TEST OF CURVATURE. Approx 1.05 Angstroms) Yes. 0 1 2 3 4 5 6 7 8 9 10 Use defaults (not valid for L109). coordinates. End with a blank card. using a (10**-N Angstroms) criteria. Bohr. Estimate force constants using valence force field.J.FC(I. only recognized by 103).001) Step size for ABM method in Trudge for isodensity method. Use semiempirical force constants. L105. 3 . Use unit matrix (default for L105.I=1. L113.NVAR) (8F10.6) (L103 only). L112.J).J). L117: Scaling Factor for Determining Overlaps of VDW atoms -1 0 N Turn off scaling Default is 1. and radians).000 + N*(0. Read from checkpoint file in internal coordinates.010 1. (5I3. 0 1 2 DEFAULT (TEST for z-matrix or cartesian TS but not for LST/QST or for minimum). Read cartesian forces and force constants from the checkpoint file are convert to internal Read cartesian forces followed by cartesian force constants (both in format 6F12.0) (L103 only). How close to get to the isosurface in search.J=1.0**-N L121: Whether to read in initial velocities: 0 1 2 3 Default (same as 1) Generate random initial velocity Read in initial cartesian velocity (Bohr/sec) Read in initial MW cartesian velocity (sqrt(amu)*Bohr/sec) IOp(1/10) L103. DON'T TEST.1 -N 0 N Yes. L109. (not valid for L109).8) from input stream. Read I. L114: Input of initial Hessian: All values must be in atomic units (Hartree. Read ((FC(I. BOMB THE JOB IF THE SECOND SECOND DERIVATIVE MATRIX HAS THE WRONG NUMBER OF NEGATIVE EIGENVALUES.F20. Use unit matrix throughout. followed by a blank line. using a default criteria (0.0D-6 (N=20) 2. Second derivative matrix calculated analytically. TEST.I).

7 for L103 TS. D2Corr and L115. L121: Multi-time step parameter (NDtrC. Powell (not in L103). safeguarding positive definateness (not inL103 or L115) D2Corr (New.1/N IOp(1/12) L103: OPTIMIZATION CONTROL PARAMETERS 0 1 USE DEFAULT VALUES READ IN NEW VALUES FOR ALL PARAMETERS (SEE INITBS) IOp(1/13) L103.NDtrP) 0 NN MM00 No multi-time stepping Iterate density constraints NN times per step Do gradient once every MM steps IOp(1/14) L103: Max. only in L103 and L115).L115: Type of Hessian Update: 0 1 2 3 4 5 6 7 8 9 Default (9 for L103 minimization.L114.0 N 0. no quadratic step). 0 N Default (0 for TS.1 * N 4 . Allow N -.L113. BFGS (not in L103) BFGS. otherwise BFGS). only in L103 and L115). Powell for L113 and L114).05 (N=2) 0.L109: ABORT IF DERIVATIVES TOO LARGE -1 or 0 N No force test at all. use initial Hessian). D2Corr (BFGS) D2Corr (Bofill Powell+MS for transition states).linear only starts with the N+1st. FMAXT = 0. number of bad steps to allow before attempting a linear minimization (i. D2Corr (New if energy rises. 1 for minima). D2Corr (Old.e. D2Corr (No update. IOp(1/15) L103..

IOp(1/16) L103. Include MM atoms in internal coordinate definitions (no microiterations). AND PROCESSING CONTINUES.L114: MINIMUM ALLOWABLE MAGNITUDE OF THE EIGENVALUES OF THE SECOND DERIVATIVE MATRIX. Skip MM atoms in internal coordinate definitions and do microiterations the old way. Skip MM atoms in internal coordinate definitions and do microiterations the new way. Read the AddRedundant input section for each structure. No optimization Do optimization in cartesian coordinates. THE SIZE OF THE EIGENVALUE IS REDUCED TO THE MAXIMUM. / N IOp(1/18) L103: Coordinate system. Microiterations for pure MM. 0 N EIGMAX = 25. in L120. will occur.1 * N IOp(1/17) L103. Do full optimization in redundant internal coord.0001 EIGMIN = 1. Do optimization in Z-matrix coordinates.L114: MAXIMUM ALLOWABLE MAGNITUDE OF THE EIGENVALUESOF THE SECOND DERIVATIVE MATRIX.L113. done in L402.L113. 0 1 10 20 30 40 50 100 1000 2000 3000 10000 20000 100000 0000000 1000000 2000000 3000000 4000000 Proceed normally Second derivatives will be computed as directed on the variable definition cards. Do not define H-bonds Define H-bonds with no related coordinates (default) Define H-bonds and related coordinates Reduce the number of redundant internals Define all redundant internals Old definition of redundant internals. IF THE LIMIT IS EXCEEDED.0 HARTREE / BOHR**2 OR RADIAN**2 EIGMAX = 0. IOp(1/19) L103: SEARCH SELECTION 5 . Do full optimization in redundant internal coords with large molecular tools. SIMMILAR TO IOp(16) 0 N EIGMIN = 0. in L103. Default (2000000). Do full optimization in pruned distance matrix coords.

0 1 NORMAL MODE. Read in a re-ordering vector from the input. Linear as usual. L106. STEEPEST DESCENT AND LINEAR ONLY WHEN ESSENTIAL. LINEAR AND STEEPEST DESCENT.L114: EXPERT SWITCH. EXPERT MODE: CERTAIN CUTOFFS USED TO CONTROL THE OPTIMIZATION WILL BE RELAXED. DXMAXT. No linear search. RFO without linear.L114: Search Selection: 0 1 P-RFO OR RFO STEP ONLY (DEFAULT) P-RFO OR RFO STEP FOR "WRONG" HESSIAN OTHERWISE NEWTON-RAPHSON IOp(1/20) L101. L113. Newton-Raphson only. 0 1 2 Yes. L109. EIGMAX AND EIGMIN. L108. RFO and linear. GDIIS and linear GDIIS only. Quadratic if curvature is correct.L113. Assume reactant order equals product order. RFO if not. First-order simultaneous optimization. L110: INPUT UNITS 0 1 2 3 ANGSTROMS DEGREES BOHRS BOHRS DEGREES RADIANS ANGSTROMS RADIANS IOp(1/21) L103. L115: KIND OF SEARCH: 0 1 6 BOTH DIRECTIONS AND GENERATE SEARCH VECTOR FORWARD DIRECTION AND GENERATE S. Quadratic if curvature is correct. Newton-Raphson and linear. IOp(1/22) L107: Whether to reorder coordinates for maximum coincidence.0 2 3 4 5 6 7 8 9 10 11 13 Default (same as 6). VECTOR . RFO if not. THESE INCLUDE FMAXT.

Energy + Forces + Force constants IOp(1/24) Whether to round tetrahedral angles. -1 0 1 2 3 Read in values Default (same as 1). round angles within 0. VECTOR 8F10. gradient 1. No. VECTOR 8F10. 0 1 2 Energy only.6 FORWARD DIRECTION AND READ S.d-7).d-7. 1000*I+100*J+10*K+L) 7 .d-6) IOp(1/27) = IJKL (i. Energy + Forces. Normal accuracy for HF (energy and gradient both 1.001 degree.d-5. VECTOR 8F10. Gradient 10**-(MM).d-4) Fine grid accuracy for DFT (Energy 1. gradient 1. VECTOR FORWARD DIRECTION AND READ S. Standard grid accuracy for DFT (Energy 1. Yes.6 IOp(1/23) L112: Derivative availability.e. IOp(1/25) Wether SCRF is used with numerical polarizability: 0 1 No. Yes.6 BOTH DIRECTIONS AND READ S. 0 1 2 Default (Yes). VECTOR BOTH DIRECTIONS AND GENERATE S. IOp(1/26) Accuracy of function being optimized: -NNMM Energy 10**-(NN). VECTOR 8F10.2 3 4 5 6 7 BACKWARD DIRECTION AND GENERATE S. the field in /Gen/ must be cleared each time.6 BACKWARD DIRECTION AND READ S.

same as -1. Normal (6 or 5 for linear molecules).1*J*R + 0. The values of the TS coord are estimated by interpolating the sturcture of R and P. The union without QST. QSTRad = 0. of the two redundant coordinates are taken as the redundant coords for the TS. 7 10 000 100 200 1000 2000 8 Get all input (title. K = 1-9 J = 1 J = 2 J = 3 J = 4 I = 0-9 Interpolation of initial guess of TS between R and P (TS=0. By model builder. Set all optimization flags to optimize. GET CARTESIAN COORDINATES ONLY FROM THE CHECKPOINT FILE. GET Z-MATRIX AND VARIABLES FROM THE CHECKPOINT FILE. Read optimization flags in format 50L1 after the z-matrix. R and P are used to guide the QST optimization of the TS.g.1*(10-J)*P. The third one is the initial guess of the transition structure.1 L= 2 Input one structure. N. Do not abort job if model builder generates a z-matrix with too many variables. Default. L= 3 TS. charge and multiplicity. default J=5) LST constraint in internals QST constraint in internals LST constraint in distance matrix space QST constraint in distance matrix space Control parameters for climbing phase of QST (e. model B. default QSTrad = 0. either initial guess of the minimizing structure or transition structure Input 2 structures. The first one is reactant the second one is the product. R and P are used to guide the QST optimization of the IOp(1/28) L103: Number of translations and rotations to remove during redundant coordinate transformations: -2 -1 0 N 0. Print details of the model building process.Transition state searching using QST and redundant internal coordinates L= 0. By model builder. the first one is the reactant. Default (same as 100). structure) from the checkpoint file. BY Z-MATRIX BY DIRECT COORDINATE INPUT (must set IOp(29) in L202). but read new values for some variables from the input . Get Z-matrix from the checkpoint file. model A.05) Input 3 structures. Abort job if model builder generates a z-matrix with too many variables.01*I. the second one is the product. IOp(1/29) L101: SPECIFICATION OF NUCLEAR CENTERS 0 1 2 3 4 5 6 stream.

IOp(1/30) L103: ARE THE READ-WRITE FILES TO BE UPDATED? THIS OPTION IS SET FOR THE LAST CALL TO 103 IN FREQUENCY CALCULATIONS IN ORDER TO PRESERVE THE VALUES OF THE VARIABLES FOR ARCHIVING. Rebuild the coordinate system. hack to make IRCs work with Cartesian input).3000 4000 5000 00000 10000 20000 100000 200000 Purge flags except the frozen variables. Default. Mark Z-matrix constants as frozen variables rather than wired-in constants. (2+3) Purge all flags but keep the coordinate definition. 0 1 YES NO IOp(1/32) TITLE CARD PUNCH CONTROL. 0 1 DON'T PUNCH. PUNCH. It also suppresses error termination on large gradients. Do not retain symbolic constants. 9 . but retain the redudant internal coordinate definitions. same as 10000. IOp(1/33) L101: L102 L103 L106 L109 L110 L113 L114 0 1 OFF ON DEBUG PRINT IOp(1/34) L101 L102 L103: DEBUG + DUMP PRINT 0 1 OFF ON IOp(1/35) RESTART (L102-L112). Generate a symbolic z-matrix using all Cartesians if none is present on the checkpoint file (a Same as one.

0 N -1 -N>1 Use internal default (0. Path step size in L115. FIRST POINT OF A RESTART.0001*N (angstroms in L106. 0. electric field au in L111). 10 . 0 1 NORMAL CHECKPOINT OF OPTIMIZATION. In L110 and L111: differentiate energy N times. SUPPRESS CHECKPOINTING. 1 for others). L110. L107. In L103: Initial entry of guided optimization using N levels. IOp(1/38) Entry control option (currently only by L106. 0 1 N>1 . 0. IOp(1/36) CHECKPOINT.01 Angstrom in L110.0 1 NORMAL OPTIMIZATION. and L105). In L106: differentiate wrt electric field. (L106. GET GEOMETRY.001 Use step-size of 0. L111. IOp(1/39) Step size control for numerical differentiation. Read stepsize (up to 2 for L106. In L106: differentiate Nth derivatives once. Initial entry. L110. L108. FROM THE CHECKPOINT FILE. WAVEFUNCTION. In L106: differentiate wrt nuclear coordinates. Use step-size of 0. L109. L103. L109. In L106: differentiate wrt field and nuclear.0001*N atomic units everywhere. L109. ET. THIS IS THE LAST POINT AT WHICH ANALYTIC SECOND DERIVATIVES WILL BE DONE.005 A in L109. N0 000 100 200 Continuation of run. and L112 but not L102. DELETE THE D2E FILE AND THE BUCKETS AND TRUNCATE THE READ/WRITE FILES. IOp(1/37) D2E CLEANUP (obsolete) 0 1 NO CLEANUP. 0. au in L111). free-format. L110. L111).001 Angstroms in L106.

Just update. terminated Read in replacment radii for selected atoms as pairs (I.Rad).IOp(1/40) L113. L114: Hessian recalculation. 0 1 2 3 Dipole Quadrupole Octapole Hexadecapole L117: How to define Radii 0 1 2 10 20 30 100 200 Default is 11 Use internally stored Radii.Rad) or (Symbol. Initial 11 by a blank line. IOp(1/41) Step number of optimization from which to take geometry. L116: Whether to read initial E-field: 0 1 2 Start with 0.0D-10 1. Read in replacement radii for selected atom types as pairs (IAn. L115: Number of points along the reaction path in each direction. 0 N 1. Force CHELPG (Breneman) recommended radii. Read from checkpoint file. .0. L117: Cutoff to be used in evaluating densities. centers will be on atoms Read-in centers and radii on cards Force Merz-Kollman radii (Default) Force CHELP (Francl) recommended radii. Read from input stream. -1 0 N Pick up analytic second derivatives every time.0D-N IOp(1/43) L116: Extent of Reaction Field. Default is 6. Recalculation the Hessian every N steps. The default. -1 for the initial geometry IOp(1/42) L103. execpt for CalcAll.Rad). terminated by a blank line.

points which clearly lie in another basin. 0 NM 0. This parameter should be chosen with the parameter Cont in mind 0 NM 10.radius of spheres to be placed around attractors to "capture" the gradient trajectories.1 N. from the outset.used in Trudge only to eliminate. Maximum distance between a nucleus and its portion of the isosurface .M = NM/10 IOp(1/44) IRC Type 0 1 2 3 L117: Default (same as 3).M au = NM/10 L121: Seed for random number generator (ISeed) -1 0 N Use system time initialize iseed (Note each run will give different results) Use default seed value (ISeed = 398465) Set random number seed to N IOp(1/45) Read isotopes in L115. 12 . Read Isotopes.0 au N. Mass-weighted. IOp(1/46) Order of multipoles in numerical SCRF: 0 1 2 3 Dipole Quadrupole Octapole Hexadecapole. Cartesian. 0 1 Do not read isotopes. Internal. The final radius is then automatically optimized separately for each atom.

normal calculation Two layers Three layers Default (20) Include electrostatics in model systems using MM charges. The default is MK charges. IOp(1/53) L120: Action of each invocation of L120: 0 1 2 there.IOp(1/47) Number of redundant internal coordinates to allow for. Default: 1 3/2/1 dont/QST3/QST2 optimize TS structure (for QST input). 0 N Default: 50000 N. No electrostatics included in the model systems Do full square for testing. Default: unimolecular IOp(1/52) L101 and L120: Type of ONIOM calculation: 0/1 2 3 00 10 20 100 N000 One layer. 13 Do nothing Set up point MM on rwf from initial data Set up point MM on rwf from initial data and restore point MM on chk file if ONIOM data is present . IOp(1/48) IRCMax control. 1 20 Do IRCMax Include zero-point energy. 3 Restore point M from data on the rwf. Use atomic charge type N-1 during microiterations. Default: 1 2/1 unimolecular/bimolecular reaction. Default: 1 2/1 dont/do optimize product structure. IOp(1/49) Options to IRC path relaxation (IJKL) L K J I 2/1 dont/do optimize reactant structure.

Yes. L115: IRC optimization. Amber allowing any symbol. gradient. hessian) of point NN of the ONIOM Next point to do is MM. Dreiding/UFF. Calc Level High | | | Mid | | | Low S M L 1--3--6 system size 2--5--8 4--7--9* IOp(1/54) Whether to recover initial energy during IRCMax from chk file: 0 1 No. IOp(1/57) Whether to produce connectivity: 14 . for use with parameters in input stream. Use displacements to find the next geometry.4 5 6 7 NN0 MM000 Integrate energy Integrate energy and gradient Integrate energy. Save necessary information (some rwf's. grid. 0 1 Default. IOp(1/55) L103: Options for GDIIS: ICos*1000+IChkC*100+IMix*10+Method form. Amber. use gradients to find the next geometry. gradients. and hessian Restore point MM from RWF but do not create a new model system. energy. NN = MaxLev**2 + 1 (currently 17) to restore real system. IOp(1/56) Set of atom type names to parse: 0 1 2 3 Accept any.

Connectivity input is in terms of z-matrix entries. IOp(1/60) Interpret extra integer and fp values in z-matrix as scan information. 2 for regular) New versions. No. read from checkpoint file. Yes. 0 1 2 Default (No). 15 . NN NN fragments. otherwise 3). IOp(1/58) IRCMax control in L115. Yes. including dummy atoms. but with NBasis and NBsUse for the high-level calc on the model system. NN < 50. Useful for treating the full system as having electrons only on the QM atoms. IOp(1/61) How ONIOM should leave the rwf at the end of each geomtry: 0 1 2 Default (1).0 1 2 3 4 5 10 100 Default (4 if reading geom from chk file and connectivity is there. read from input stream Yes. IOp(1/59) Update of coordinates in L103 0 1 2 Default (1 for large opt. generate connectivity. IOp(1/62) Counterpoise control. Yes. read from rwf file. Read modifications. Yes. No. Normal: leave the rwf set up for the low-level calculation on the real system. Old version. MOMM: leave the rwf set up for the real system.

UFF. Read modifications to parameter set. hard-wired has priority.2*NFrag -. Use only soft. MMFF (NYI). 2=Gradient. With soft parameters. If IOp(67)=3. corresponding to the 'classes' in FncInf. Use soft and hard-wired. Use the last when there are equivalent matches. Dreiding. then the default is to apply soft parameters with higher priority.lone fragments IOp(1/64) Molecular mechanics force field selection: 0 1 2 3 4 5 6 7 000 100 200 300 0000 1000 00000 10000 20000 None. soft has priority. AMBER. 0 1 10 100 1000 10000 Do all (default) Non-bonded Stretching Bending Torsion Out-of-plane 16 . MM3 (NYI). abort when different parameters match to the same degree. Use soft and hard-wired. IOp(1/65) Control of which terms are included in MM. Lowest 2 digits then have no meaning. Supermolecule Fragments with ghost atoms NFrag+1 . Use the first when there are equivalent matches. Do not read modifications to parameter set. Quartic fitting field (NYI). Use only hard-wired. MM2 (NYI).IOp(1/63) Step in counterpoise calculation: MNN NN = 0 1-NFrag M = order of derivatives (1=Energy.

Generate here. 0 1 2 00 10 20 000 100 200 Default (2. Don't do QEq. IOp(1/70) L118 Type of sampling (Nact) 0 1 2 3 4 Defalt (same as 3) Orthant sampling Microcanonical normal mode sampling Fixed normal mode energy Local mode sampling ( now only Nact = 0 or 3 OK ) IOp(1/71) Whether to print out input files for each structure along an IRC: 0 1 No. Do for all atoms regardless of typing. 0 1 2 3 Default: 2 if reading geom from chk file. IOp(1/72) L103: Algorithm choice for microiterations.IOp(1/66) Whether to generate QEQ charges. Copy from chk file. 17 . else 1. Pick up non-standard parameters from chk file. reading from input if requested by IOp(64). Default (10) Do for atoms which were not explicitly typed. or to use the values already there. Default (100) Do for atoms which have charge specified or defaulted to 0. Yes. Do for all atoms regardless of initial charge. IOp(1/67) Source of MM parameters. over-written the values in AtChMM. 1==> 221) Do QEq.

L121: Lagrangian constrain method for ADMP (ICType) Half*Gamma*Tr[(P*P-P)**2] + Lambda*[Tr(P)-Ne] + Eta*Tr(P*P-P) 0 0) 1 2 3 4 5-7 8-11 Use Lambda and Eta only. (Gamma=0) Use Lambda, Eta, Gamma. Gamma = .2 Use Lambda, Eta, Gamma. Gamma = 1. Constraints for scalar Mass case: Use exact constraint Sum(ij)[Vij*(P**2-P)ij] Iterative Scheme same as 4. Different initial guesses. 7 is default for scalar mass case. Mass-weighting constraints. Documentation maybe found in DVelV1. 10 is default. Default Same as 7 if no Mass-Weighting (IOp(76) < 0) Same as 10 if Mass-Weighting (IOp(76) >

Constraints for tensorial Mass:

IOp(1/73)

L103: NInit for microiterations. L121: Initial Kinetic energy of the Nuclei (EStrtC) 0 N>0 N<0 Default (.1 Hartree) N*micro-Hartree 0.0 Hartree

IOp(1/74)

Charge scaling for charge embedding in ONIOM. IJKLMN 6th through 1st nearest neighbors of current layer scaled by I*0.2, J*0.2, etc. 0 ==> 5 (no scaling); all layers are scaled by at least as much as ones farther out. The default is 500. M L0 K00 Factor for charges one bond away from link atom Factor for charges two bonds away from link atom Factor for charges three bonds away from link atom IJ etc. The actual factors used are: 0: 1.0

1: 0.0 2: 0.2 3: 0.4 4: 0.6 5: 0.8 6-9: 1.0

IOp(1/75)

ADMP control flag (ICntrl) 0 1 2 3 00 10 20 Standard ADMP Read converged density at every step Fix the nuclear coordinates Test time reversability (MaxStp must be even) Default (20). Read stopping parameters from input. Do not read stopping parameters.

18

IOp(1/76)

+/- XXXXZYYYY = Ficticous electron mass (EMass) YYYY IOp(76)>0 IOp(76)<0 Z XXXX BoxMas=0 Default (1000) YYYY*.0001 AMU MW core functions more than valence functions. YYYY*.0001 AMU. Use uniform scaling for all basis functions (Note YYYY > 9999 makes no sense) Mass-weighting option. If IOp(76)<0, Z is meaningless. If PBC: Mass of Box Coordinates (BoxMas) = XXXX*.0001 AMU Box coordinates not propagated (default).

IOp(1/77)

Initial Kinetic energy of the density matrix (EStrtP) (For UHF, Alpha and Beta each get half this energy) and Option Number to compute initial kinetic energy. Format of Input: XXYYYY (six digits) IWType = XX N = YYYY (For UHF, Alpha and Beta each get half this energy) 0 N>0 Default (0.0 Hartree) N*micro-Hartree IWType is used to figure out how the initial velocity is is computed (in gnvelp).

If XXYYYY < 0 : Initial velocity = 0.0 Hartee (i.e., currently same as N=0 above)

IOp(1/78)

Sparse in L121 -N 0 1 Sparse here with cutoff 10**(-N), full elsewhere Use full matrices or spase based on standard settings. Use sparse fixed form

IOp(1/79)

IRCMax convergence in L115 Stopping criteria in L118 and L121.

IOp(1/80)

L106: 0/1/2 Cartesian/Normal mode/Internal coordinate differentiation. 2 is NYI. L118: .eq.1 to surpress the 5th order correction after surface hop has been made in Trajectory Surface Hopping calculations. Needs also IOp(10/80=1) Nuclear Kinetic Energy Thermostat Option. (Currently only Velocity scaling is implemented) 19

0 11XXXXX 1000000

No Thermostat. Velocity scaling, but only for the first XXXXX simulation steps. (This options is useful, Velocity scaling, all the way through the simulation.

if thermostating in only required during equilibration.

IOp(1/81)

Nuclear KE thermostat in ADMP -- temperate is checked and scaled every IOp(81) steps.

IOp(1/82)

Temperature for nuclear KE thermostat in L121.

IOp(1/83)

Whether to read in frequencies for electric and magnetic perturbations. 0 1 2 Default (No). Yes. No.

IOp(1/84)

Differentiation of frequency-dependent properties. 0 N No. Mask for which properties on file 721 will be differentiated.

IOp(1/85)

Band gap calculation in PBC ADMP: 0 1 2 Default (No). Diagonalizae Fock matrix to get band gap, evolution, etc. No.

IOp(1/86)

Printing for NMR for ONIOM. 0 1 20 Default (1). Print tensors and eigenvalues.

as required by IOp(80). K. IOp(1/89) Maximum allowed deviation from average nuclear KE during ADMP.etc: 0: SCF 1: MP first order 2: MP2 3: MP3 4: MP4 5: CI one-particle 6: CI 7: QCI/CC 8: Correct to second order IOp(1/88) Whether to read in atomic masses (isotopes): 0 1 2 3 4 Default (1 if geometry read from input. 4 if geometry read from chk) Use most abundant isotopes. Integrate densities specified by following digits: Density to use from gridpoint 1 Density to use from gridpoint 2 etc. Read isotopes from rwf. IOp(1/90) To read in the velocity in cartesian coordinates Nuclear Kinetic Energy Thermostat Option.M. IOp(1/87) ONIOM integration of density. for backwards compatibility.2 Print eigenvectors as well.L. Read isotopes from chk. 21 . Integrate current densities. Average energy (in microhartree) to be maintained during Simulation. Read isotopes from input. in Kelvin. The temperature and pressure are read first. 0 1 2 K0 L00 M000 Do not integrate.

Default(15). full diagonalization. Quadratic micro-iterations. Regular micro-iterations. direct /w MM Hessian incore. 97. Regular non-coupled macro step. Coupled macro step. Coupled macro step. IOp(1/101. 103. full diagonalization. Quadratic micro-iterations. 95. IOp(1/94. direct /w full Hessian incore. direct /w prepared Hessian incore. Do not use dynamic convergence criteria for the micro-iterations. Quadratic micro-iterations. Coupled macro step. Quadratic micro-iterations. N3. N4 IOp(1/105) Reaction direction 00 22 Default (Same as 10) . N2. 98) IOp(94): IOp(95): IOp(96): IOp(97): IOp(98): <0 0 1 2 3 4 5 10 20 30 40 50 Davidson control for quadratic micro-iterations (see MMOpt2) RFO/Davidson control for quadratic micro-iterations (see MMOpt2) Davidson control for coupled QM/MM macro step (see MMOpt2) RFO/Davidson control for coupled QM/MM macro step (see MMOpt2) Control of quadratic micro-iterations and coupled QM/MM quadratic macro step. Also in microhartree.IOp(1/91) Thermostat Option. Coupled macro step. fully direct. direct /w raw MM Hessian incore. 102. 104) Phase control in L115 and L118: N1. 96. IOp(1/92) Maximum allowed deviation from average nuclear KE specified in IOp(81). fully direct.

10 20 0 1 2 00 10 20 000 100 200 300 0000 1000 2000 Forward direction Reverse direction Damped-Velocity Verlet (DVV) options for Dynamic Reaction Path Following Default (Same as 2) Use DVV Do not use DVV Default (Same as 10) Follow the rxn path in the forward direction Follow the rxn path in the reverse direction Default (Same as 200) Time step correction not used Time step correction used but not to recalculate current DVV step Time step correction used and current DVV step recalculated Default (Same as 1000) Use DVV stopping criteria Do NOT use DVV stopping criteria IOp(1/106) Damping constant for DVV Dynamic Rxn Path following (v0) 0 N Default v0=0.0001 IOp(1/109) Force-Velocity angle for DVV stopping criteria (Crit2) 0 N Default (90 Degrees) Use N Degrees 23 .0001 IOp(1/107) Error tolerance for DVV time step correction (Error) 0 N Default Error=0.04 (N=400) v0 is set to N*0.0001 IOp(1/108) Gradient magnitude for DVV stopping criteria (Crit1) 0 N Default (N=15) N*0.003 (N=30) Error=N*0.

0 N Default (standard temperature. IOp(1/114) Scale factor for harmonic frequencies for use in thermochemistry and harmonic vibration-rotation analysis. N/1000000. 0 N Default (1 atomosphere.IOp(1/110) Scaling of rigid fragment steps during microiterations. N/1000 atmospheres. Effectively disables the scaling Default (50) Scale up or down to maximum change in a variable of N/1000 IOp(1/112) Temperature for thermochemistry. IOp(1/113) Pressure for thermochemistry. unless read in). 0 1 2 -n Do not scale Scale with 1/NRA Scale with 1/Sqrt(NRA) Scale with 1/n (NRA = number of atoms in fragment) IOp(1/111) Step-size to use with steepest descent when L103 is having trouble: -N -1 0 N Scale up to RMS step of N/1000 if DXRMS is less. unless read in). 24 . 0 N Default (1 unless specified by IOp in overlay 7 or read in). N/1000 degrees.

using the z-matrix for connectivity info. Do not print the distance matrix. ANGLES WITHIN 0. 0 1 2 00 10 20 30 000 100 200 300 0000 1000 2000 Default: same as 2. DO NOT TEST FOR SUCH ANGLES. Print distance matrix. Default: same as 20. Default: same as 100. Print dihedral angles. 0 1 2 Default (print if 50 atoms or less) Print Don't print IOp(2/12) CROWDING ABORT CONTROL 25 . Do not print the angle matrix.Overlay 2 IOp(2/9) Printing of distance and angle matrices. Print dihedral angles. using a distance cutoff for connectivity info.001 DEGREE OF 109. IOp(2/11) PRINTING OF Z-MATRIX AND RESULTANT COORDINATES. Print the angle matrix.471 WILL BE SET TO ACOS(-1/3). Default: print only for small cases Do not print the cartesian coordinates in the input orientation Do print the cartesian coordinates in the input orientation IOp(2/10) TETRAHEDRAL ANGLE FIXING 0 1 2 Default (don't test). Use cutoffs instead of the z-matrix for determining which angles to print. using z-matrix connectivity if possible. Do not print dihedral angles.

then confirm symmetry Recover the previous symmetry operations from the rwf. Abort the run for zero atomic distances only Abort the run if any atoms are within 0. Default (same as 2).12) IOp(2/14) Internal coordinate linear independance. ATOMS WILL NOT TAKE THE ATOMIC NUMBERS. Do not perform the test. 0 1 2 3 10 linear independance check.0 1 2 3 Default (same as 1).12). but is useful for debugging the derivative transformation routines. SO THEY ARE NOT PUNCHED. with tight cutoffs. 100 Abort the job if the number of z-matrix variables is not exactly the number of degrees of freedom (i. Note that symm is still called. Leave symmetry in whatever state it is presently in. THE ATOMIC NUMBERS AND COORDINATES ARE PUNCHED IN FORMAT (I2.. Call Symm once with loose cutoffs. This is not correct for the job if they are dependant. If internal coordinates are in use. IOp(2/15) SYMMETRY CONTROL. symmetrize the resulting coordinates. Unconditionally turn symmetry off.3E20.5 A. IN 'ATOMS' FORMAT (3E20. IOp(2/13) PUNCH COORDINATES. this is not a full optimization). Don't even call symm. and confirm that the new structure has framework group. test the variables for linear independance and abort the Compute nuclear forces as well as second derivatives for the test. and will determine the However. the same symmetry. same as 0 for molecules but turns on assignment of space group ops. IN FORMAT SUITABLE FOR COORD INPUT TO Gaussian. Perform the test.e. 0 1 2 NO YES. NOTE. YES. but do not abort the job. Do not abort the run regardless of 0 distances. -1 0 1 Turns on symmetry. the molecule is not oriented. 2 3 4 5 Bring the molecule to a symmetry orientation. for PBC. 26 . but then disable further use of symmetry.

d-8). 0 1 2 3 Abort the job. but get symmetry info from the chk. 0 N>0 N<0 Default (determined in the symmetry package. using the new symmetry.d-7). use same tolerance for orientation. IOp(2/19) Largest allowed point group. Default (10) Do re-orientation for PBC. 0 N>0 N<0 Default (determined in the symmetry package. 10**-N. 10**N. IOp(2/20) Number (1-3 for X-Z) of axis to help specify which subgroup of the type specified in IOp(19) to use. Keep going and leave symmetry on. currently 1. Turn on symmetry operations for PBC. IOp(2/18) Tolerance for non-distance comparisons in symmetry determination. currently 1. Keep going. IOp(2/17) Tolerance for distance comparisons in symmetry determination. Suppress re-orientation for PBC. IOp(2/16) action taken if the point group changes during an optimization.6 00 10 20 100 Same as 5. 10**N. Keep going and leave symmetry on. as Hollerith string. 27 . 10**-N. using the old symmetry. use same tolerance for orientation.

format (50I2) IOp(2/40) Save (initial) structure and possible constraints in rwf 698: 0 1 2 3 Default (No). IOp(2/30) Read in vector of atom types (for debugging). 0 1 2 3 Default (1) No. Default (no constraint unless reading constraint from chk file). Yes. IOp(2/41) Force constants for Harmonic constraints.IOp(2/29) Update of coordinates from current Z-matrix. Read in structure from input stream. Yes. Yes. but remove z-matrix. 0 1 No Yes. Pick up structure from rwf698 on the chk file. N/1000 Hartree/Bohr**2 28 . -2 -1 0 N Read in force constants for each cartesian coordinates No constraints.

Ahlrichs TZV basis sets. UGBS basis set.Ne 6-311++g(3d2f) on Na . 17 18 19 20 21 22 23 24 25 26 Stuttgart/Dresden ECP basis sets. G3MP2large basis set. Dunning/Caltech basis sets. B-Ne 2MWB.6}Z) and augmented if IOp(7)=10. (VARIOUS IMPLEMENTATIONS MAY OMIT SECOND ROW ATOMS.p) on P.Ne 6-311+g(3d2f) on Na . He 6-311+G(2df) on Li . GENERAL--SEE ROUTINE GenBas FOR INPUT INSTRUCTIONS. USE IOp(8) TO SELECT 5D/6D. type selected by IOp(6) (=0-4 for V{D. whether intended for use in expanding AOs (IOp(5)) or in expanding the density (IOp(82)). 6 7 8 9 10 11 12 13 14 15 16 LANL ECP basis sets. IOp(3/6)=5 for MTsmall basis set.) SEE IOp(6) FOR DETERMINATION OF THE NUMBER OF GAUSSIANS IN THE INNER SHELL. G3large basis set. IOp(6) specifies type. Literature citations in SDDPot. Type selected by IOp(6).5-31G.6-31G (VALENCE FUNCTIONS ONLY) MINIMAL STO-NG EXTENDED LP-N1G (VALENCE BASIS FOR CORELESS HARTREE-FOCK PSEUDOPOTENTIALS) EXTENDED 6-311G (UMP2 FROZEN CORE OPTIMIZED) BASIS for first row.Overlay 3 IOp(3/5) TYPE OF BASIS SET. Stevens/Basch/Krauss/Jasien/Cundari ECP basis sets for H-Lu.Ar CBS basis #4 -. Literature citations in CEPPot. Type selected by IOp(6) for H-Ar. Ahlrichs SV basis sets. rest LANL1MB. CBS basis #1 -.p) on H . The same numbers are used for all basis sets.Ar CBS basis #2 -.Core correlation basis set Dunning cc basis sets.p) on H.6-311++G(2df. IOp(6) selects options.Q.Large APNO basis set CBS basis #6 -. S. MIDI! basis sets.9p)-->(631111. Cl CBS basis #5 -. MacLean-Chandler (12s.52111) for second row. EPR-III basis sets. 29 . 0 1 2 3 4 MINIMAL STO-2G TO STO-6G EXTENDED 4-31G.5. EPR-II basis sets.T.2p) on H . use daggers if any polarization CBS basis #3 -.6-31G.Si 6-31+G(df.Be 2SDF. Coreless: Li.6-31+g(d. 5 SPLIT VALENCE N-21G (OR NN-21G) BASIS FOR FIRST OR SECOND ROW ATOMS.6-31+G(d.

N-31G. Phys. Phys. When IOp(5)=8 (Dunning bases) this option selects the type: Dunning full double-zeta. CEP-31G. N-21G. Used for counterpoise calculations.27 28 DGauss basis sets. LANL2 ECP (where available. where one wants to modify the basis differently during different steps. DZ.B. . LP-31G FOR LI. 3546 (1981). READ THE GENERAL BASIS FROM THE RW-FILES AND MERGE WITH THE COORDINATES IN BLANK COMMON TO PRODUCE THE CURRENT BASIS. selected by IOp(6) Auto-generated.LP-N1G. When IOp(5)=9 (CEP basis) this option selects the type (H-Ar only): 0 1 30 CEP-4G.AL LP-41G FOR OTHER ROW1 AND TWO ATOMS. 1662 (1981) and J. for density basis (generated here from 2) Read from the alternate file and remove functions/ECPs for inactive atoms.STO-NG-VALENCE. MBS. IOp(3/6) NUMBER OF GAUSSIAN FUNCTIONS N STO-NG. Read the general basis from the checkpoint file. 85. LANL1 ECP. for density basis (generated here from 1) Same as 2. SHC ECP on second row). Chem. Dunning valence double-zeta. When IOp(5)=6 (LANL basis and potentials) this selects the type: 0 1 2 3 0 1 2 LANL1 ECP.MG. If non-standard ECPs are in use. otherwise LANL1). MBS. WAG basis (Dunning VDZ on first row. DZ. LANL2 ECP (where available. See Rappe. THIS OPTION IS USED TO CONTROL WHERE THE BASIS IS TAKEN FROM: 0 1 2 3 4 1x READ GENERAL BASIS FROM THE INPUT STREAM. NOTE IF IOp(5)=3 AND IOp(6)=8 . otherwise LANL1). 85. and Goddard. Chem. Same as 1. This option is useful when doing general basis geometry optimizations or properties using a wavefunction on the checkpoint file. DEFAULT OPTIONS IOp(6)=0 IF IOp(5)=0: N=3 STO-3G IF IOp(5)=1: N=4 4-31G IF IOp(5)=2: N=3 STO-3G (VALENCE) IF IOp(5)=3: N=3 IF IOp(5)=5: N=3 WHEN IOp(5)=7 (GENERAL BASES). Smedley.NA. they are read along with the basis set information. useful only for density basis sets.BE. J.

CBS-Q d(f). Functions from extended Huckel theory. P-FUNCTIONS ON HYDROGENS.d.d. When IOp(5)=17 (Stuttgart/Dresden ECP bases) this option selects the type according to: 6 7 0 1 2 3 SDD SDD for Z > 18. When IOp(5)=26 (Coreless basis) this selects the choice of basis (the same ECPs are used regardless): Default (3) Primitives which match the ECPs.2 CEP-121G.2d on 2nd and later atoms. 31 . 2 SETS OF P-FUNCTIONS ON HYDROGENS. 2 D-funcs. Three sets of p-functions and one set of d-functions. ONE SET OF D-FUNCTIONS AND ONE SET OF F-FUNCTIONS ON HEAVY ATOMS (indicates an extra tight 2df with ccp basis sets. Three sets of p-functions. on heavy atoms (scaled up/down by a factor of 2 from the standard single D value). TWO SETS OF D-FUNCTIONS AND ONE SET OF F-FUNCTIONS ON HEAVY ATOMS. 1d on 1st row atoms.p polarization basis Tight d for VnZ+1 (W1 theory) A SET OF DIFFUSE SP FUNCTIONS ON HEAVY ATOMS. Three sets of d functions and two sets of f-functions. When IOp(5)=7 (DGauss basis sets): 1 2 3 4 5 DGDZVP DZVP2 DGTZVP DGA1 (fitting basis) DGA2 (fitting basis) IOp(3/7) DIFFUSE AND POLARIZATION FUNCTIONS. Three sets of d functions and one set of f-functions. D-FUNCTIONS ON HEAVY ATOMS (2ND ROW ONLY FOR 3-21G). Augment non-hydrogens only (cc basis sets only). 0 1 2 3 4 5 6 7 8 9 10 20 100 200 300 400 500 600 700 1000 NONE.p) -. D95 and no ECP otherwise. 2d. VSTO-4G basis for 1st row. ONE SET OF P-FUNCTIONS AND ONE SET OF D-FUNCTIONS ON HYDROGENS. TWO SETS OF P-FUNCTIONS AND ONE SET OF D-FUNCTIONS ON HYDROGENS. A DIFFUSE S FUNCTION ON HYDROGENS. along with LP-31G potential. Three sets of d functions.

FORCE 10F. Li-Ne=0. D. Write over the dipole length integrals (518). 32 . IOp(3/10) Modification of internally stored bases (default 12000): 0 1 10 100 1000 2000 10000 20000 100000 200000 300000 DEFAULTS STO-NG STANDARD SCALE-FACTORS.IOp(3/8) SELECTION OF PURE/CARTESIAN FUNCTIONS. Add ghost atoms to /B/ so that every shell is on a separate center. 0 -1 N Usual place (572). He=1. Split S=P=D= AO shells into S=P. F. FORCE 6D. Do not split AO S=P=D Uncontract the AO basis.7. 0 SELECTION DETERMINED BY THE BASIS N-31G N-311G N-21G* STO-NG* LP-N1G* LP-N1G** GENERAL BASIS 1 2 10 20 FORCE 5D.95)/3 None.2. shells. 6D/7F 5D/7F 5D 5D 5D 5D 5D/7F IOp(3/9) Where 308 should store dipole velocity integrals. Uncontract the density basis Uncontract both basis sets. Massage the data in Common /B/ and Common /Mol/. Split S=P AO basis shells into separate S and P shells.65*I-4. Na-Ar=(0. FORCE 7F. Read in general basis data in addition to setting up a standard basis.325*(IA-1). the values for H-Ar can be determined by Slater's rules: H=1. For VSTO-nG. Store in RWF N. Do not split S=P AO shells.

15 1.CHEM.01 LP-N1G SCALE=1.15 FIRST ROW ATOMS ATOM B C N O F 1S 1.59 12.68 4.98 1.98 0.47 16.75 1.88 3.75 1.65 9.43 17.55 2.70 1.48 3.0 FOR LI-AR (INNER AND OUTER) STANDARD POLARIZATION EXPONENTS FOR N-31G* AND N-31G** BASES 33 .50 1.00 1.00 2SP 1.64 10.00 3SP 0.67 6.31 5.05 2.00 1.83 5.00 1.33 J.68 5.40 2SP 3SP 0.70 1.00 1.61 11. 2686 (1963) OUTER SHELL HAS BEEN SELECTED ON THE BASIS INNER SHELLS ARE BEST ATOM VALUES OF NUMEROUS OPTIMIZATION STUDIES ON VARIED SMALL MOLECULES.03 1.50 15.00 2SP 1.56 13.99 0.90 4.79 6.24 1.00 1.53 14.12 1.36 4.90 2.PHYS.04 0.98 0.95 2.00 1.01 1. N-31G (ALSO N-31G* AND N-31G**) STANDARD SCALE-FACTORS HYDROGEN H 1S 1.20 1S* 1. 38.00 1.98 1.10 2.00 SECOND ROW ATOMS ATOM P S CL 1S 1.00 1.02 1.72 1.99 1.00 3SP* 1.67 7.69 3.80 1.74 1.26 6.69 2.00 2SP* 1.ATOM H HE LI BE B C N O F NE NA MG AL SI P S CL A 1S 1.00 0.25 2.66 8.

starting with electric field. RAFFENETTI '2' INTEGRAL FORMAT. RAFFENETTI '3' INTEGRAL FORMAT. SUITABLE FOR USE WITH THE OPEN SHELL (UHF) SCF.multiply moments by fudge factor for charged species. ZINDO/S. INDO/2. ATOM NA. to account for sparkles. 0 1 2 3 9 REGULAR INTEGRAL FORMAT IS USED. blank terminated.4 0. but not yet accepted by them. RAFFENETTI '1' INTEGRAL FORMAT IS USED. SUITABLE FOR USE WITH OPEN SHELL RHF SCF AND THE POST-SCF USE ILSW TO DECIDE BETWEEN RAFFENETTI 1 AND 2.09 0.1 0. MG AL-CL VALUE 0. IOp(3/12) Flag for semi-empirical runs.ATOM H LI BE B C-NE VALUE 1. SCRF calculation -. 0 -1x -6 No.39 IOp(3/11) CONTROL OF TWO-ELECTRON INTEGRAL STORAGE FORMAT. IOp(3/14) Addition of electrostatic integrals to core hamiltonian. ZINDO/1. CAN ONLY BE USED WITH THE CLOSED SHELL SCF. 34 . CNDO/2.8 STANDARD POLARIZATION EXPONENTS FOR STO-NG* BASIS. up through 34 elements (hexadecapoles) in free format. PROCEDURES.6 0.2 0. Read coefficients of field. The magnitude is specified by IOp(3/15). IOp(3/13) Nuclear center whose Fermi contact terms are to be added to the core hamiltonian. translation vectors and d functions properly: 1 2 3 MNDO/AM1.

xxy. IOp(3/15) Magnitude of electric field.yyyy. zzxy field third derivatives in format (3D20. LANL2 potentials.10). ECPs if defined with the basis set. Read components of moments off rwf 521. yyyx. (These correspond to dipole. read 12 cards with x.yyxz.xy.zzzz. Dresden/Stuttgart potentials .yyyz. and xxxx.SDD for Z > 18.-5 -4 -3 -2 -1 Read components of electric field only from /Gen/ on checkpoint file. and hexadecapole perturbations).xyz field second derivatives. LANL1 potentials.yy.yyz. 1-34 Just component number n in the above order with magnitude given by IOp(3/15).xxz.xxxy.xxzz.xzz.zzzy.SDF 35 . no ECP otherwise. Read components of electric field only from /Gen/. N N * 0.xxyy. No. Yes. Stevens/Basch/Krauss CEP potentials. and the electric field is stored in Gen(2-4). xxxz.yz electric field gradient. yzz.xxyz. The nuclear repulsion energy is also modified appropriately.yyzz. read if general basis. unused READ IN FROM CARDS (SEE PINPUT FOR DETAILS) Dresden/Stuttgart potentials .zzzx. based on IOp(3/5). Read components of moments off rwf 521 on chk file. Default.xyy.xz. USE INTERNALLY STORED 'CORELESS HARTREE-FOCK' Goddard/Smedley SECE/SHC potentials.yyy.z components of electric field.0001. IOp(3/17) SPECIFICATION OF PSEUDOPOTENTIALS -1 0 1 2 3 4 5 6-7 8 9 10 11 Read potential in old format.SDD combination Dresden/Stuttgart potentials . IOp(3/16) Pseudopotential option 0 1 2 Default.y.zz. D95V. Yes. octopole.zzz. xxx. quadrupole. followed by xx.

55296 words on Cray).12 13 14 15 16 17 18 19 20 Dresden/Stuttgart potentials .MWB (third set) Pseudopotentials for all coreless basis. Alternative potentials for coreless basis. 0 N Default (3200 integer words).SHF Dresden/Stuttgart potentials .MHF (second set) Dresden/Stuttgart potentials . No longer used.CHF).MDF Dresden/Stuttgart potentials . N integer words. 0 N Default (Machine dependant.MHF (first set) Dresden/Stuttgart potentials . WITH A SUBSTITUTION POTENTIAL OF TYPE N WHEREVER SUCH A SUBSTITUTION POTENTIAL EXISTS. IOp(3/21) Size of buffers for integral derivative file.MWB (first set) Dresden/Stuttgart potentials . 16384 integer words on VAX. N integer words. IOp(3/20) Size of buffers for integral file. PRINT DON'T PRINT IOp(3/19) SPECIFICATION OF SUBSTITUTION POTENTIAL TYPE 0 N DONT USE ANY SUBSTITUTION POTENTIALS REPLACE THE STANDARD POTENTIAL OF THIS RUN (EG. 36 .MWB (second set) Dresden/Stuttgart potentials . IOp(3/18) PRINTING OF PSEUDOPOTENTIALS 0 1 2 PRINT ONLY WHEN INPUT IS FROM CARDS or if GFPrint was specified.

IOp(3/25) NUMBER OF LAST TWO-ELECTRON INTEGRAL LINK. We are re-using integrals produced earlier in the current calculation. Read in a list of basis function numbers in Format (10I5). USE OLD very inaccurate CUTOFFS IN LINK 311. IOp(3/26) ACCURACY OPTION. IOp(3/23) Disable use of certain basis functions. -2 -1 0 1 >0 Use integrals from a previous job read /IBF/ from the checkpoint file.0. Print in more readable format. DO ALL INTEGRALS AS WELL AS POSSIBLE in L311. 0 1 NO PRE-CUTOFF. Print as GenBas input. on the RWF. Use looser cutoffs in L314. 37 . LINK NUMBER. use the /IBF/ already WE ARE NOT USING TWO-ELECTRON INTEGRALS. Direct SCF. STO-3G. Print old-fashioned table. 0 1 2 10 20 DEFAULT. Sleazy. terminated by a blank line. INTEGRALS ARE COMPUTED TO 10**-10 ACCURACY. 0 1 10 100 1000 Default (don't print). TEST. Used only in L312.IOp(3/22) CONTROL OF THE PRE-CUTOFF IN THE TWO-ELECTRON D-INTEGRAL PROGRAM. PRE-CUTOFFS DESIGNED FOR THE 6-31G* BASIS. IOp(3/24) Printing of gaussian function table. DO ALL INTEGRALS AS WELL AS POSSIBLE in L314. and set their dialgonal core Hamiltonian elements to +100. Print shell coordinates. 0 1 Use all basis functions. TEST.

TWO-ELECTRON INTEGRAL SYMMETRY IS TURNED ON. 10**-N. 0 N DISCARD INTEGRALS WITH MAGNITUDE LESS THAN 10**-10. No. All integrals with d's -. SYMMETRY HAS BEEN TURNED OFF ELSEWHERE. AND SET2E WILL ALSO TURN IT OFF HERE. THE SET2E WILL INTERROGATE ILSW TO SEE IF THE SYMMETRY RW-FILES EXIST. IOp(3/29) Accuracy in L302: 0 N Default (10**-12). NOTE. IOp(3/32) Whether to check the eigenvalues of the overlap matrix: 0 1 2 38 Default (4). Yes. d and higher elsewhere. IOp(3/30) CONTROL OF TWO-ELECTRON INTEGRAL SYMMETRY. . IOp(3/28) Special SP code control. use IsAlg. IF THEY DON'T. HOWEVER. SP integrals in link 311.L311 does nothing. DISCARD INTEGRALS WITH MAGNITUDE LESS THAN 10**-N. 0 1 TWO-ELECTRON INTEGRAL SYMMETRY IS TURNED OFF.IOp(3/27) HANDLING OF SMALL TWO-ELECTRON INTEGRALS. IOp(3/31) USE OF SYMMETRY IN COMPUTING GRADIENT (Obsolete). 0 1 2 Default.

IOp(3/35) LAST INTEGRAL DERIVATIVE LINK (No longer used in overlay 3). Quadrupole. COMBINATION OF 1 AND 4. Dipole. and reduce expansion space if linear dependence is found (NYI). 0 N WHATEVER LINK STARTS WRITING THE INTEGRAL DERIVATIVE FILE SHOULD ALSO CLOSE IT. PRINT ONE-ELECTRON INTEGRALS.3 4 5 Yes. 0 1 2 3 NO DUMP. IS THE NUMBER OF THE LAST TWO-ELECTRON INTEGRAL DERIVATIVE PROGRAM. 0 1 3 4 5 6 NO INTEGRALS ARE PRINTED. IOp(3/36) Maximum order of multipoles to compute in L303: -1 0 1 2 3 4 00 10 20 30 None Default (dipole). using SVD to reduce expansion space. and use Schmidt orthogonalization to reduce expansion space. COMMON/B/ IS DUMPED AT THE BEGINNING OF EACH INTEGRAL LINK. ADDITIONALLY. Compute absolute overlap over contracted functions. Default (same as 20). ADDITIONALLY. COMBINATION OF 1 AND 3. Octopole. Hexadecapole. PRINT TWO-ELECTRON INTEGRALS IN STANDARD FORMAT. 39 . IOp(3/34) DUMP OPTION. THE INTEGRALS ARE PRINTED (STANDARD FORMAT). PRINT TWO-ELECTRON INTEGRALS IN DEBUG FORMAT. Yes. Compute absolute overlap over both contracted and over primitive functions. CONTROL WORDS PRINTED (AS USUAL). Do not compute absolute overlaps. Yes. IOp(3/33) INTEGRAL PACKAGE PRINTING.

PRISM. neglect three center two-electron integrals as well. 0 40 No NDDO . 0 1 2 Default (No). IOp(3/38) Algorithm for 1e integrals: 0 1 2 00 10 20 Default in 302. neglect 1e integrals with diatomic differential overlap. neglect four center integrals. IOp(3/39) Initialization of force and force constant rwfs. No. neglect three center one-electron integrals. Yes. same as 1.IOp(3/37) Whether to sort integrals in L320. neglect 2e integrals with diatomic differential overlap. IOp(3/40) Neglect of integrals: 0 1 2 3 10 20 30 keep all integrals. Default in 308. PRISM. 0 1 Initialize. Rys. IOp(3/41) Various semi-empirical methods. Leave alone. same as 1. Explicit spdf code. Do only overlap and not other 1e integrals.

J=1.I=1. for testing. Default parameters (same as 5). Read from Chk.12).18) Read parameters from rwf. Do CNDO/2.J=1.3). 41 . geometric mean for NDDO/1 Arithmetic mean in NDDO.3). Do ZINDO/1. Use Slater's rules scale factors. 1000000 Do Harris functional. 1400000 Regular SCF with separate K. 1300000 Second-order XC (NYI). followed by ((HDiag(J. 1500000 J as usual but NDDO for K. Default (unit overlap matrix). IOp(3/43) Solvent charge distribution to add to Hamiltonian: 0 1 2 3 4 None Read charges and DBFs from input stream in input orientation Read from RWF.12). GNDDO/1 parametrization. Do ZINDO/S. Use STO-3G scale factors. Load all the integrals into memory. Read parameters for atomic numbers 1-18 in the order Scale (D20. Read parameters from chk. Do INDO/2.I=1.I). Use the real overlap matrix. 1100000 Do Harris functional scaling atomic densities for current charge and multiplicity. Use the unit matrix for the overlap. followed by ((Beta(J.1 00 10 20 000 100 200 300 400 500 0000 1000 00000 10000 20000 100000 200000 300000 400000 NDDO Default use of NDDO beta parameters (arithmetic mean for indo parameters. 1200000 First-order XC. IOp(3/42) How to form NDDO core hamiltonian in L317: 0 1 2 Default (same as 1). Geometric mean in NDDO. or read-in parameters). Original INDO/2 Beta and HDiag Parameters. Read the integrals sequentially.I). Same as 1.18) (Format 3D20.

5 10 20 Read charges and DBFs from input stream in standard orientation Force units of Angstroms for coordinates. If negative.T and S unchanged. Force use of only the OS path for all calculations. IOp(3/45) Transformation matrix in L318. Use expanded matrix logic for PBC exact exchange.no (ab|xx) and no NDDO on 2e and V ints only -. Use unit matrix (for debugging). 0 1 2 use S**-1/2. just orthogonalize functions on the same center. 0 1 2 3 4 5 6 keep all integrals. no. Force units of Bohr for coordinates. neglect four center and 3 center (ab|ac) integrals. NDDO approximation -. IOp(3/44) integral rejection using L318. Order of multipoles in SCRF for L303.0) integrals. neglect four center transformed integrals. only 1e. yes IOp(3/47) Flags for use in PRISM and CalDFT throughout the program. neglect four center and three center (0. the perturbation is computed separately and stored in the third and fourth matrices in the core Hamiltonian rwf. -1 Bit flags: 0 1 42 If bit 0 is set (use AllowP array) then read in a list of allowed paths. Do not transform 2e integrals. IOp(3/46) Whether to abort the job if badbas detects an error: 0 1 2 Default (yes). .

Skip consistency checks for XC quadrature Do not do extra work to use cutoffs better. Do allocation for parallel XC but run sequentially. No longer used. Turn off use of Sqrt(P) in density-based cutoffs. Forbid use of gather/scatter digestion even for small numbers of density matrices. Turn off vFMM. The default is diagonal only on vector machines. Use Mura radial grid instead of Euler-2 grid. Reserved for more control of scatter/gather. Default is to use replicated Turn off Schwartz during FMM/NFx calculations. Reverse normal choice of Scat20 vs. replicated Fock matrices. Reverse normal choice of diagonal/canonical sampling in Prism and PrmRaf. Make all shells large in XC quadrature. Turn on angular offsets in XC grid generation. Do not symmetry reduce grid points on unique atoms. Do allocation for parallel 2e integrals but run sequentially. Trace input and output using Linda/subprocess.2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Reverse choice of whether to precompute distance matrix during numerical quadrature. currently only in PrismC. IOp(3/48) Options for FMM: RRLLNNTTWW RR: LL: NN: TT: WW: Range (default 2) LMax (default from tolerance) Number of levels (default 8) Tolerance (default 18) IWS (default 2). 43 . currently only affects CalDFT. Forbid use of FoFCou. Force single matrix code in CPKS. Force square loops. Force all near field in FMM. Turn on use of precomputed XC weights. even if not doing FMM. Turn off MP-based cutoffs in FF part of NFx. matrices only on Fujitsu and NEC. Force use of FoFCou. Make all atoms large in XC quadrature. Do nuclear contribution in FoFCou even for non-PBC Do not use special Coulomb algorithm in FoFCou.

No Cutoffs 10**-N. Default box length. Uncontract all shell pairs. Do not save as many multipole expansions as possible in memory. based on geometry (but minimum for molecules 3. IOp(3/52) Turn off normal evaluation of ECP integrals. IOp(3/53) Accuracy in ECP integral evaluation: 0 -1 N Default. Old routines will be used. Split primitives for better boxification. Set by PsmSet to indicate whether the NAtoms test for defaulting FMM was passed. 44 . UseUAB. Apply symmetry to derivative distributions (NYI). so L302 does not do ECP ints. Do not default to FMM. Turn on FMM print. Turn off parallelism in FMM (does not use parallel logic). IOp(3/51) Parameters for NF exchange and box length (MMMMNN): 00 NN 0000xx MMMMxx no NFx NFx range NN (R+n with n=NN-1). Force FMM on. Convert to sparse storage under FoFCou for testing. Box length MMMM/1000 Bohr. if 128 set.0 Bohr).IOp(3/49) More options for FMM: 1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 Indicates whether FMM can be used by FoFCou. Set up for parallel FMM but run loops sequentially. ECP integrals are evaulated in L302. 0 1 Default: if needed. Default UseUAB/Use 256.

IOp(3/57) No. 10**(-N). cutoff 5 x 10 ** (N+100) Yes. intermediate accuracy (5x10**-7) Yes. No Yes. of core electrons for Stuttgart/Dresden ECP's. IOp(3/58) Cholesky control options. crude accuracy (5x10**-5) Yes. default accuracy (10**-10). IOp(3/59) Threshold for throwing avay eigenvectors of S: 0 N Default (10**-4) 10**-N. cutoff 10**(-N) IOp(3/56) Cutoff for intermediate matrices during sparse operations: 0 N 100 times smaller than storage cutoff.IOp(3/54) Type of core density to use with ECPs: -1 0 1 2 None Default (None) Non-relativistic Relativistic IOp(3/55) Use of sparse storage: N<-100 -3 -2 -1 0 N Yes. 45 .

Yes. IOp(3/64) Set value for ILSW derivative flag. F(Mu. IOp(3/65) Number of k-points: -1 46 Just Gamma point. IOp(3/62) Maximum allowed error in S over orthogonalized basis functions: 0 N Default (10**-9).Nu) cutoff). (Defaults to no F(Mu. Lambda are basis functions on different atoms. set to N. Orthogonalize and remove primitives with 0 or small coefficients. Nu are basis functions on the same atom. -2 -1 0 N Set to zero Set to -1. 10**-(N). Leave alone.IOp(3/60) Control of orthogonalization and simplification of ccp basis sets. (Defaults to 15 angstroms). . Mu.Lambda) atom--atom cutoff criterion (angstroms).Nu) atom--atom cutoff criterion (angstroms) Mu. Orthogonalize and remove primitives with 0 coefficients. IOp(3/61) Sparse Semiempirical Hamiltonian Cutoffs in L302: XX XX00 F(Mu. Only active if IOp(3/39)=0. IOp(3/63) Debug option to test point charge FMM. 0 1 2 Default (1). 0 1 No.

SCI-PCM. IEF-PCM. Read setting from checkpoint and modify them by reading from the input stream. No. with standard values.N -N About N points. IOp(3/71) IDeriv level flag (for SCRF setup). otherwise off). C-PCM. Onsager. Yes. Flag for macroiterations (IPCM). Read from rwf. D-PCM. 0 1 2 3 4 5 0100 1000 2000 2100 2200 2300 3000 4000 10000 Default (1) Use defaults. Yes. Read setting from the input stream. IOp(3/67) Electric-field dependent functions: 0 1 2 3 Default (on if already present in basis read from rwf or chk. IOp(3/70) SCRF flag. IVC-PCM Cramer/Truhlar solvation model. Old logic for NRecip=N. Generate COSMOTHERMO output. Set to N. Read setting from checkpoint. 47 . IOp(3/66) Over-ride setting of NThInc in lineary dependence cutoff: -1 0 N 0 Don't change. with read-in values.

HCTH93. B1LYP. Default. IOp(3/74) Type of exchange and correlation potentials: -24 -23 -22 -21 -20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 00 00 01 02 03 04 05 06 48 O3LYP. VWN 80 (LSD) correlation VWN 80 (LSD) + Perdew 86 correlation . PBE1PBE mPW3PBE. B98.IOp(3/72) Solvent type flag (for SCRF setup). Becke "Half and Half": 0. Becke 3 with Perdew 86 correlation. Perdew 81 + Perdew 86 correlation. B1B95. HCTH407. BA3PBE. B97-1. HCTH147. Perdew 81 correlation. mPW1PBE. Becke "Half and Half" with LYP/VWN correlation. No correlation. B97-2. Lee-Yang-Parr correlation. same as 100.5 LSD Do only coulomb part. Becke3 with Perdew 91 correlation. Vosko-Wilk-Nusair method 5 correlation. PBE3PBE. BA1PBE. skip exchange-correlation.5 HF + 0. LG1LYP. mPW91PW91. mPW1LYP. IOp(3/73) ONIOM system flag (for SCRF setup). Becke3 using VWN/LYP for correlation.

IOp(3/75) Number of radial and angular points in numerical integration for DFT: 0 IIIJJJ Use a special grid designed for efficiency (default). -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 O3LYP coefficients. with a=0. III radial points. Becke "Half and Half" 0. 200 is Hartree-Fock-Slater.5 Xc + Corr Coefficients of 0 and 0 (no exchange). Hartree-Fock-Slater exchange (Alpha = 2/3). IOp(3/76) Mixing of HF and DFT. mPW91PW91 coefficients.0 and 1. PW91 exchange Gill 96 exchange PW86 exchange mPW exchange PBE exchange BA exchange VSXC exchange So 100 is Hartree-Fock. respectively.81.8 b=0. same as B1B95. Becke 3 coefficients: aLSD + (1-a)HF + b(dBx) + VWN + c(LYP-VWN). B98 coefficients.5 HF + 0.07 08 09 10 11 18 19 20 100 200 300 400 500 600 700 800 900 1000 1100 1200 OS1 correlation PW91 PBE VSXC Bc96 VWN5+P86 LYP+VWN5 for scaling KCIS Hartree-Fock exchange. B97-2 coefficients. and 402 is BLYP. Note that Becke actually used Perdew correlation rather than LYP.0 for DFT and HF. JJJ angular points. HCTH coefficints. B97-1 coefficients. LG exchange.72 c=0.7) Becke 1988 exchange. 49 . Coefficients of 0. X-alpha exchange (alpha= 0. 205 is Local Spin Density.

In L510. 0 for both. Use SS weights with XCal = M/1000. 1 to set up for CAS-MP2 or 2 to do spin-orbit calculation. 0 for both. 0 -1 -2 -3 -M<-3 N Default (SS weights) Use SS weights. Sign is applied to the local term. IOp(3/80) Range for microbatching in DFT. 1000000000 turns of microbatching logic. Use Becke weights with default cutoff of 30 au. 50 . Default (coefficients of 1 and zero as determined by IOp(3/42) IOp(3/79) Range cutoff in Becke weights. Sign is applied to the local term. IOp(3/77) Mixing of local and non-local exchange: -1 0 MMMMMNNNNN MMMMM/10000 non-local plus NNNNN/10000 local. Use Savin weights. Use Becke weights with cutoff N Bohr. DFT or mixed in accord with IOp(3/76) Mixture of MMMM/10000 DFT exchange and NNNNN/10000 HF exchange. Default (coefficients of 1 and zero as determined by IOp(3/42) IOp(3/78) Mixing of local and non-local correlation: -1 0 MMMMMNNNNN MMMMM/10000 non-local plus NNNNN/10000 local. Negative to turn off screening of basis functions and grid points.0 MMMMMNNNNN Default: pure HF.

-1 0 N None. Gross-Kohn form. Also static limit. IOp(3/85) Pure vs. Keep all generated functions. Do not split shells Split F and higher shells away from S=P=D. Cartesian. N Discard functions with L>=N. where MaxTyp is the highest AO angular IOp(3/84) Equivalent of IOp(3/7) for density basis. 51 . including 7=General basis. 0 1 2 Default (pure for read-in basis). but returning zero for imaginary response contributions. Pure. For auto-generated basis sets: 0 1 2 10 20 Default (22) Use all products of AOs.IOp(3/81) Frequency-dependence (if any) for XC functional. IOp(3/82) Fitting density basis set for Coulomb in DFT. Cartesian functions in density basis. Keep all functions with angular momentum up to MaxTyp+1. Use only AO primitives squared in fititing basis. Same numbering of basis sets as for AO basis. Default (-1). for debugging. None (static limit). For auto-generated basis sets: -1 0 momentum. See comments for IOp(3/5) and IOp(3/6) 28=Generate automatically from AO basis. IOp(3/83) Equivalent of IOp(3/6) for density basis. 0 1 2 3 Default (same as 1).

Solve iteratively with no preconditioning Solve iteratively with diagonal preconditioning. IOp(3/88) Modification of internally stored density basis.. Massage the data in Common /B/ and Common /Mol/. forming Z = modified A^-1. solving iterative with direct products. Use fits. IOp(3/87) Discard density basis functions based on angular momentum: 0 N No. Form inverse matrix once. with A formed to generate preconditioning.. Use fits. Iterative. This is also done if requested Split S=P density basis shells into separate S and P shells. Discard density basis functions with angular momentum >= N. Form A' over neutral distributions via multiplies by A. Split S=P=D=. Read in general basis data in addition to setting up a standard basis. Add ghost atoms to /B/ so that every shell is on a separate center. IOp(3/89) Set up for density fitting. no formation of A. Discard basis functions with angular momentum >= N. in IOp(3/10). solving iterative with stored A. Solve iteratively with symmetric block-diagonal preconditioning. F. . D. Do not split density S=P=D. .. Use fits. Do not split S=P density shells. 1 otherwise). shells. 0 1 2 3 4 5 6 7 1xx 2xx 3xx 4xx 52 Default (102 if a fitting set has been included and pure DFT is being used.IOp(3/86) Discard basis functions based on angular momentum: 0 N No. density shells into S=P. Form A' over neutral distributions via direct products. 0 1 10 100 1000 2000 10000 20000 None.... Do not use density fits.

IOp(3/93) Nuclear charge distribution: 0 1 2 3 4 10x Default (1. unless scalar relativistic) Point nuclei Single s Gaussians using formula of Quiney et. for debugging. DBFs have normalization as for AOs. 53 . AOs have J-normalization. for debugging. default MM=09. IOp(3/92) Whether read-in basis sets are in terms of normalized primitives? 0 1 2 3 10 20 30 Default (12). Solve non-iterative using precomputed A'^-1. IOp(3/90) Thresholds for density fitting MMNN 10**(-MM) on iterative solution. AOs have normalization as for AOs. IOp(3/91) Scalar relativistic core Hamiltonian: 0 1 2 3 4 Default (1) Non-relativistic. Same as 2 but exponents are 100x smaller. al Very tight single s Gaussians. AO coefficients are for raw primitives. default NN=06. Douglass-Kroll-Hess 2nd order. DBFs have J-normalization. 10**(-NN) on generalized inverse. Put all functions into a single block in forming the preconditioning matrix.5xx 6xx 1xxx Solve iteratively with non-symmetric block-diagonal preconditioning. Include nuclear charge distributions in DBF set. DBF coefficients are for raw primitives. RESC. Douglass-Kroll-Hess 0th order.

. -1 otherwise). based on number of cells having overlap with cell 0. N Bohr. At least N. -N 0 N At least N cells in each direction. IOp(3/97) Number of PBC cells for exact exchange: 0 N As many as look significant. IOp(3/94) Range of PBC cells in Bohr. IOp(3/98) Maximum number of density matrices in PBC. Multiply usual range by M. Based on range estimate (IOp(3/94)). IOp(3/99) Whether to set up precomputed quadrature grid in L302: 0 54 Default (2 if doing DFT. IOp(3/95) Minimum number of PBC cells. At least N.Mxxx Use method M to handle nuclear charges during density fitting. 0 N Default. 0 N -M default (100). No more than N matrices. IOp(3/96) Number of PBC cells for DFT: 0 N As many as look significant. At least N cells total.

Default is Max(NDBF/2. storing grid parameters. storing only grid parameter Yes. Reuse. and point coordinates. Default is Max((1000. IOp(3/100) Minimum Number of PBC cells for PBC-MP2 0 N Same as for HF exchange. storing grid parameters and weights. Yes. Default is 6 (using 5 previous solutions plus the current right-hand side to generate the initial guess).-1 1 2 3 No Yes. Negative to use projected equations rather than least-squares. weights. N. IOp(3/105) Re-use of PBC cell data. IOp(3/102) Number of density fittings solutions to save from previous SCF iterations. 0 1 Default (re-use if present).1000). 55 . IOp(3/104) Maximum number of iterations during iterative density fitting.NDBF+100). IOp(3/103) Maximum number of vectors allowed in expansion space during iterative density fitting. No more than N total. IOp(3/101) Maximum range of cells -N N No more than N in each direction No limit.

IOp(3/106) Override default number of atoms threshold for turning on FMM (for debugging). This number is scaled up appropriately if symmetry is in use. 56 . Read from chk file. 0 N Default (60) N atoms for the C1 case. to compensate for the loss of some symmetry with FMM.2 3 Do not reuse.

Renormalize and orthogonalize intermediate SCF results which are on the RWF. Default (1 for PBC without alter. IOp(4/6) FORCED PROJECTION WHEN GUESS IS READ FROM CARDS (401). Re-use orbitals not Fock matrices. This uses the Harris functional unless atoms heavier than Xe are present. chosen via IOp(11).Overlay 4 IOp(4/5) Type of guess: 0 1 2 3 4 5 6 7 8 it. 9 it.1 corresponding to 1. 57 . Projected ZDO guess. in which case Read guess from the checkpoint file. Read generalized density specified by IOp(38) from the chk file & generate natural orbitals from Huckel is used. Don't check MOs for othornormality. Renormalize and orthogonalize the coefficients which are currently on the read-write files. on the checkpoint file.3.5 * V.3. SUPPRESS PROJECTION. Read generalized density specified by IOp(38) from the rwf file & generate natural orbitals from Default. 100 1000 00000 10000 20000 Convert Guess=Check to Guess=Restart or to generating guess depending on what if anything is Use the simultaneous optimization recipe: S**-0. EVEN WHEN BASES ARE IDENTICAL. Re-use Fock matrices instead of orbitals. Huckel guess (only valid for NDDO-type methods). Default MO checking (yes). otherwise 2).2. IGuess values of 10-14 are generated internally and are the sparse versions of 0 and 5-8. Note that variable IGuess here has 4. 0 1 2 00 10 20 000 100 200 FORCE PROJECTED GUESS. Read intermediate SCF results which are on the checkpoint file. EVEN WHEN BASES ARE IDENTICAL. Guess from model Hamiltonian. Suppress orthogonalization.4 above.2. FORCE PROJECTED GUESS. Check MOs for othornormality. Default orthogonalization (perform) Schmidt orthogonalize guess orbitals.

IOp(4/7)

SCF CONSTRAINTS (401,402,403). -1 0 1 2 3 4 5 6 Ignore ILSW and determine on the fly. USE ILSW TO DETERMINE. REAL RHF. REAL UHF. COMPLEX RHF. COMPLEX UHF. COMPLEX, BUT USE ILSW TO DECIDE WHETHER RHF/UHF. REAL ROHF.

IOp(4/8)

ALTERATION OF CONFIGURATION (401). 0 1 2 10 100 DO NOT ALTER CONFIGURATION. READ IN PAIRS OF INTEGERS in free format INDICATING WHICH PAIRS OF MO'S ARE TO BE INTERCHANGED. PAIRS ARE READ UNTIL A BLANK CARD IS ENCOUNTERED. Read in a permutation of the orbitals. READ ALTERATION INFORMATION FROM THE READ-WRITE FILE. Use alpha orbitals for guess for both alpha and beta.

NOTE IF THE CONFIGURATION IS ALTERED ON AN OPEN SHELL SYSTEM, TWO SETS OF DATA AS DESCRIBED ABOVE WILL BE EXPECTED, FIRST FOR ALPHA, SECOND FOR BETA.

IOp(4/9)

SCF symmetry control (401). 0 1 2 3 4 5 10 20 100 200 1000 58 Default, same as 104 Read groups of irreducable representations to combine in the SCF. These are read before any orbitals and before alteration commands. Use no symmetry in the SCF. Pick up the symmetry mixing information from the Alteration read-write file. use the full abelian point group, as represented by the symmetry adapted basis functions produced by link 301. Initial guess orbital symmetries are assigned. (Use symmetry in SCF if possible, but do not assign initial guess abelian symmetries). Localize all occupied orbitals together and all virtual orbitals together Localize the orbitals within the selected or defaulted symmetry. Assign orbital symmetries for printing in full symmetry. Do not assign orbital symmetries in full symmetry. Force the guess orbitals to have the Abelian symmetry. This option can cause the symmetry

adapted basis function common blocks to be modified.

IOp(4/10)

Orbitals to mix to form complex guess (401). 0 1 Mix the HOMO with the LUMO. Read from cards (2I3) pairs of integers indicating which pairs of orbitals are to be mixed. Reading

is terminated by a blank card. NOTE THE SAME CONSIDERATIONS FOR OPEN SHELL SYSTEMS WHICH APPLIED IN IOp(8) APPLY HERE, ALSO.

IOp(4/11)

Type of Guess (401): For iterative ZDO Guess: -1 0 1 2 3 4 5 6 Force old path using old Huckel. Best available (6,4 in order of preference). Old Huckel. CNDO. INDO. New Huckel. Iterative extended Huckel. Harris, converted to IGuess=3 and IZDO=3 here.

For unprojected single diagonalization guess: 0 1 2 3 000 100 200 300 400 500 1000 n0000 MMMM00000 Use functional MMMM Default (same as 1). Use bare core matrix. Dress core Hamiltonian with QEq-based density. Use Harris Functional. Default, same as 2. Use SG1 and 10^-6 accuracy in Harris guess Use FineGrid and 10^-8 in Harris functional. Use UltraFine and 10^-8 in Harris functional. Use user's IRadAn and 10^-8 in Harris functional. Use (199,974) and 10^-12 in Harris functional. Save energy in Gen(43) for Harris functional. Force IDoV=n in HarFok.

59

IOp(4/13)

MIXING OF ORBITALS (401). 0 1 NO MIXING. LUMO = LUMO + HOMO (ALPHA) AND LUMO = LUMO - HOMO (BETA). NOTE THAT THIS WILL USUALLY DESTROY BOTH SPACIAL AND ALPHA/BETA SYMMETRY. THE MIXING IS DONE AFTER ANY ALTERATIONS.

IOp(4/14)

Reading of specific orbitals (401). 0 1 No. Yes. For alpha orbitals, read one card with the format for the orbitals, followed by zero or

more sets of IVec (I5) -- vector to replace. If IVec is -1, all NBasis vectors follow. (Vector(I),I=1,NBasis) -- vector in the specified format. Input is terminated by IVec=0. For beta orbitals, the same format as for alpha is used. Note that if alter is also specified, the replacements are read before the corresponding alterations (thus the order is alpha orbitals, alpha alterations, beta orbitals, beta alterations).

IOp(4/15)

Spin-state for initial guess (401). 0 N UHF). Use multiplicity in /Mol/. Use multiplicity N. This is useful for generating guesses for open-shell singlets or unusual

spin states involving orthogonal orbitals by treating them as high-spin in the guess (which only does

IOp(4/16)

Whether to translate basis functions of read in guess (401). 0 1 2 3 Default (same as 2). Use the basis functions as is. Translate to the current atomic coordinates. Translate to the current atomic coordinates, and determine an overall rotation to provide to the read- in orbitals.

IOp(4/17)

Number of open-shell orbitals (not electrons) in 402. 0 N 60 #open electrons. N. Number of electrons in the CAS space.

Use slater determinants. Pulay. INDO. Use pseudo-diagonalization.IOp(4/18) Number of orbitals in CI in 402. CNDO. 61 . Use determinant basis with Sz=b/2. IOp(4/20) Type of model (402): (This is also tested in 401 to see whether atomic number greater than 102 are special flags). Camp-King. No 3/4. No Pulay (DIIS).default full CAS. 3/4 point on unless Pulay or Camp-King. 3/4. MNDO. IOp(4/21) SCF type (402). L405: Trucation level for excitations -. L405: Number of orbitals in the CAS space. No pseudo-diagonalization. use pseudodiagonalization). AM1. no Camp-King. 0 1 2 3 4 5 Default (AM1). Flags for MCSCF (L405): 1 10 100 1000 Read options from input stream. MINDO/3. No Camp-King. 0 1 2 10 20 100 200 1000 2000 Default (no Pulay. Default is number of open shells. Just list configurations. IOp(4/19) L402: Spin change in CI (default based on multiplicity).

(For straight semiempirical calculations). NRow in L405. More flags for MCSCF: IFlag2 IOp(4/23) Number of iterations (402. Single determinant. but don't convert from Lowdin orbitals. (For Opt=MNDOFC). Write formatted file of symbolic matrix elements. Do 1st derivatives analytically if possible. /Mol/. 0 1 2 3 10 Default (don't update). . NDiag in L405. RHF/UHF from IOp(5). ROHF (NYI). IOp(4/25) Wavefunction (402). 2nd derivatives. 0 N Default. 0 1 2 3 4 5 -N 62 Default (Same as 1). Biradical 1/2 CI (only for MINDO3. Don't update. General CI. using specified orbitals. Restart 2nd derivatives.10 binary switches in L405.AM1). 403). and ILSW on the rwf (402). with N microstates read in. eigenvalues.MNDO. (For Opt=MNDOFC). Update. IOp(4/22) Derivatives? (402). Update. General CI.AM1). Yes. Update second force array instead of first. 0 1 2 12 100 1 No. IOp(4/24) Whether to update orbitals. Closed-shell 1/3 CI (only for MINDO3. N.MNDO.10000 100000 Write unformatted file (NDATA) of symbolic matrix elements.

IOp(4/29) NC in L405. 0 1 NO DUMP. Equal occupations of the lowest N virtuals and high N occupieds.15 IOp(4/28) SCF Convergence (10**-N. 63 . IOp(4/33) PRINTING OF GUESS. No Yes. TURN ON ALL POSSIBLE PRINTING. 0 1 2 NO PRINTING. default 10**-7). PRINT THE MO COEFFICIENTS. IOp(4/34) DUMP OPTION. IOp(4/31) Root to solve for in CI (402) (Default is 1). -2 -1 N Yes.05 au (according to ZDO) of the HOMO & virtuals within 0. mix all equally. mix valence orbitals and an equal number of virtuals. mix valence occupieds with 0. Yes.IOp(4/26) Whether to mix orbitals in generated guess density: 0 -3 au. PRINT EVERYTHING.

Yes. Copy on disk is used. This type of computation can only be done in a CAS comp. IOp(4/43) IDiEij = Switch for direct matel calculation. This will be the only way to print configs in a direct matel calc. since there can be many thousands in a large CAS. Old S parameters. 0 1 No. Configs printed as normal. IOp(4/38) Generalized density to use for natural orbitals: N Density number N. with all matels calculated here and stored on disk. Eij's calculated here and stored on disk. reformat ZIndo integrals and wfn into rwf. Overlap assumed to be unity. 0 1 For normal route. Old Si parameters. A flag is automatically sent to L510 to tell it to compute the remaining atels directly.IOp(4/35) Overlap matrix. 2 As option 1. Also L510 must use Lanczos. The configurations will not be listed unless see below. IOp(4/44) 1 Prepare input for Mp2 implies IOp(21)=10 Slater Det. 0 1 2 Default (copy on disk is used). For direct route. option generates data for use in MC-SCF generation of zeroth order H note: for b=0 ie no unpaired spins forces use of Clifford Algebra Spinors 64 . but all configurations are printed. IOp(4/37) Selection of old MNDO parameters in L402: 0 1 2 Defaults. IOp(4/36) ZIndo reformating.

Use Lewis dot structure guess density. This is a normal GVBCAS calculation.instead of simple determinants IOp(4/45) Ipairs = number of GVB pairs in GVBCAS. L405 performs a CAS on the inner space. implicitly. 0 n Default. Yes. but is smaller subsequently. IOp(4/46) CI basis in CASSCF: 1 2 3 10 Hartree-Waller functions for singlets Hartree-Waller functions for triplets Slater Determinants Write SME on disk IOp(4/47) Convert to sparse storage after generating guess. IOp(4/48) Whether to do (sparse) Conjugate Gradient methods in 402: 0 1 2 No. Use threshold 10**-N. 65 . There are n pairs: 2*n extra orbitals and electrons will be added into the active space later. -1 0 N No. L510 will execute normally. Default (-1 if sparse is turned on) Yes. normal CAS calculation. IOp(60-62) Over-ride standard values of IRadAn. Use diagonal guess density. No pairs. Yes. and sets up L510 to compute extra matels etc. IRanWt. -n There are n pairs: 2*n orbitals and electrons of the specified CAS are to be considered to be GVB type orbitals when generating configs / matels. use the Lewis dot structure to generate a sparse guess directly. This is the GVBCAS test mode. This occupies as much space as a full CAS in this link. and IRanGd.

Turn on non-bonded terms. Turn on angle bending.0. Turn on all terms.IOp(4/63) Flags for which terms to include in MM energy: 0 1 2 10 100 1000 10000 100000 Default (111111) Turn on all terms. initial threshold 5 x 10 ** (N+100) IOp(4/66) Dielectric constant to be used in MM calculations. initial threshold 5x10-5. No variable thresholds. 0 N Eps = 1. initial threshold 10**(-N) Yes. Yes. 0 1 N N N<-100 Default: Yes. 10**-N. IOp(4/64) Cutoff for MM non-bonded term. 0 1 66 Default (211 if UFF. Eps = N / 1000. 10**-N. IOp(4/67) Whether to use QEq to assign MM charges. Turn on inversions/improper torsions Turn on torsions. r**-2 Coulomb. 1==> 221) Do QEq. . 0 N Default (no cutoff). Turn on bond stretches. 2 otherwise. r**-1 Coulomb. IOp(4/65) Tighten the zero thresholds as the SCF calculation proceeds.

IOp(4/68) Convergence criterion for microiterations in L402: 0 N Default. Do not do the extra guess. Do the extra guess and store as the initial Fock matrix. Just generate orbitals from the Harris guess. IOp(4/72) Irreps to keep in MCSCF CI-wavefunction. Save the Harris guess as an initial Fock matrix. All IOp(4/80) The maximum conjugate gradient step size (MMNN) 67 . IOp(4/69) Whether to do a new additional guess in addition to reading orbitals from the rwf: 0 1 2 3 00 10 20 Default: yes if no Guess=Alter. Default (200) Do for atoms which have charge specified or defaulted to 0. 20 otherwise). Do for all atoms regardless of typing. Do the extra guess regardless. 0 IJKLMNOP List of up to 8 irrep numbers to include.2 00 10 20 000 100 200 Don't do QEq. Default (20) Do for atoms which were not explicitly typed. 10**(-N). Default (10 for PBC. Harris guess. IOp(4/71) Write out AM1 integrals in 402 0/1 No/Yes. Do for all atoms regardless of initial charge. and not a small geometry step.

IOp(4/110) Scaling of rigid fragment steps during microiterations. (defaults to no F(Mu. No preconditioning.Lambda) atom--atom cutoff criterion (angstroms) Mu. Lambda are basis functions on different atoms. (MM=00 defaults to 20 cycles. F(Mu.0000 MMNN No maximum step size Step size of MM. Don't use CG DIIS Use CG DIIS. MM=01 turns DIIS off) F(Mu. Maximum Number of purification cycles per CG iteration. Nu are basis functions on the same atom. IOp(4/82) Conjugate-Gradient Parameters MM NN00 00000 10000 000000 100000 0000000 1000000 Maximum Number of CG cycles per SCF iteration. (defaults to 4 CG cycles). (defaults to 15 angstroms). Polak-Ribiere CG minimization Fletcher-Reeves CG minimization Use diagonal preconditioning in Conjugate-Gradient.NN IOp(4/81) Sparse SCF Parameters MM NN00 PP0000 Maximum number of SCF DIIS cycles.Nu) cutoff). (defaults to 3 cycles). 68 .Nu) atom--atom cutoff criterion (angstroms) Mu.

Force Rys only. Compute SCF Fock matrices. Cutoffs for 10**-10 accuracy. forbidding incore Force conventional Something obsolete use option NNNNN in control of 2e integral calculation. HGP df (no sp done here at all).Overlay 5 IOp(5/5) Direct SCF control (L502. Form full Fock matrix every time. HGP d. Read the integrals off disk OR Incore (Acording to Memory) Compute 2e integrals(Using FoFDir). The default for first derivatives. Compute 2e integrals and store in-core. HGP sp. 0 1 2 3 4 NNNNNx 0000000 1000000 2000000 Default (same as 1).: ICntrl = Algorithm control: 0 1 2 3 4 5 6 7 8 10 20 30 40 100 200 Default for MCSCF is (1522). Calcs).(How to Obtain the Integrals) 0 1 2 3 4 5 6 NNNNNx Incore or Direct(FoFDir) according to available Memory. Rys f (no sp done here at all). NNNNN=ICntrl with values as below. No cutoffs. Read the integrals off disk. Compute 2e integrals (Use TrnDir + FoFDir = 4 Can. L508). Force FoFDir. Cutoffs for high accuracy. Rys df (for debugging). Force HGP only. Use option NNNNN in control of 2e integral calculation. Compute 2e integrals. HGP spd. BraKet only. Default -. Rys f. 69 . rest not done here. Compute 2e integrals and forbid in-core.incremental Fock matrix formation only for direct SCF. BraKet up to L=8.only in L502. Do not compute operator matrices. L510: Direct MCSCF control (L510). The default for integrals or second derivatives. Form delta-F each iteration -. Sleazy (10**-6) Cutoffs.

300 400 500 600 700

Compute CIS operators Compute WTilda terms. GVB: DA==>FJ,FK. Compute regular integrals and load into R03, in canonical form if IOpcl=0 and square form if Compute raffenetti integrals: IOpCl=0 IOpCl=1 IOpCl=2 IOpCl=3 IOpCl=4 1000 2000 3000 4000 Load R1. Load R1 and R2. Load R1 and R03. Load R1, R2, and R03. Load R2 and R03. Do not compute forces. Compute forces. Make derivative Fock matrices Make derivative Fock matrices and form contributions to polarizability derivatives (ie 6 sets of forces will be returned in FXYZ, and 3 extra sets of densities must be supplied in PA,B). 5000 10000 Compute forces using including CIS 2PDM terms. Compute second derivatives.

IOpcl=1. NMatS is used as the dimension of R0 if IOpcl=1.

0000000 1000000 2000000

Default -- incremental Fock matrix formation only for direct SCF. Form full Fock matrix every time. Form delta-F each iteration -- only in L502.

IOp(5/6)

Convergence (RMS density except in L506 (SQCDF), L508(rms rotation gradient), and L510 (Energy)). 0 N 10**-8, except 10**-7 for PBC. 10**-N.

L510: CONVERGENCE CRITERIUM (ACC) FOR THE ENERGY IN THE MCSCF 0 N Acc = 10**(-8) Acc = 10**(-N)

IOp(5/7)

Maximum number of iterations. 0 128, except 512 in L503 and L508.

L510: MAXIMUM NUMBER OF ITERATIONS TO BE DONE (MaxIt) 0 N -1 MaxIt=64 (Default Value) MaxIt=N It does only a CI calculation. Options other than the standard SCF ones:

70

IOp(5/8)

SELECTION OF THE PROCEDURE OF DIRECT MINIMIZATION (L503). 0 1 2 4 5 STEEPEST DESCENT WITH SEARCH PARAMETERS DEFAULT STEEPEST DESCENT WITH SEARCH PARAMETERS READ (SEE BELOW) CLASSICAL SCF (ROOTHAAN'S METHOD OF REPEATED DIAGONALIZATION CONJUGATE GRADIENTS WITH SEARCH PARAMETERS DEFAULT CONJUGATE GRADIENTS WITH SEARCH PARAMETERS READ: MAX. NUMBER OF SEARCH POINTS (I1) MIN. NUMBER OF SEARCH POINTS (I1) INITIAL STEPSIZE, TAU (G18.5) SCALING FACTOR FOR SUBSEQ. TAU (G20.5) Q (G20.5) Search method (L508). 0 1 2 3 00 10 20 000 100 200 Default (123). Steepest descent. Scaled steepest descent. Quadratic convergence (after rotation gradient is sufficiently small). Default linear search (full search). Do a full linear search to locate a minimum. Do a linear search only if the energy goes up after the initial step. Default handling of wrong curvature (switch direction). Reverse direction if curvature in NR step direction is wrong. Take pure NR steps, even if curvature is wrong.

Flags for L510: 1 10 100 1000 10000 1000 IRdF2, read damping coefficients. IFrzCI, freeze CI coefficients after 1st iteration. Read unformatted symbolic matrix elements from NDATA instead of rwf. Read in damping factors from cards. Use Levy damping. Read Fock matrix restriction matrix.

IOp(5/9)

SWITCH TO CLASSICAL SCF AFTER DENSITY MATRIX HAS ACHIEVED A CERTAIN CONVERGENCY (L503 only). 0 1 2 NO YES, CRITERION DEFAULT 10(**-3) YES, CRITERION READ IN (FORMAT G16.10)

Number of pair iterations (L504, L506). -1 None; coefficients are frozen at initial values (L504: causes coefficients to be read in in order 71

11 12 22). 0 L510: 5.

IOp(5/10)

IVShft Level shifting: -N -2 -1 0 N Dynamic level shifting to achieve a gap of -0.001*N Dynamic level shift to a default goal (same as -200) No level shifting. Default: -200 for diagonalization calculations, -1 for sparse diagonalization replacements, and if energy DIIS is turned on. Shift by 0.001*N

IOp(5/11)

3 and 4 Point Density extrapolation control (L501,L502, L503 has only 4 point, L505). 0 1 BOTH 3-POINT & 4-POINT EXTRAPOLATION PERFORMED WHEN APPLICABLE. THREE-POINT EXTRAPOLATION IS INHIBITED, BUT THE PROGRAM WILL STILL PERFORM FOUR-POINT EXTRAPOLATION WHEN POSSIBLE. (IE. DISABLED).

IOp(5/12)

Whether to allocate only two N**2 arrays for RHF: 0 1 2 Default (No). Yes. No.

Number of GVB pairs (L506). If non zero, the number of orbitals in each pair is read in format (30I2). Each pair consists of the highest available occupied from the guess (after high spin orbs are accounted for) and the lowest available virtuals. If <0, pair coefficients are read; otherwise standard initial values are used.

IOp(5/13)

Action on convergence failure (L502): 0 1 2 72 Default (2). Continue the run even on non-convergence. The ILSW flag for convergence failure is set. Terminate on non-convergence.

(Calculates directly Nact*(Nact+1)/2 Fock matrices by contracting the AO integrals with the Density matrices. 0 1 2 CALCULATION IS PERFORMED (PROVIDED OF COURSE THAT ENOUGH SPACE EXISTS IN THE RW-FILES). CALCULATION IS PERFORMED. Fit the converged density even if fitting is not in use during the SCF. ADMP. use IOp(17)). Number of rows in an initial transformation of MO. State average density matrices. Skip core-valence Fock matrix elements. CONTINGENT ON SPACE. (More input from cards See below) IOp(5/14) Special functions in L502: 0 1 10 20 100 matrices. 0 1 2 ON BESSEL CRITERION ON STRONGER INDIVIDUAL-OVERLAP CRITERION OFF None. IRdNLp. L503). Also redoes the fit at the end even if using fits during SCF.L510: MCSCF flags: 2 10 100 XX000 Generate MOs using UHF natural orbitals. CONTROL OF ANNIHILATION OF SPIN CONTAMINANTS (L502). Turn the current RHF run into a uhf run at the end of this link. first cycle: use initial AO densities. 73 . Skip valence-virtual Fock matrix elements. Use full diagonalization method rather than Lanczos. Just recompute band structure from stored real-space Fock matrix. REORDERING OF THE ORBITALS (MAINTAINING CONTINUITY OF THE WAVEFUNCTION ALONG THE SEARCH PATH. CALCULATION IS BYPASSED. Use Generalized energy-weighted density routines regardless. Terminate after computing the 2e terms at the first iteration.turned on automatically in FoFDirINFC Number of Frozen Core Orbitals IRdNT. Skip core-valence Fock matrix elements. Do not use GEW routines even for CP. 200 1000 2000 10000 ADMP. later cycles: transform the density from L121 before calculating the energy and Fock L510: Flags for MCSCF: 1 10 100 1000 10000 100000 Skip valence-valence Fock matrix elements.-For testing purposes. (Obsolete. AND THE SYSTEM RW-FILES FOR THE APPROPRIATE DENSITY MATRICES ARE UPDATED (USEFUL IF ONE WANTS A POPULATION ANALYSIS).

CEM. By orbital least change. Lanczos starting vector in L510: -1 74 Read in eigenvector. Default (use Abelian symmetry in diagonalization). SNRDMS. keep overall wavefunction the same as the initial guess. KyDiag. but doing the minimal amount of orbital switching to accomplish this. 0 1 2 By eigenvalue. Sign Matrix Method. 0 1 2 3 00 10 20 Default (1 for L502. keep occupation of each irrep the same as the initial guess. 0 1 NO YES Selection of OCBSE vectors (L506). No. Pseudo-diagonalization whenever possible. 2 for L501 and L506). PDM. By energy least change. Use Abelian symmetry in diagonalization. Force formation of the Fock matrix using sparse storage. Yes. Do not use Abelian symmetry in diagonalization. CONTROLS THE AUTOADJUSTMENT OF TAU (L503). INHIBIT PERFORMANCE OF MINIMIZATION OF ALTERNATE WAVEFUNCTION PROVIDED BY SECOND ORDER PROCEDURES (L503). DiagD. 4 for sparse). CGDMS.(ILzVec=-1) See below .IOp(5/15) Apply Abelian symmetry constraints on orbitals. 0 1 DONE TAU IS KEPT FIXED IOp(5/16) Diagonalization method (L502): 0 -N 1 2 3 4 5 6 7 8 1xx 2xx Default (1 for full matrices. Yes. Pseudo-diagonalization with real diagonalization every Nth cycle. Force formation of the Fock matrix using full storage.

symmetric. Lowdin orthogonalize C+O and V. 2nd order iteration using RFO type step + level shift. SET TO ZERO IF Abs(F(I. Same as 30 in 502 but 20 in 508. L510: MCSCF flags. Use full 2nd order convergence. Force the density matrix to have full symmetry at the first iteration. be symmetric. Don't orthogonalize.D-10. Virtuals obtained by Schmidt orthogonalization to occupieds.0 IOp(5/17) CONDITION OFF-DIAGONAL TERMS OF THE Fock MATRIX (L503). symmetrize Fock matrices and require proper density matrix symmetry.D-5 FUZZY AND DEGEN READ IN (2D20. 2nd order iteration at end. 1012 for 508). replicate integrals so that density matrices and wavefunctions need not If 2E symmetry is on.14) Selection of virtual orbitals (L506). If 2E symmetry is on. 2nd order iteration using RFO type step + level shift and prepare for non-direct CPMCSCF 75 . Prepare for CPMCSCF(FREQ): Direct method with no storage of Hessian.O. then Schmidt. 0 1 2 5 10 100 200 300 400 500 Orthogonalize C. then Schmidt.O. VECTORS 0 1 FUZZY=1. Bad for 1st order method.V by separate Lowdin.J)). 0 1 Virtuals obtained by diagonalization of hamiltonians. DEGEN=2. Don't use natural orbitals each iteration. symmetrize them to ensure that they are exactly Do not symmetrize the density matrices. Force the density matrix to have full symmetry at every iteration. DEGEN) M.0 C(N) = 1. Always use tight convergence. DELETE COUPLING TERMS BETWEEN ALMOST DEGENERATE (DELTA E . Tighten convergence by an extra factor of 10.FUZZY. Default (1000) If the density matrices pass the symmetry test. Warning!! should be used only for large jobs where Hessian does not fit in memory.LE. Just schmidt. in preparation for CPMCSCF.LE.0 N C(1) = 1. Use of symmetry (in L502 and L508) and linear equation convergence (in L508): 0 1 2 3 10 20 30 40 100 200 0000 1000 2000 Default (1032 for 502. Choose LinEq convergence based on orbital gradient. If 2E symmetry is on. choose between replicating integrals and symmetrizing the Fock matrix based on whether the current density matrix is symmetric.

Do not save integrals (same as 0x). Force Square in-memory storage. Forbid in-core: force re-reading of integrals even if they fit in 2 buffers if conventional. Force allocation for 1 or 2 buffer case conventional case (VV. L503: CUTOFF CRITERIA IN SYMMETRY DETERMINATION OF M.D-4.SPAN ARE CONSIDERED TO BE ZERO 0 1 STHRS=1. 40000000 Do SA and prepare for SA Second Derivative Computation (terms involving 2nd order orbital rotation derivatives not included) IOp(5/18) L502: Mixing when doing damping: -3 -2 -1 0 N MO damping at all iterations. 10000000 Use State Average density matrices. Dynamic selection of density damping based on band gap and DIIS error. Force Lower-triangular in-memory storage.LE. do not convert to in-core if direct and enough memory for in-core is available.(the weights 8F10. 10**-N.O. scaled steepest descent is used): 0 N Default (1.J)). Turn off damping.J)).d-2).(IRdCIV) Use full diagonalization method rather than Lanczos. L502.ne. SPAN=5. L508): -1 0 1 2 3 4 1x 2x 3x 76 Choose the best given amount of memory available.S.O. Default (-1 unless reoptimizing during Stable=Opt). otherwise 1.IBuf2E). (100-N)/100 old density.STHRS ELEMENTS Abs(F(I. N/100 new density. Save generated integrals on disk (file 610). 30000000 Do SA and prepare for Gradient of Energy difference.8) 20000000 Do SA and prepare for SA-CPMCSCF. SYMMETRY IS DETERMINED IF LARGEST OFF-DIAGONAL M.10000 100000 1000000 Attempt to control root flipping in CI. L506.D-7 STHRS AND SPAN READ IN (2D20. Force computation of raff 1 and 2 integrals even for RHF. Read CI vector and use it every iteration. .GE. FOCK-MATRIX ELEMENT Abs(F(I. IOp(5/19) Over-ride integral storage control (L501.14) Damping (L506) Maximum rotation gradient for Newton-Raphson in L508 (above this value. 2 if possible.

no more than 1/3 virtual component for any occupied at each iteration. MCSCF flags: 2 10 Generate MOs using UHF natural orbitals. 77 . IOp(5/22) Use of DIIS extrapolation (L501. do a final unextrapolated diagonalization after convergence is reached. 0 1 ABORT RUN VIA LNK1E. maximum virtual mixing for MO damping: For density damping: 0 N Default (Damp if error > 0. 0 1 2 Default (no). L504). IRdNLp.001) Damp if error > 10**-N For MO damping: 0 N Default. L502. No. Extrapolation control in L506. L502. Yes.T. CONTINUE RUN. 0 Default (1042) for calculations using diagonalization (2) for calculations using sparse diagonalization replacements. IOp(5/21) DIIS error for density damping. Orbital rotation control (L506).PRINT F(1). 0 1 NO YES IOp(5/20) Final non-DIIS iteration (L501.END 2I2) (L503). L504). Maximum N/1000 virtual component. ACTION IF OTEST DETECTS PROBLEMS (L503). just quit when extrapolated convergence is reached. N Rotations are turned on when SQCDF is below 10**(-N). (READ ONE CARD WITH START.

IOp(5/25) Rotation application (L506). with "pFON" version of Fermi broadening. Freeze all closed.1 2 3 4 5 10 20 30 40 1xx Nxxx Mxxxx No. commutator. high spin and first natural orbitals. used (DIIS error/10^-N) for Use print level M in DIIS. L509). Read from rwf. Regular DIIS Energy-based mixing Energy DIIS when DIIS error has increased significantly or is above threshold Energy DIIS when DIIS error has increased significantly. Orbital mixing control in L506.2) . IOp(5/26) Type of calculation (L504). otherwise. Optimize only 2nd and higher naturals. Read from chk. 0 1 Optimize all orbitals. Yes. IOp(5/23) Flag for later points of an optimization. 0 1 2 Read from input stream. weight of energy DIIS in method 4. Yes. Apply rotations sequentially. mixture of energy and Use energy DIIS when commutator gives huge coefficients. Yes. Switch from energy to commutator when error is 10^(-N) in method 3. 0 1 Default (exponentiate rotation angles). 3 78 3rd root of CAS(2. with "FON" version of Fermi broadening. with Fermi broadening as well. IOp(5/24) Orbital freezing (L506). Yes. deciding on the fly between the two forms. so that pair and hamiltonian information can be reused (L506.

IOp(5/27) Whether to do closed-shell calculation in L502. Integrals with degree of contraction greater than or equal to N are done are regular integrals. IOp(5/28) L510: Root of CI to use in MCSCF (IState) 0 -1 N Defaults to Istate=1 Read IState from cards (see below) IState = N IOp(5/29) Use of rafinetti integrals during direct SCF. -1 0 1 N All integrals done as Raffenetti. No Yes (used for RHF direct SCF). Force closed shell. All integrals are done as regular integrals. IOp(5/30) Whether to symmetrize final orbitals using abelian symmetry operations (L502. 0 1 Default (Yes). if mulitplicity 1).2).2) GVB(1/2) Orthogonal open-shell singlet. Closed/Open control for L511: 0 1 2 Default. the unpaired orbitals are assumed to be high spin. ROHF Triplet (a debugging option). L505. Number of hamiltonians to read in (L506). 0 1 2 Default (Yes. 79 . If -1. Default: let FoFDir decide. Yes.2 1 0 -1 -2 Excited singlet as 2nd root of CAS(2. Force UHF. error if Multip>1. not needed in L506). If zero. closed for Multip=1. an open-shell singlet is assumed. GVB as CAS(2. It will never use Raffenetti for SCF.

Foc formation. BUT AT THE END OF EACH ITERATION. No longer used. Use general DBF logic only if the DBF rwf is present. L510): 0 1 2 3 4 5 6 00 N00 I000 00000 10000 20000 Default (No). Yes. REGULAR SYSTEM DEFAULTS APPLY HERE. SAME AS IOp(33)=2. The default is CoarseGrid for iterations and SG1 for final energy. N. IOp(5/32) Sleazy SCF (L502. Use approximation I. 0=normal 1=Linear approximation to Xc. O. BUT ADDITIONALLY THE DENSITY MATRIX IS PRINTED. then one iteration with next grid up. THE EIGENVALUES AND THE M. SAME AS IOp(33)=3. BUT ALL MATRIX TRANSACTIONS ARE PRINTED (BEWARE!!! MUCH OUTPUT EVEN ON SMALL MOLECULES. Decide between 1 and 4 based on details of the calculation. ONLY SUMMARY RESULTS ARE PRINTED (WITH POSSIBLE CONTROL FROM THE 'NO. IOp(5/33) PRINT 0 1 2 3 4 IOp(33) PRINT OPTION. Force use of general DBF logic.PRINT' OPTION). Force use of 1c instead of general DBF logic. IOp(5/31) How many vectors to form at a time during microiterations in L508 (NYI) and L509: 0 N Default (3 in L509). use loose integral cutoffs.) IOp(5/34) DUMP OPTION. SAME AS IOp(33)=1. Thresholds similar to DGauss for convergence and integrals. 80 .2 No. convergence on either energy or density and always do incremental No. No longer used. Yes. COEFFICIENTS ARE PRINTED AT THE END OF THE SCF. doing an inexpensive pass 0 and then full accuracy in pass 1. Do iterations with sleazy XC grid.

do integrals 3 digits more accurately than current convergence. allows different default actions for PBC. IOp(5/37) Frequency at which to do full Fock formation instead of incremental (L502). Don't checkpoint. Yes. VarAcc allowed. IOp(5/39) New On-Fly symbolic matrix element generator. decide based on details of problems. VarAcc forbidded because of guess=read. Yes. Checkpoint. IOp(5/38) Whether to vary integral cutoffs during direct SCF: 0 1 2 3 4 5 6 Default (same as 1). Every Nth cycle. do integrals at same accuracy as convergence until final iteration. -1 0 N Do not do incremental Fock formation. then 2 digits more Converge to 10**-5 with integrals good to 10**-6 first. 1 2 3 Hartree-Waller functions for singlets Hartree-Waller functions for triplets Slater Determinants 81 . IOp(5/36) Whether to checkpoint after every SCF cycle. REQUIRE 'NOFULLDIAG' Remember: the first digit indicating the type of function to be used. 0 1 2 Default (No). must be set.IOp(5/35) Whether basis is orthonormal (L501. Default (every 20 for direct). 0 1 2 Default (checkpoint only if direct). L502). No. No. Yes. accurately. then full convergence.

Default is DON'T EXCLUDE any integral Use sum of the first IState roots of a Reduced Hamiltonian as guess for Lanczos Use IState-th root of a Reduced Hamiltonian as guess for Lanczos Save first IState roots on disk for Davidson (this option will automatically call Davidson instead Print S**2 Print S**2 and its orbital components of Lanczos) IOp(5/40) Use of reaction field. Option xx is used also by Lanczos if IOp(39)=10000n or 20000n Maximum dimension of reduced Hamiltonian used as guess Default=Min(NSec. only used now for Onsager and control of details of SCIPCM -N 0 N 00000 10000 20000 30000 40000 50000 Multipoles of order N. store field in Gen(2-4) Default (same as 10000). Default is DON'T EXCLUDE any Use cutoff = 10**(-yy) on the product Integr*DenMat.xx0 integral yy000 100000 200000 300000 1000000 2000000 Use cutoff = 10**(-xx) on integral value to exclude contributions. Maximum allowed change is 10**N larger than RMS.e. Multipoles of order N.50) Maximum dimension of iterative subspace. L510: xx yy00 zz0000 k000000 Davidson options. Update surface every iteration in pass 1 only. Update surface and restart DIIS when within 10**-2 of convergence. WARNING !!! ratio (zz+k-1)/k must be equal to n. IOp(5/41) Whether to converge on maximum density change as well or instead of RMS: 0 N -1 -2 N0 in L510. convergence only on maximum). Default=60 Number of vectors provided in input BEWARE !!! Davidson will execute zz updating per iteration. Update surface on pass 2 iterations only. Maximum allowed changed is same as RMS (i. value must be 0<=ll<=20 Default=0 82 22.. increment field in Gen(2-4) No. Default=IVEC Reduction factor between number of guess vectors provided and number of vectors wanted at the end (1<=k<=9). Same as 3. Converge only on RMS density change. Converge on energy to 10**(N)*RMS-density-accuracy Also control of iterative diagonalization . Update surface every iteration. Default=1 (no reduction) ll0000000 Davidson iteration at which to scale back the number of vectors WARNING !!! For overflow reasons. but re-use 1e matrix instead of surface terms. number specified in nroot=n.

IOp(5/47) In L510. 1 2 Prepare data for Mp2 (l906 obsolete) Compute transition spin density and SO coupling IOp(5/48) Options to be passed to CalDFT: 83 .IOp(5/42) Number of orbitals to localize in L510 1 n Localize all active orbitals Localize first n (strongly occupied!) orbitals IOp(5/43) L509: Whether 5th order terms are treated explicitly 0 1 2 Default: set to 1 All 5th order terms are treated implicitly (Debug option) 5th order GG and WG terms are explicitly computed in L715 L510: DFT corrections to MCSCF on last iteration 0 1 2 No Yes. If negative reads the vector from rwf.7: the displacement in geometry in internals in Angs. 1 to set up for CAS-MP2 or 2 to do spin-orbit calculation.8 f10. If positive reads vector from input 4f20. 10 Include the CSF contribution to the orbs for the DerCpl. Uses MC-SCF density to compute B88 + LYP energy (These are hard-wired since they were the only choices that gave sensible results) Replaces diagonal elements of MC-SCF CI with B88 + LYP energy IOp(5/45) Numerical Derivative Coupling calculation(for testing) 0 1 No Yes (Needs NonStd root and two cards in input stream): i3 the other vector which coupled with iVec.

0 1 Yes canonicalize. -1 0 N None Default. or 22 if sparse is set in ILSW). Linear storage (only in Fock formation if diagonalization). Square storage (only in Fock formation if CG). L510: configuration cutoff for mp2 0 i 84 . L510: Option for using lanczos in CPMCSCF calculations 0 1 2 No Yes Use lanczos except for the last iteration IOp(5/50) L510: Option for setting the maximum number of lanczos iterations in Direct CPMCSCF IOp(5/51) L510: Canonicalize MC-SCF orbitals by diagonalization of Core and Virt Fock operators.1 Float(1/i) .) IOp(5/52) Amount of memory to allocate to stashing integrals. L510: Option for using reorthogonalization procedure in Lanczos 0 1 No Yes IOp(5/49) Use of sparse storage and Conjugate Gradient optimization instead of N**2 memory and diagonalization. 0 1 2 10 20 Default (11. Speeds up convergence in CP-MCSCF Do not Canonicalize (turn this on to maintain compatability with previous versions of code. Diagonalization Conjugate gradient.N Control flag for CalDFT is N. N words. also none.

IOp(55) Number of steps over which the SA coefficients are brought from (0. IOp(56) IOp(57) IOp(58) Threshhold for switching on. Default is 0.0) to (0. default parameters Dielectric constant |N|/1000 IOp(5/55) How many HOMOs and LUMOs to solve for after CG: 0 N None. (Or reversed. L510: see below IOp(5/56) A0 for Onsager SCRF. N of each.050 Threshhold for switching off.0-1. then start after instability searches. N N/1000 Bohr.0.001 .5-0.1 unless doing stable=opt. the calculation is started with coefficients (0.5-0. The state averaging will be switched off (graduately) when the degenerate region is left.lt. Usefull for optimizations or trajectory calculations where only a part of the surface is (nearly) degenerate. Energy difference smaller than IOp(56)*0. when the SA calculation graduately is switch off.IOp(5/53) PCM input and solvent type. in link 1003 a normal frequency calculation is performed when the optimization is in 85 . L510: See below IOp(5/57) First iteration at which to level shift and do FON. L510: See below IOp(5/55-58) L510: Switching on a State Averaged calculation graduately.) When IOp(55).001 .075 When set. Energy difference larger than IOp(57)*0. the number of steps for switching on/off is -IOp(55). 0 Default . which is usefull when a trajectory is started at a degenerate region.5).5). N>0 N<0 Solvent type N. Default is 0. In that case.

0-1. IRanWt.0/0.2/0. for a hypothetical example with nroot=2: cycle 4 5 6 7 8 old SA 0. use N. Turn off FMM here regardless.Threshold for a determining an adiabatic hop threshold=n*0.2/0. instead of a State Averaged second derivatives calculation.9 0.5 crash new SA 0.6 0. this can cause jobs sensitive to SA to fail to converge The new method checks the gap each cycle and will decide to increase or decrease the SA based on that result.3/0.7 0.1/0.a region of (0. only used with IOp(97)=11 IOp(5/60-62) Override standard values of IRadAn.3/0.4/0. Yes.0).8 0.1 Egap <limit <limit >limit >limit >limit Note: trajectory option set in overlay 10: IOp(10/99)=n -. and IRanGd.0001 (checks Offdiagonal element). regardless of weather that continues to decrease after the switch point.8 0. IOp(5/64) Over-ride default value of FMFlags 0 N No. 0 1 Use global default.2/0.8 0.7 0.0) (default) Allways SA second derivatives THESE OPTIONS MUST BE SET IDENTICAL FOR OVERLAY 10! (only IOp(10/55) needs to be set in overlay 10) IOp(5/59) IOp(59) 0 1 (Default) No action Turn on new more flexible SA options: The old method switches SA on when you reach certain energy gap. 86 .0-1. 1: 2: Normal computation when (0. IOp(5/63) Whether to do FMM.5/0.

IOp(5/74) Type of k-point integration: 0 1 2 3 9 10 20 90 100 200 Default (911). -2 -1 0 N None Start at a high temperature (limited only by DIIS error). 6000K for PBC) N degrees IOp(5/71) Number of steps to apply simulated annealing (L502): 0 N Default -. or temperature for broadening (PBC and IOp(74)=[1-4]xx). 0 N No.10 steps FON / 20 steps PFON N steps IOp(5/73) Options for ADMP: 0 1 2 Default (2) Use Lowdin basis for CP orthonormal transform.IOp(5/65) Over-ride NFx parameter. Use Cholesky basis for CP orthonormal transform. Yes. IOp(5/70) Maximum Initial temperature for FON (non-PBC). Occupy lowest NE at each k point regardless of the energies Improved LT with quadratic corrections Original LT method No concern for corrections Smearing Marzari method I Smearing Marzari method II 87 . Default (3000K = 10 milliHartrees for non-PBC. Use LT method (interpolation) Occupy entire points (used together with broadening) Full points for insulators. temperature broadening for metals. use N.

a value will be chosen in ZInLT1. 0 Default. No preconditioning. IOp(5/80) The maximum conjugate gradient step size -1 0 MMNN No maximum step size Default maximum (. Don't use CG DIIS Use CG DIIS Polak-Ribiere CG minimization Fletcher-Reeves CG minimization Use diagonal preconditioning in Conjugate-Gradient.NN IOp(5/81) Conjugate-Gradient Parameters MM NN00 00000 10000 000000 100000 0000000 1000000 Maximum Number of CG cycles per SCF iteration. (defaults to 4 CG cycles). and beta orbitals for fractional occupation. (defaults to 3 cycles). Maximum Number of purification cycles per CG iteration. alpha orbitals.G. IOp(5/82) C.300 400 500 900 First order Hermite-Gaussian of Paxton and Methfessel Gaussian smearing Classical Fermi-Dirac broadening No broadening (this will be Gaussian broadening with small T) IOp(5/75-78) Number of alpha electrons. beta electrons.8) Step size of MM. Convergence criterion 0 N 88 Defaults to 10**(-7) 10**(-N) . IOp(5/79) Range around Fermi level where temperature distribution will be applied if broadening is turned on for PBC.

89 . Level shift only if the HOMO-LUMO gap is zero or insignificant (>-0. Level shift to a maximum of the Goal. Turn off 3 and 4 point extrapolation if DIIS is on. up to twice the goal No longer used.1). Just continue as usual if energy goes up. IOp(5/86) Strategy options 000000 0 1 2 3 4 10 20 100 200 1000 2000 3000 4000 5000 6000 N0000 100000 Default (101100). Reduce DIIS space when energy rises from previous cycle.IOp(5/83) Maximum SCF DIIS vectors 0 N Default (20). IOp(5/85) Over-riding of maximum cycles for XQC. Level shift as much as necessary for HOMO>LUMO. Turn off level shift after energy rises. Reduce DIIS space whenever energy is above the lowest energy. Yes. Turn on dynamic level shift from the beginning Turn on dynamic level shift only after FON is over. Reduce DIIS space when energy goes above the lowest energy. Restart using Fock matrix in L502. -1 0 N Default for first step (128). Default (1 except during Stable=Opt. then 4). 0 1 No. Level shift only if the HOMO-LUMO gap is zero. Level shift to a maximum of 2*Goal. No. Limit is N cycles. Use SCF DIIS with N vectors IOp(5/84) Restart in L509.1) Level shift only if the HOMO-LUMO gap is zero or insignificant (>-0. Keep level shift after energy rises.

this and ZFormV should be moved to L302. separate Coulomb and NFx exchange for PBC). 10**-N. 1 2 Yes. Yes. do preliminary AM1 calculation and compare with guess from previous step in geometry optimization. IOp(5/88) L510: controls the amount of printing. Retain 3 and 4 point extrap. set in FoFDir/FoFCou/CalDSu based on accuracy part of IOp(5). if DIIS is on. Typically 10^-10 for molecules and 10^-12 for periodic systems.200000 off. IOp(5/92) Whether to use FoFDir or FoFCou for exact exchange: 0 90 Default: normal processing based on FMM for non-PBC. do preliminary AM1 calculation. IOp(5/91) Control option for chebyshev sparse control. IOp(5/90) Whether to generate sparse guess here. The energy is only checked after FON has been turned IOp(5/87) Accuracy criterion in Fock matrix formation: 0 N Default. . variable=kPrint 0 1 2 3 Print only summary information Print the a(t) vector and probability for each csf Print almost everything for debugging Print everything for debugging warning! this is a lot of stuff and you will only be able to do a few cycles IOp(5/89) Linearly dependent basis control for PBC.

01) or the lower state (=(1-x)*0. L510: has different meaning depending on if you are using IOp(97)=22 or IOp(97=23) If IOp(97)=22. FoFCou for Coulomb. which determines what type of gradient calculation to do: xxN Choose the basis variable=mBasis 0 1 2 3 xNx (Default) same as 1 Use a(t) basis Use mcscf basis orthogonal to a(1) and a(2) Use currnet mcscf vectors to check the code must be used with mHTest=2 91 Do with diagonalisationof Ecc variable=mTDGrd: .75 in ES (or 0.1 fempto seconds If IOp(97)=23. N iterations. value=xxx. variable=IWait. Value = xxxzyy. wait time is xxx*0.25 in GS) and if Zyy=125. where: Zyy yy Z xxx Threshold for a hop down determined by probability of being on the upper state (=x*0. separate FoFDir for exchange. separate FoFCou/NFx for exchange.125 in GS How long to wait before checking for a hop after going through a hop. L510: flag hopping controls starting and stoping options (x=0 or x=1) variable=iBack for first hop being up: xxxx0 xxxx1 variable=iStNow xxx0x xxx1x variable=iSpace xx0xx xx1xx variable=iEnd x0xxx xNxxx 0xxxx Nxxxx Use full space CI basis Use reduced space in projection of alpha number of cycles to carry on before stopping the timedep code after exiting the hopping region Default stop 6 iterations later Stop time dep on cycle N after exit (if IOp(92) is negative then stop immediately) No effect Stop time dep on cycle N allowable values 1-9 Use Energy gap criteria to start timedep Start timedep imediately Hopping down (forward) Hopping up (backward) variable=iFcTD to stop the time dependent code on cyle Z IOp(5/93) Number of initial iterations for which damping is allowed: 0 N Default (10). threshold =0.1 2 FoFCou for Coulomb. threshold for a hop up (after a hop down) = 0.01) Determines a variable Ulimt which lies between 01 and 99 Determines if this probability is halved after the first hop default Zyy=25.

05 switch timdep off only used with IOp(97)=22 and IOp(97)=23 IOp(5/95) Option for using Davidson in RFO calculations 0 1 Yes No use Lanczos (not recommended) IOp(5/96) Over-ride IRadAn for CPHF-like step in L509. We also need to define the maximum number of holes in the Ras1 space (ie the number of electrons that can be excited out of the Ras1 subspace. Ras2 and Ras3. default xxx=4 ie deltaE<0. Ras3 orbitals are empty. and for pass 0 grid in L502. Use grid N. and the maximum number of electrons in the Ras3 space: zzyyxxww. 0 N Use default. RAS control in L510.01 (checks energy gap). where ww 92 Number of Ras1 orbitals . on the model system L510: Threshold for turning propagation method on and off yyy xxx The first three digits determine the energy gap for turning off The last three digits determine the energy gap for turning on Threshold =xxx*0. The CAS active space is subdivided into three RAS active subspaces. Ras1. default yyy=5 ie deltaE>0. variable=mHTest IOp(5/94) PCM/ONIOM calculation 0 1 2 Standard PCM calculation PCM/ONIOM calcn. Ras1 orbitals are doubly occupied. In the reference space. on the real system PCM/ONIOM calcn.0 1 2 N00 0 1 2 information (Default) same as 1 Only check Ecc if MCSCF energies are almost degenerate Force check on Ecc by diagonalising it (Default) no testing Testing construction of ecc Testing construction of the ecc or exx portion of hessian using the mcscf vector 3 Calculate but do not use TD gradient Testing options.04 switch timdep on.

No.2. for detailed control of the propagation conditions see options 92-96. Needs option 80 and option 99. hop is based on the probability of being in a specific state Propagation of the wavefunction but no hopping. 93 .8 for 1.xx yy zz Maximum number of holes in Ras1 Number of Ras3 orbitals Maximum number of electrons in Ras3 IOp(5/97) Whether to update precomputed grid data with timing information.3d IOp(5/100) Tight convergence during CGDMS: 0 1 2 Default (No). Vector following or Root following hop alone (Needs option 80) Make a hop based on the secular equation (adiabatic hop) this option also includes a hop decision Debugging Option: Propagation of the wavefunction is switched on. 0 1 2 10 11 21 Default (Yes. 0 Default: 32. No. but the hop is determined by based on the vector following method (diabatic hop) (Needs option 99 and 80) the diabatic or adiabatic criteria (whichever determines a hop first). IOp(5/99) Grid for numerical k-integration in FT-LT method. Hopping threshold during trajectories with L510. the molecule continues in a mixed state (Needs (Needs also option 80) (detailed control is determined by options 92-96) option 71 and 80) IOp(5/98) Whether to save eigenvalues and orbitals at all k-points. Yes. Yes. 0 1 2 Default (No).12. if available). 22 23 Propagation of the wavefunction. No. Yes.

IOp(5/101) SDif test on numerical accuracy of PBC diagonalization. N. 0 -1 N>0 Default (10) No test Abort of SDif is larger than N. L510 Notes These options must be set in multiple links: L1003 iop(97) iop(55-59) iop(80) yes yes yes yes yes no no no yes L510 L118 These options must be set for the following links: IOp l118 l510 l1003 55-58 no yes yes 80 yes no yes 97 no yes yes 98 no no yes 99 no no yes 94 . IOp(5/102) Maximum number of configurations for CAS-MP2: 0 N Default (1000).

ONE CARD PER CENTER WITH THE X.Z2 -. Constrain the dipole in fitting charges. Read in a set of cards specifying a grid of points at which the electric potential will be computed.KTape The coordinates of N points in Angstroms will be read unit LTape in format (3F20.Z1 N2. POTENTIAL.XO. Evenly spaced rectangular grid. STARTING AT EACH POINT. Read in centers at which to evaluate the potential from the rwf. READ IN COORDINATES AS FOR 1. with N2 values in each record.NEFG. B. EVALUATE THE PROPERTIES ONLY AT EACH ATOMIC CENTER. READ ADDITIONAL CENTERS.Z-coord. IF MORE THAN ONE OF THESE IS REQUESTED.Y1. N1 records will be written to unit KTape.XX-EFG YY-EFG. or potential. Only one card is needed: N. do not default cube.YO. Read grid. 8 16 32 128 Do potential-derived charges.LTape.XY-EFG.Potential X-field. containing: X-coord. If KTape is 0.Y-field.number of increments and vector.12). Two forms of specifications are allowed: A. An arbitrary list of points. potential and field (NEFG=2). LOCATED THE NEAREST STATIONARY POINT IN THE ELECTRIC.12).Y-coord. 0 1 2 4 NO ADDITIONAL CENTERS. The potential (NEFG=3).number of increments and vector. Thus if NEFG=3 for each point there will be 4 cards written per point.X2. Y AND Z COORDINATES IN ANGSTROMS (FREE FORMAT).Overlay 6 IOp(6/15) SPECIFICATION OF ADDITIONAL CENTERS.output unit and coordinates of one corner of grid. -. -. and field gradient (NEFG=1) will be computed and written along with the coordinates to unit KTape in format (4F20.X1. it defaults to 51. 95 . THE LISTS ARE IN SEPARATE INPUT SECTIONS IN THE ORDER LISTED BELOW.XZ-EFG YZ-EFG Note that either form of grid should be specified with respect to the standard orientation of the molecule.ZO N1.Z-field. field. Three cards are required: KTape.Y2.ZZ-EFG.

IOp(6/20) How to do electrostatic-potential derived charges: 0 -1 1 2 3 00 10 20 30 100 96 Default (1).Rad) or (Symbol. This is like IOp(9) in L9999. IOp(6/17) DEBUGGING CONTROL (L602). terminated . same as 1. COMPUTE ONLY THE NUCLEAR CONTRIBUTION. Force CHELP (Francl) recommended radii. of points. Read in replacement radii for selected atom types as pairs (IAn. so rotate. COMPUTE ONLY THE ELECTRONIC CONTRIBUTION. one per line. Merz-Kollman point selection CHELP point selection. CHELPG point selection. use as-is. COMPUTE ONLY THE CONTRIBUTION OF SHELL N. 0 1 Use full accuracy in calculations at specific points. Exact is already in z-matrix orientation. Exact is still in standard orientation. Force Merz-Kollman radii. Default radii are those defined with the selected method.IOp(6/16) Cutoffs in L602. 0 1 2 -N COMPUTE ALL CONTRIBUTIONS TO SELECTED PROPERTIES. Force CHELPG (Breneman) recommended radii. 0 1 2 Default. Read a list of points at which to fit. IOp(6/18) Whether to update dipole rwf 0/1 yes/no.Rad). IOp(6/19) Whether to rotate exact polarizability before comparing with approximate (which will be calculated in the standard orientation). but use sleazy cutoffs in mapping a grid Do all points to full accuracy.

Use all available density matrices. No. IOp(6/24) Frozen core: -N 0 1 2 Freeze N orbitals. Default (Yes). Optimize density basis set. Read in density basis functions and compute populations. Read density matrices from . L604): -1x +1x -5 -4 -3 -2 -1 N.chk file. gradients and divergence. 97 . Yes. terminated by a blank line.chk file.by a blank line.0 Read density matrices from . All available transition densities. Density for the excited state given by IOp(29). Density values. IOp(6/21) Operation of L603: 0 1 2 Default (same as 2). Density values. Use the density matrix for the current method. or the HF density if the one for the current method Use the density matrix for method N (see Link 1 for the numbering scheme). L602.ge. 200 1000 10000 Read in replacment radii for selected atoms as pairs (I. Density values and gradients. Fit united atoms (heavy atoms only) rather than all atoms. is not available. IOp(6/22) Selection of density matrix (currently only in L601. Transition density between the states given by IOp(29) and IOp(30).Rad). Use only active atoms in the fit. IOp(6/23) Density values to evaluate over grid in L604: 0 1 2 3 Default (same as 3).

Do bonding Mulliken Populations. O(N**4)).requires 2e integrals. Total. Do Minimal population analysis. classically (including self terms . IOp(6/26) Which density to use in L602 and L604: 0 1 2 3 4 Default (same as 1). Spin. IOp(6/27) Choice of population analysis: 0 1 2 10 20 100 Default (12). and only available for HF.requires 2e integrals. Yes. Do Mulliken populations. IOp(6/28) Mark SCF density as current density. Yes. Beta. Yes: mark as well. 0 1 No: save SCF density. No.IOp(6/25) Whether to compute coulomb self-energy in L601: 0 1 2 O(N**5)). quantum mechanically (no self terms . IOp(6/29) Excited state to use if requested by IOp(22). Alpha. 98 . but do not mark. Don't do bonding Mulliken Populations. Don't do Mulliken populations.

attractor interaction lines. Yes.11' IOp(6/35) WHAT TO DO: 0 1 2 4 10 100 Determine attractors. 1: Do not print the atomic occupancies. default 6).. Use spin density. using alpha and beta separately for UHF. default 9).99. This precludes calculations of atomic property derivatives with respect to nuclear displacements. Transition denstiy between state IOp(29) and state N. 10. Compute charges of AIMs (IDoAtC). where MItLoc ITlLoc IDcInt IPrLoc MItLoc*NOrb*(NOrb-1)/2 is the maximum number of iterations in localization of (spin)orbitals (1.IOp(6/30) 2nd excited state for transition density: 0 N Transition denstiy between state IOp(29) and g.. L606: naming of RPAC interface file. Yes. and cage points. Localized (spin)orbitals with atomic occupancies less than 0. IOp(6/31) Whether to determine natural orbitals from densities: 0 1 2 3 4 5 No. 0 1 Make this a scratch file. 99 .**(-ITlLoc) is the convergence criterion for (spin)orbital localization (1. Compute kinetic energies and multipole moments of AIMs (IDoPrp).01*IDcInt are interpreted as lone pair MOs rather than bond MOs (1.9..9. Store only alpha NOs. using total density.s. Store only beta NOs. default 10). IOp(6/32) CONTROL PARAMETERS FOR COVBON in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 10000*MItLoc+1000*ITlLoc+100*IDcInt+IPrLoc... Compute energies of electrostatic interactions between AIMs (IDoPot). ring points. Name this file 'rpac. Determine zero-flux surfaces (IDoZrF). Compute atomic overlap matrices (IDoAOM). 0: Print the atomic occupancies of localized (spin)orbitals (default). L605..

Number of grid points in critical point search (1.200 400 1000 Compute other atomic matrix elements (IDoAMa)..**(-IPrNDe) in evaluations of electron density and its derivatives (0 default 7). No other properties can be calculated. augment valence electron densities with nonrelativistic core contributions Abort if pseudopotentials have been used (IHwAug=3).99. default 7). Note that IDoSRe should be set to 1 in order to obtain correct results! Also note that analytical polarizabilities have to be available but force constants have to be absent! 2000 10000 20000 100000 200000 400000 1000000 2000000 3000000 Compute derivatives of atomic properties with respect to nuclear displacements as well (IDoNuD). Length of Fourier expansion for the trial path (1. and IEpsIn=100. This option sets IPrNDe=5... where MxBpIt SBpMax NGrd LookUp Maximum number of iterations in trial path determination (1. If necessary. (IHwAug=1). which is a default anyway (IHwAug=0). Include zero-flux surface relaxation terms in all atomic matrix elements (IDoSRe) Compute derivatives of atomic properties with respect to electric field (IDoSeP). default 2). Use numerical instead of analtyic integration.. Note that analytical force constants have to be available! Compute localized orbitals and bond orders (IDoLoc).. zero occupancies... default 10). Use numerical instead of analtyic integration and use reduced cutoffs. augment valence electron densities with relativistic core contributions. default 20). default 100)..**(-IPrNAt) in integrations over atomic basins (099. Compute atomic orbitals in molecule (IDoAOs). Neglect primitive contributions below 10. 1: Do not ignore (spin)orbitals with Neglect primitive contributions below 10. where INoZer IPrNDe IPrNAt 0: Ignore (spin)orbitals with zero occupancies (default). RNGPNT. IOp(6/38) CONTROL PARAMETERS FOR ZRFLUX AND OIGAPI in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 100 .999. IOp(6/37) CONTROL PARAMETERS FOR ATINLI. IPrNAt=5. Maximum value of the control sum (1.9. AND CAGPNT in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 1000000*MxBpIt+100000*SBpMax+1000*NGrd+LookUp. 99. If necessary.99. IOp(6/36) CONTROL PARAMETERS FOR NEGLECT OF ORBITALS AND PRIMITIVES in L609: 10000*INoZer+100*IPrNDe+IPrNAt. Reduce accuracy so atomic charges can be computed more rapidly (IQuick).

9..99. IHowFa is the maximum distance in the Runge-Kutta integrations along gradient paths (1.9.. 10*INInGr is the initial number of grid points in theta and phi in the adaptive integration subroutine (1...9.. default 2).9.. 0. N layers. IOp(6/40) Control of link 607...100000*INStRK+10000*IHowFa+1000*IGueDi+100*IPraIn+10*IRScal+IRtFSe INStRK IHowFa IGueDi IPraIn IRScal IRtFSe 10*INStRK is the number of steps in the Runge-Kutta integrations along gradient paths (1.9. default 5). This link must have been invoked with IOp(40) = 0 or 1 prior to invoking it with IOp(40)=2.9.9.0001*IEpsIn is the target for integration error (1..*IRtFSe is the safety factor used in the intersection search (1. default (1. whose energy can then be computed by one of the SCF links. must be >=4.. 0 1 2 Default NBO analysis ..*IPraIn is the cut-off for zero-flux surfaces (1. 0 N Default (4). (1... default 2). default 2). IOp(6/41) Number of layers in esp charge fit. default 6).**(-IGueDi) is the initial displacement from the critical point in the Runge-Kutta integrations (1.**(-5-IToler) is the tolerance for the intersection search (1. Read input data to control NBO analysis... 5+INInCh is the initial number of sampling points in the intersection search (1. IEpsSf IEpsIn INTrig IEpsSf is the safety factor used for patches with surface faults in the adaptive integration subroutine (1.. 10*INTrig is the number of sine and cosine functions in the trial function for surface sheets 10.. 10.9. IRScal is the scaling factor in the nonlinear transformation used in the intersection search 10... 101 .... default 2).9. IOp(6/39) More CONTROL PARAMETERS FOR ZRFLUX AND OIGAPI in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 1000000*IToler+100000*INInGr+10000*INInCh+1000*IEpsSf+10*IEpsIn+INTrig IToler INInGr INInCh 2). default 2). default 5). default 6). Delete selected elements of NBO Fock matrix and form a new density. default 2).9.. default 2).. 10. 3 Read the deletion energy produced by a previous run with IOp(40)=2 and print it.don't read input.9.

same as 2. 0 N Default (0. 0 N Default (1).10**-3 Read from input. N*10**-4. IOp(6/45) Number of points per bohr**3 for Monte-Carlo calaulation of molar volume -1 0 N Read from input Default (20) N points .IOp(6/42) Density of points per unit area in esp fit. points per unit area. 50 is recommended. IOp(6/46) Threshold for molecular volume integration.for tight accuracy.4/Sqrt(#layers)) 0. 0 -1 N Default . IOp(6/44) Type of calculation in L604: 0 1 2 3 10 Default. 102 .01*N. Compute the molar volume Evaluate the density over a cube of points Evaluate MO's over a cube of points Skip header information in cube file. IOp(6/43) Increment between layers in MK charge fit.

Yes. on non-SCF densities. N*0. IOp(6/51) Whether to apply Extended Koopman's Theorem (EKT): 0 N -1 -2 Default (No).IOp(6/47) Scale factor to apply to van der Waals radii for the box size during volume integration: 0 N Default. N. 103 . 10**-N IOp(6/49) Approximate number of points per side in cube in l602/l604: 0 N -1 -2 -3 -4 Default (80) N points Read from cards.01 . up to N IPs and EAs. 6 points/Bohr. Coarse grid. Fine grid. 12 points/Bohr. IOp(6/48) Use of cutoffs 0 N Default (10**-6 accuracy for cubes. No. Yes. -N>4 Grid using 1000 / N points/Bohr. 3 points/Bohr. 1 digit better than desired acuracy for volumes). on non-SCF densities. IOp(6/52) Number of radial integration points in L609: 0 N Default (100). Medium grid.for debugging. all possible IPs and EAs.

0 N Default (100000). read in name for output file. Must be an unpruned grid.IOp(6/53) Distribution of radial points in L609: 0 N Default (cubic) Polynomial of order N. N. IOp(6/56) Whether to read in density matrix from input stream in L608. Yes. IOp(6/57) Whether to generate data over a grid using the total SCF density: 0 1 2 3 No. IOp(6/55) Number of inner angular points in numerical integration in L609: -1 0 N 0 (no inner sphere) 302 N point Lebedev grid (see AngQad). IOp(6/58) Grid to use in generating tables of density and potential. Yes. 0 Default (99001). Yes. 104 . IOp(6/54) Maximum number of domains. 0 1 No. also read in name for input file with a different grid and compare. Output in the form of data statements.

of isotropic value. Do correct energies and 0th order approximation Do correct energies and 0th-1st order approximations Do correct energies and 0th-2nd order approximations IOp(6/60-62) Over-ride standard values of IRadAn.IOp(6/59) Approximations to Exc -1 0 1 2 3 Test superposition of atomic densities using L608: Do correct energies. Do test as usual. of diagonal tensor elements and isotropic value. Default (1 pmm). Yes. of all tensor components. N/1000 ppm IOp(6/70) Control of L610. and IRanGd. Yes. -1 0 N Zero. IOp(6/65) Threshold for printing of NCS contributions. IRanWt. Suppress test. IOp(6/64) Natural Chemical Shielding Analysis: 0 1 2 3 No. Yes. IOp(6/63) Suppress number of electrons test in XC quadrature in L608 (for debugging with small grids): 0 1 2 Default (do test). 105 .

NMR data is available. if open-shell. Yes. IOp(6/74) Whether to use Gaussian connectivity in choosing Lewis structure for NBO. Suppress re-orthogonalization. Use.IOp(6/71) XC functional in L610. reading isotopes IOp(6/73) Whether to save orbitals from NBO: 0 1 2 3 10 Default (No).D-6). Save NLMO occupieds and NBO virtuals. regardless of other terms. and other terms are being computed No. Save NLMOs in place of regular MOs. Don't use. IOp(6/76) Threshold for linear dependence in L607. 0 N 106 Default (1. . Save NBOs in place of regular MOs. IOp(6/72) Whether to read isotopes for hyperfine interractions and do hyperfine terms in L602: 0 1 2 3 4 Default (1). IOp(6/75) model for CM2 charges. Yes. Yes. 10**(-N). 0 1 2 Default (use if present and choose is selected in NBO input).

IOp(6/80) Whether to calculate Lowdin charges and Mayer bond orders.IOp(6/77) Restraint in charge fitting in L602: 0 -1 N None. Yes. IOp(6/82) Tensors for hyperfine spectra. 0 1 2 Default (No). if doing other orbital results). Yes. for the top 5 occupieds and lowest 5 virtuals. Yes. No. 2. Yes. for all orbitals. IOp(6/81) Print kinetic energy of orbitals? 0 1 2 3 Default (yes. 107 . No.d-4 N * 10^-5. IOp(6/78) Use MOs instead of density in AtmTab. Use density. Use MOs. No. 0 1 2 Default (2). IOp(6/79) Whether to calculate Hirshfeld charges. 0 1 2 Default (No).

2 Default. compute if there are 100 or fewer atoms Compute QEq tensors and for open-shell systems compute isotropic and anisotropic splitting Do not compute tensors.0 1 tensors. 108 .

Overlay 7 IOp(7/6) operation of link 705 (NYI). No. Yes. Print both HP modes and internal displacements. Form the ecp contribution to the SCF forces. do hindered rotor analysis. 0 1 2 10 20 Default (12). IOp(7/9) Whether to rotate derivatives back to the z-matrix orientation. Print normal mode displacements in redunant internals. Print higher precision normal modes. with most common isotopes. Default scale factor is #N (1=HF. 0/1 yes/no. 109 . with read-in isotopes.91671. If M=2. Yes. IOp(7/7) USE OF INTERNAL COORDINATES.12. 1/1. If M=3. Do not form 1e derivative matrices. only harmonic thermochemistry. Increment the 1e derivative matrices with ecp terms. but neglect first derivatives in conversion of second derivatives to internal coordinates. 3=CBSQ=0. Do not the ecp contribution to the SCF forces. 2=CBS4=0. 0 1 2 YES NO Yes. IOp(7/8) Harmonic frequency calculation: 0 1 2 3 10 20 30 Nxx Mxxx No. Read hindered rotor parameters from input.91844) If M=1.

Scalar Rys DF. FoFDir HGP df. Rys f. FoFDir: HGP spd. FoFDir Rys df. FoFDir: Berny SP. IOp(7/13) Contraction with two-particle density matrices: 0 1 2 3 4 110 Default (same as 1). Old vector Rys SPDF. Use generalized density number N for both the one-electron integral derivatives and the corresponding 2PDM terms. old vector Rys DF. IOp(7/11) control of integral derivative algorithm: 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Default use IsAlg to decide.IOp(7/10) First/second derivative control. IOp(7/12) Selection of density matrix. FoFDir: HGP sp. Use external 2PDM. FoFDir: HGP spdf. PRISM df. Generate 2PDM from CIS square 1PDM (for debugging) . Berny SP. Berny SP. do both. Berny SP. Use both HF and external 2PDM. Berny SP. Scalar Rys SPDF. FoFDir: Rys spdf. FoFDir: PRISM spdf. Berny SP. FoFDir HGP d Rys f. do only second derivatives. FoFDir Rys df (same as 7). 0 1 2 do only first derivatives. Rys df. 0 N Usual SCF density. Use HF 2PDM. Berny SP.

Electric field Jth order. Default. 111 . IOp(7/14) State for CIS gradients. Use sleazier cutoffs in L703. while 6-7 imply use of the generalized density derivatives in L701. 0 IJK 1000 Default (1st order nuclear and electric field). N. and 11 are valid in overlay 7. DO AS ACCURATELY AS POSSIBLE in L703. Use sleazier cutoffs in L708. DO AS ACCURATELY AS POSSIBLE in L702. IOp(7/16) Number of translations and rotations to remove during redundant coordinate transformations: -2 -1 0 N 0. same as -1. Defaults to 1. Only 1. IOp(7/15) The nature of the perturbation(s). USE THE ORIGINAL 'BERNY' VALUES in L702. The types of derivatives are given by IOp(15). Generate simulated density derivatives. 10. IOp(7/18) Derivative accuracy option: 0 1 2 10 20 100 200 Compute to 10**(-8) accuracy.5 6 7 10 Generate 2PDM from CIS square 1PDM and use HF/Z 2PDM as well. Contract with external 2PDM derivatives. given by IOp(15) 0-5 imply use of the generalized density in L701. DO AS ACCURATELY AS POSSIBLE in L708. Normal (6 or 5 for linear molecules). The types of derivatives are Leave the external 2PDM on the disk instead of deleting it. Magnetic Field Ith order. Nuclear Kth order. Form derivative 2PDM from CIS and HF derivative density matrices.

same as 2. 0 2 COMPUTE AS NORMAL. PSCF freq=numer). 0 1 00 10 Just use the forces in IRWFX. Use FX in conversion of force constants to internal coordinates. Default method for initial force constants IOp(7/30) USE OF SYMMETRY IN OVERLAY 7: 0 1 USE (SUBJECT TO AVAILABILITY). Second.IOp(7/25) Type of derivatives available. Third. Normal + Generate estimated initial force constants. Normal NUCLEAR REPULSION ONLY (USEFUL FOR TESTING). 0 1 2 10 First. (HF freq. Compute HF forces from D2E file and increment both FX and FXYZ (non-O11 PSCF grad and HF freq). 0 1 2 6 00 Normal. DON'T USE. Use FXYZ in conversion of forces constants to internal coordinates (PSCF opt with HF 2nd deriv). IOp(7/28) SKIP OPTION TO DEFER INTEGRAL EVALUATION TO L703. Read derivatives from checkpoint file (in Z-matrix orientation). DO ALL GRADIENT INTEGRALS IN L703 IOp(7/29) MODE OF USE OF L716. 112 . IOp(7/31) Handling of forces contributions.

12 format.false. IOp(7/42) 1PDM: 0 N Use SCF total density. and derivatives in redundant internals only in compressed form.12 format in compressed form. Punch energy. first and second derivatives in both cartesian and internal coordinates. 0 1 Evaluate integrals correctly.IOp(7/32) PUNCH OPTION. and derivatives in cartesians and redundant internals. IOp(7/41) NDDO flag. Force constants and third derivatives are punched in 4E20. IOp(7/43) 2nd order simultaneous optimization flag. Punch nuclear coordinate derivatives. (other 2nd derivative options must also be set appropriately) 113 . IOp(7/40) Neglect of integrals (only option 1 works in Overlay 7): 0 1 2 3 10 20 30 Keep all integrals. 0 1 2 3 4 5 1x None. Use generalized density N. Do punch only if second derivatives are available.16. Neglect 2e integrals with diatomic differential overlap. 0 1 . Do only overlap and not other 1e integrals. coordinates. Neglect three center two-electron integrals as well. Neglect four center integrals. Apply NDDO approximation.8. . Neglect 1e integrals with diatomic differential overlap. forces and lower triangular force constants in format 6F12. Punch energy in format D24. Punch energy. coordinates.true. Punch energy. Forces are punched in 3D20. one card per atom. Neglect three center one-electron integrals.

Update forces for the self-consistent reaction field (SCRF) method Update forces for a uniform electric field. with forces done the usual way for CIS or MP2 2nd derivatives. IRanWt. 114 . Compute optimimum lambdas. IOp(7/60-62) IOp(60-62) Over-ride standard values of IRadAn. Project using forces if the RMS force is larger than 1. IOp(7/52) Whether ECP integrals should be done in L701 as usual. IOp(7/53) Convert forces over shells to field-dependent dipole and forces over atoms (for debugging): 0 1 10 No. Update forces for a uniform electric field. Project the reaction path and compute 3N-7 frequencies.2 . The point is a stationary point. No.true. (debugging option: compute fifth order WG and GG terms in L715) IOp(7/44) Handling of an applied electric field. Project using the Newton-Raphson step. Yes. and IRanGd. 0 1 Yes. IOp(7/45) Controlling the projection of the reaction path. -1 0 1 2 Do not add electric field terms to forces. 0 1 2 3 Do not project.d-6.

0 1 Use global default. IOp(7/70) Do vibro-rotational analysis: 0 1 2 Default (No). Lines Lorenzians Both IOp(7/65) Harmonic constraints with respect to initial structure during geometry optimization. Yes. 0 1 2 3 Default (1). -1 0 1 No. Default (Yes. IOp(7/72) Read additional parameters for anharmonic computations 0 No 115 . if ref structure is present and has non-zero force constants). Currently lots of hacks to determine where we are in the process instead of different values of this option. Turn off FMM here regardless.IOp(7/63) Whether to do FMM. IOp(7/64) Type of simulated spectrum in output. Yes. IOp(7/71) Do vibrational 2nd order perturbation: 0 1 No Yes. No.

The threshold is automatically lowered for each mode until 90% of the absolute displacements are included. Clear real density and do -1/2 P(Fit)*Jx*P(Fit). IOp(7/75) Threshold for printing redundant internal contributions to normal mode displacements: 0 N -1 Default (10%) 10**-N Zero (all printed). Yes.1 Yes IOp(7/74) Non-equilibrium PCM gradients: 0 1 No. Turn off 1c logic for 1c DBF case. 10^-12 for PBC. 10**(-N). Prohibit FoFCou. Copy fit density over real density and do P(Fit)*Jx*P(Fit). IOp(7/87) Accuracy in FoFDir/FoFCou/CalDSu: 0 N Default. IOp(7/77) Debuging options for DBFs: 0 1 2 3 4 Normal processing. IOp(7/76) Over-ride use of FoFCou in L703: 0 1 2 Normal processing. 10^-10 for molecules. Omit subtraction and do P(Fit)*Jx*P. 116 . Force FoFCou.

(IA/JB). 0 1 2 3 4 5 6 7 BUCKETS FOR MP2: (IA/JB).Overlay 8 IOp(8/5) Whether to pseudocanonicalize ROHF orbitals. otherwise same as 3. IOp(8/8) Whether to delete MO integrals in L811. BUCKETS FOR CID OR MP3: (IJ/AB). IOp(8/9) Debug control (L802): 117 . BUCKETS FOR SEMI-DIRECT MP4DQ.(IJ/AB). -1 0 Yes. Full transformation if this is consistent with MaxDisk. IOp(8/6) Bucket selection. CISD. THE COMPLETE SET OF TRANSFORMED INTEGRALS. 0 1 2 Default (No).(IA/JB).(IK/KL). 0 1 Test that SCF has convergd. Full transformation if this is consistent with MaxDisk. BUT INCLUDES (IA/BC). CISD or MP4SDQ or MP4SDTQ. QCISD. Do not test SCF convergence (mainly used for testing). BD: (IJ/AB). No. IOp(8/7) SCF convergence test. Yes. BUCKETS FOR STABILITY: (IA/JB). otherwise same as 4.(IJ/KA). No.(IJ/KL).

virt-rep2. 1 <= N <= 89 selects frozen-core type N: The largest noble gas core is frozen. L811): 0 1 2 3 10 20 30 100 200 1000 2000 3000 4000 00000 10000 20000 Operate normally. A negative value for the end deletes the top virtuals. Default (10000) Do not symmetry compress transformed integrals. If IOp(37) is 0. The next to the largest noble gas core is frozen. Accumulate MP2 force constant terms in direct fashion Write the MO basis first derivative ERI's to disk Force fully in-Core algorithm (L804 only).0 -N Operate normally. Force transformed integrals in Core algorithm. Do symmetry compress transformed integrals (buckets) (This will cause windowed MOs. .) 30000 100000 Symmetry compress transformed integrals only if RHF. except that the outer sp electrons of 3rd row and later alkalai and alkalai earth elements are retained. core virtuals are also frozen.occ-rep2. Force sorting for output bucks. virt-rep1.. Force N orbitals per pass. The window is recovered from rwf 569. Force output bucket in Core antisymmetrization. Use all MOs. The largest noble gas core and main group d's are frozen. (Upper triangle of symmetry compressed integrals for IOp(6)=5 or 4 only!) Reorder MOs.. a card is read in indicating the start and the end. eigenvalues and symmetry assignment vectors according to ther representations IOp(8/10) Window is selected as follows: -N 0 N 1 2 3 4 90 91 92 93 94 000 118 Use the top N occupieds and lowest N virtuals. For basis sets with double-zeta cores. Force semi-direct mode 3 if IOp(6)=3. Eigenvalues and symmetry assignment vectors will be put in correspondence with vectors. Read a list of orbitals to freeze. Generate and test RInt3 array (L804).. Default (200)... Default.. Direct Transformation Control (L804. The window is recovered from file 569 on the checkpoint file. G2 frozen-core: the largest noble gas core and main group d orbitals are frozen. reordered in the order of representations like occ-rep1. Force semi-direct transformation. The window is specified by IOp(37-38). Force semi-direct mode 4 if IOp(6)=3. Force semi-direct mode 2. Force semi-direct mode 1. same as 4.

Use overlap with atomic core orbitals from Harris to choose core orbitals. Suppress such a test (CPHF may still be done for such a case). 0 1 No. Kill the job if there is no correlation energy. IOp(8/12) Calculation of frozen-Core contributions. orbital energy. kill a frozen-core job if there there is significant core-valence mixing in the canonical orbitals 00 10 20 Default. IOp(8/11) MO coefficient.K over deleted orbitals and add to Core-Hamiltonian. if there is only 1 electron or 1 virtual spin-orbital. IOp(8/15) Control of DRT output. Kill the job if any mo coefficients are greater than 1000. Also..001.10x 20x 30x Use orbital energies to choose core orbitals. 0 1 Output to Gaussian system buckets. This is done when IOpCl=0 or when IOpCl=1 and the calculation is rohf or gvb. IOp(8/14) Control of drt input. 0 1 2 Default. Use overlap with atomic core orbitals from Core Ham to choose core orbitals. Read 'old-style' drt input cards. 0 1 Take necessary input from Gaussian data structures. 119 . Calculate 2 J . same as 2 Just print a warning message.g. same as 10.0 or the smallest difference between occupied and virtual orbital energies is less than 0. Output transformed integrals for DRT-CI calculation. IOp(8/13) Control of output. and number of electrons test. e. Used to select output mode.

0 N Default let program decide. except canonical in frozen-active blocks. 0 N Default excitation level = 2.x = -Sab. Canonical. Uij.x = -Sij. (only if integral derivative file is being written) Excitation level for SDGUGA-CI. IOp(8/16) Maximum number of orbitals per pass in L811. Uij.x = 0.x.0 1 2 Write DRT output to RW-files.x . Excitation level = N.x = -Sij. Do both. Uij. IOp(8/18) Which type of derivative transformation to do in L811: 0 1 2 3 4 5 6 Default. . Non-canonical.x = -1/2 Sij. Nuclear Kth order. Uab. Uij.x/2 Canonical virtuals.x = (Fij.x = -1/2 Sij.x.x) / (Ei-Ej) Note that this blows up for degenerate orbitals and is intended primarily for debugging. Write DRT output to Fortran unit 'drttap'.x/2 IOp(8/19) The nature of the perturbation(s) in L811: 0 IJK Default (1st order nuclear and electric field). Electric field Jth order.EjSij. Uij. Non-canonical. same as 3. Magnetic Field Ith order. Integral Block Size = N. IOp(8/17) Specification of integral block size for GUGA CI programs. IOp(8/20) Which terms to include in L811: 0 120 Default (same as 11).x Uji. Non-canonical. Canonical occupieds.

IOp(8/28) Hack number of occupieds for full ci using links 921 or 922: -1 0 N Transform all orbitals (after freezing Core) as occupieds (i. independant of MAXDISK. -3 -2 code. IOp(8/22) These options control the in-Core post-SCF link. set NOA=NOB=N for purposes of transformation). subject to the limit imposed by MAXDISK (IOp(27)). L805. Transform N orbitals (after frozen Core) as occupieds (i. This will determine the number of times AO integrals and derivatives are evaluated unless overridden by IOp(31).e.. Do as many in a batch as can be overlapped with sorting space for half transformed integrals. Do one batch. Default (-1) N evaluations and hence N coarse tiled batches (1.1 10 MO derivative times integral term. but use multi-batch logic.6 are the currently implemented options) 121 Use as much as desired. This only applies if the integral derivatives are not stored... IOp(8/27) Maximum amount of disk to use in L804 and L811: 0 N Unlimited.e.. MO times integral derivative term. N words. N triplets. IOp(8/29) Maximum number of perturbations per batch in L811: (only applies if integral deriv file is written) -3 -2 -1 0 1 N Do not use batching logic. Look there for more information. transformation). Default (same as 1). Do a single atom at a time (minimum disk usage). -1 0 N Use as much as needed for maximum efficiency. set NOA=NOB=NROrb in No. Requested disk usage. Use an amount which is similar to the maximum disk usage in other parts of the MP2 frequency .

no otherwise). IOp(8/35) Output format for closed-shell and debugging control: (only for when integral derivative file is written) 0 1 2 10 Default (consistent with integrals for open-shell.=j a<=b) for closed-shell AA and AB consistent with integrals. Default on whether to make Poo and Pvv for MP2. IOp(8/31) PERFORM PRIMITIVE POST-SCF OPERATIONS (NOT CURRENTLY FUNCTIONAL). Interacting only. 0 1 -1 Default. . up to 9. 0 N Let the program choose (make it as large as possible) maximum fine tile batch size.IOp(8/30) Type of window. Yes. Set up /Orb/ as indicated by IOp(10). Set up /Orb/ for a full window but then blank the wavefunction coefficients in L804. Explicit control of the fine tile batch size (largely a debugging option. (Yes if Ix is not stored. 0 1 NO YES IOp(8/32) Whether to do CI in the interacting space only. Store only the unique AB integral derivatives (gO2V2/4. only for no-Ix routines). 0 1 2 00 122 Default (Yes). All. Set up for full but zero Core MOs. No. i<=jab alpha-beta only for closed-shell). order g i. 0 1 2 Default (all spin-eigenfunctions). IOp(8/36) Whether to update force constants with the MP2 product of MP2 integral derivatives term (only applies if integral derivative file is not written). Test window. Do extra debugging computations.

If C1. Take reference MOs from disk if available. No for the 1st geometry of opt. yes otherwise. transformation. IOp(8/41) Transformation of spin-orbitals (alpha only) within occupied and unoccupied orbital subspaces by minimum orbital-deformation (MOD) method. No localization. Yes. forming pseudo-canonical alpha and beta orbitals and doing UHF Treat as RHF. IOp(8/38) Integers specifying the window to use. IOp(8/40) Handling of ROHF window: 0 1 2 Default (2). 0 1 2 3 00 10 20 Default. Population method. IOp(8/39) Localized orbital method adopted in SAC/SAC-CI. Rotate MOs to compute the displaced overlap. Boys method. No. Boys + population method. transforming only alpha orbitals. No. but save MOs. Force to rotate MOs. 0 No 123 .10 20 Yes. use unity as rotation. IOp(8/42) Whether to reorder MOs during potential surface exploration. No. Use ROMP2 approach. 0 1 2 3 Default. Force not to rotate MOs.

N -N Use N points for MP2. Default . 0 1 post-SCF. Ionization potential. N-th occupied band at the k-point for which the hoco is highest (by default) or at k-point N-th virtual band at the k-point for which the luco is lowest (by default) or at k-point specified specified by IOp(44) by IOp(44) IOp(8/46) Indicates special case of non-HF calculation. 0 N Use the k-point for which the hoco is highest and luco is lowest. K-point at which to apply correction. (for SAC-CI single point calculation) Use orbital energies in ordering Don't use orbital energies in ordering Use second moments in ordering Don't use second moments in ordering Use dipole moments in ordering Don't use dipole moments in ordering IOp(8/43) Number of Laplace points to use. Input orbitals are not canonical HF and pseudocanonical orbitals must be generated here for the 124 . IOp(8/44) K-point specification for MP2 band correction.MOs are canonical HF orbitals. IOp(8/45) Type of quasiparticle job: 0 -1 1 -N N Band gap.1 2 00 10 000 100 0000 1000 Yes. Use N points and set up for band gap correction. Yes. Electron affinity.

but blank contributions from inactive atoms.IOp(8/47) Whether 804/811 should generate results compressed over active atoms: 0 1 2 3 Default (2). Full list. No difference from 2 for overlay 8. 125 . Active atoms. Full list.

AND EUMP4 IN MP4 CALCULATIONS YES. CONFIGURATION INTERACTION WITH ALL SINGLE AND DOUBLE SUBSTITUTIONS. IOp(9/6) L913: CRITERIA FOR TERMINATION OF THE ITERATION 0 -1 N DEFAULT CONVERGENCE CRITERION AND MAXCYCLE READ IN MAXCYCLES AND CONVERGENCE CRITERION (I2. DAVIDSON RESTARTS WITH CURRENT EIGENVECTORS AS INITIAL GUESSES. COUPLED CLUSTER THEORY WITH DOUBLE SUBSTITUTIONS. FULL FOURTH ORDER PERTURBATION THEORY IN THE SPACE OF SINGLE.D18. ECISD IN CISD EUMP3 IN MP3. . COUPLED CLUSTER THEORY WITH SINGLE AND DOUBLE SUBSTITUTIONS. QCISD.13) Max N cycles. THIRD ORDER PERTURBATION THEORY. CCD. DOUBLE. MP4(SDTQ). TRIPLE AND QUADRUPLE SUBSTITUTIONS. L914: MAXIMUM NUMBER OF EXPANSION VECTORS IN DAVIDSON SCHEME 0 N 200 VECTORS N VECTORS **** NOTE: WHEN EXPANSION VECTORS EXCEED THE MAXIMUM. CI WITH ALL DOUBLE SUBSTITUTIONS. WITH THE CORRELATION ENERGY. CCSD. MP4(SDQ). MP4(DQ). ECID IN CID. DOUBLE AND QUADRUPLE SUBSTITUTIONS. MP3. FOURTH ORDER PERTURBATION THEORY IN THE SPACE SINGLE. IOp(9/7) UPDATE THE ENERGY IN COMMON/GEN/ 0 1 126 YES.Overlay 9 IOp(9/5) METHOD 0 1 2 3 4 5 6 7 8 9 CIDS. WITH EUMP3. FOURTH ORDER PERTURBATION THEORY IN THE SPACE DOUBLE AND QUADRUPLE SUBSTITUTIONS. CID. BD.

wavefunction for L914).3. default 1 in 907 and 10 in 919. L914 single point:n: 10**-4 wfn. NO IOp(9/8) L902: Constraint on output wavefunction for stability calculations (see link 902). same as 1.2 7 YES. WITH EMP4(SDQ) OR EMP4(DQ) IF SINGLES ARE NOT AVAILABLE. L914: State of interest: 0 N WE ARE NOT DOING GRADIENTS. 0 Default: L913 single point: 10**-7 energy. fast and R where possible). 10**-6 wfn. L913: Whether to use fast routines: 000 1 2 10 20 30 40 000 100 200 Default (no Slava. Default.2. 10**-8 energy. UMP43. UMP42. Original routines. UMP41R. 10**-5 wfn.3. The defaults are 22 for RCI. L913 gradient: L914 gradient: N 10**-N. follow /Orb/. FP OR CIS-MP2 WE ARE INTERESTED IN THE NTH EXCITED STATE IOp(9/9) Convergence criterion (on energy for L913. 42 for RQCI. Original code (DD1. Term and method selection for debugging in 906. DD4RQ (closed-shell). UMP42. with a full transformation). Number of roots in 907 and 919.. 11 for UCI. UMP41. UMP43.e. DD4UQ Use DD1R. 127 . and 31 for UQCI. Only active in L914. Use DD[1-3]R and UMP4xR (closed-shell). 0 No.2. 10**-6 wfn.4) for first iteration Use DD[1-3]R and UMP4xR (closed-shell) on 1st iteration Original code for 2nd and later iterations. UMp41U. Slava routines. Use DD1. IOp(9/10) Test flag in link 902 Whether to do "fake" frozen-core (i.

Force direct computation of contributions. Yes. Same as 2. Spin projection control in L913: 0 1 2 Default (1) Do basic projection. IOp(9/11) Flags for Green's function calculations: 0 1 2 00 10 100 1000 10000 Normal use of MO integrals.1 2 3 For AO usage (NYI here). and pole strength warning level on one line.QCISD(T). IOp(9/14) Non-iterative corrections: ICNonI 0 1 2 3 128 No. Read EMin. EMax. for open-shell. Force use of sort for intermediates. Yes. Read window of MOs to refine in the same format as 801. and store full /Orb/ back on disk. note number of frozen core and virtual and reset /Orb/ for full. Include triples? IOp(9/12) Test flag in l902. Fourth and fifth order singles and triples . BD(T). Normal production of intermediates (in-core if possible). . Force direct computation of contributions. Test flag in l902. but save the amplitudes. IOp(9/13) Symmetry constraint of output wavefunction from stable=opt: 0/1 yes/no. but with two ranges on the same line Force N**3 algorithm in GFSCMA. Link 909 only. Fourth-order triples (NYI).

included Z-amplitudes if necessary. suppress in-core storage. otherwise spilling algorithm in Tran4D. Functional to use in L914. Default (same as -3) Use disk storage for partially transformed integrals handling M occupieds at once. Use AO integral algorithm (L914 only). default: in-core if possible. Do gradient in L913. Force the fully in-core algorithm.4 Same as 2. Do Lagrangian in L906. IOp(9/17) Auto-adjustment of tau in L918. Try to minimize integral evaluations as for -3. Do a maximum of (-N-6) occupieds per pass. -2 -1 0 M Force a single integral evaluation (two for UMP2) using disk-based algorithm. using fully direct methods if possible. L913. L914: Control of in-core integrals for W(Tilda): -6 -3 0 1 Force in-core storage.E. using the fully out of core allgorithm. Force in-memory algorithm (fully direct MP2. IOp(9/16) L906: Control of (Semi-) Direct MP2: -N -6 -5 -4 -3 to disk. Do AO derivatives and Lagrangian in L906. IOp(9/18) ITERATION SCHEME: DE= (IN A(S)=W(S)/(DE-DELTA(S)) I. IOp(9/15) Type of derivative information generated: 0 1 2 None. IN THE FORMATION OF A NEW WAVE FUNCTION. requires 2OVN words of memory for E2. but do E4T as well. FOR METHOD = 129 . but also force use of the fully out-of-core Force a single integral evaluation as for -2. in Tran4D. (IOp(5)). but also force use of the fully out-of-core algorithm Try to minimize integral evaluations. 2N**3 words for derivatives). 0 USE DE DEPENDING ON THE METHOD USED.

IOp(9/22) Conversion factor in L919.4.GT.->DE = 0.DE = W(0)/A0. -1 0 N Read in factor in format D20. 1. Do not extrapolate.10. Use A as guess for Z. IOp(9/23) Localization of orbitals in L919. Use old extrapolation for CI. 0 1 2 3 4 5 00 10 100 Default: CI using old extrapolation. Use BFGS.3. NOTE THAT FOR PERTURBATION METHODS (METHOD=2. 10**-N. IOp(9/21) Guess for eigenvector of y-matrix in link 902. No (used in HF second derivative calculations). 130 . Default of 10**-8. IOp(9/19) EXTRAPOLATION. 0 1 Yes. ALWAYS. IOp(9/20) Whether to update the total energy with the MP2 energy in L901. 0. Use DIIS.5) DE IS NOT REALLY NEEDED SINCE THE WAVE FUNCTION FORMED NEVER GETS USED. QCISD using RLE. ALWAYS.0. FOR METHOD . Use RLE. Use scaled A as guess for Z. Reset RLE for Z iterations.1->. 1 2 W(0)/A0.

AT ONE CYCLE GIVEN BY INPUT (I3) YES. Choose configurations by simple truncation. Rettrup-Davidson RPA. Localize virtuals. Read in configurations. In-core method. AT FIRST CYCLE AND AT END IOp(9/26) NORMALIZATION OF THE WAVEFUNCTION 0 1 NORMALIZED TO A(0)=1. Correction to CIS in L914: 0 -2 -1 1 2 No CIS-DFT (in primitive energy code) CIS-MP2 (in primitive in-core program) CIS-MP2 (in MO Basis disk routine) CIS-DFT (in production code). AT EACH CYCLE YES. Out-of-core method. Jorgensen-Linderberg Hermetian RPA. Maximum order of perturbation theory in L921 and L922. SUM(S) A(S)**2 = 1 (ALL S) NOTE: PERTURBATION THEORETICAL RESULTS ARE VALID WITH NORM=0 ONLY 131 . Singlet states.0 1 2 3 00 10 20 000 100 0000 1000 00000 10000 None. The functional is given by IOp(17). IOp(9/25) PRINT PAIR CONTRIBUTION AND WEIGHT TO CORRELATION ENERGY 0 1 2 3 4 NO YES. Triplet states. Default (same as 10). AT THE END OF CI YES. Localize both. Localize occupieds.

force fully direct method by default. Scan the 'A' vector and print all coefficients having coefficients greater than 0.J. 22 otherwise). Same for DIIS extrapolation. For BFGS extrapolation. IOp(9/28) PRINTING OF DOMINANT CONFIGURATIONS.IOp(9/27) Maximum amount of disk to use in L906: -1 0 N No disk. Print all coefficients every iteration. N words. Compute the density correct to second order (NOT the same as the density corresponding to the Do not compute the density correct to second order. Do not print coefficients. Lee IOp(9/37) Maximum dimension for the QCISD extrapolation. Scan the 'A' vector and print all coefficients.default size is ten. Do not form the CI one-particle density matrix. IOp(9/36) Compute the T1 Diagnostic of T. 132 . IOp(9/30) Calculation of the one-particle density matrices: 00 1 2 10 20 Default (21 for CI.0001*N. IOp(9/31) Print vectors and matrices in 902 and 918 0/1 no/yes. The maximum dimension is 25. MP2 energy). Use as much disk as needed for a single pass. 0 -3 -2 -1 N Default (print coefficients 0.1 and above). Compute the CI one-particle density matrix.

HF.Default: energy and gradient.Converge on energy and gradient 3 . FORMAT I5 ON LAST CARD BEFORE EOF 133 . the only dimension is IOp(39) L913: Type of convergence test 0 .IOp(9/38) Minimum dimension for the QCISD extrapolation. the previous job indicates wavefunction not just expansion vectors has converged). For BFGS extrapolation. IOp(9/40) Reference wavefunction for MP2 in L906: 0 1 2 Default (HF). The maximum dimension is eight. CASSCF. L914: Pick out guesses from restart file or othogonalize guesses to the states already on restart file (IOp49 must be set to 1 or 2 for this option to be valid) 0 N -1 Just take guess from restart file Make N additional orthogonal guesses to those present Read which N states to use (free format integers) *** WARNING: The states on the restart file MUST be orthogonal to the convergence requested (ie. For DIIS extrapolation.Converge on gradient only Convergence on gradient is for extrapolated CI and QCISD procedures.Converge on energy only 2 . THRESHOLD FOR PRINTING EIGENVECTOR COMPONENTS in L914: 0 N ITHR = 1 ITHR = N WHERE THRESHOLD = GFLOAT(10)**(-ITHR) IOp(9/41) L914: NUMBER OF STATES TO SEEK WHEN USING DAVIDSON. the smallest dimension one and the default is three. 1 . OR NUMBER OF STATES TO PRINT OUT INFORMATION FOR WHEN USING DODIAG: 0 N -N DEFAULT TO 2 LOWEST N STATES READ IN PRINCIPLE COMPONENT OF N GUESSES (DAVIDSON).

a restart. MO Mapped to 3. Force Davidson not to half the number of states at iteration 2. 222.BB AB BA AA. transition densities from ground.BB AB BA AB BA METHOD -|-> FORCE DAVIDSON IN A. AO basis. DEFAULT IS: 3 (RHF REFERENCE STATE) (UHF REFERENCE STATE) MATRIX AA. IOp(9/44) Density matrix control for filling RWF 633 in L914: 0 1 2 3 states Same as 2 Do densities of each excited state Do densities and transition densities from ground Do densities.IOp(9/42) METHOD AND MATRIX BLOCKS TO WORK ON in L914 (See below) -NNN 1 2 3 0 333 BITS 1 NYI NYI 2 20 200 3 30 300 Mapped directly to NNN below.BB 10 100 AA. then don't.O. Mapped to 2. and transitions densities among all excited 134 . or 20 as appropriate. BASIS --|-> FORCE DODIAG TO FIND ALL ROOTS --|-> FORCE DAVIDSON IN M. BASIS -- IOp(9/43) How to handle subsequent Davidson Iterations in L914: 0 1 2 If this is not a restart. then half the number of states at the second iteration.O. 333. If this is Force Davidson to half the number of states at iteration 2. In-core. or 30 as appropriate.

IOp(9/45) Debug option for comparing previous results in L914. 0 1 Use Phycon to convert to eV's Use old conversion to eV's IOp(9/46) Control of Davidson Convergence in L914: <0 0 N Use Ortvec convergence only Converge on the number of roots .IOp(41) Converge on Ci Amplitudes for N lowest states IOp(9/47) Control of Davidson Iterations in L914: 0 1 2 Usual Don't do any iterations (guess=print) Stop after first iteration IOp(9/48) RESTRICTION ON TYPES OF ROOTS (DAVIDSON RHF ONLY) 0 1 2 3 GUESS ONLY SINGLETS Same as 0 GUESS BOTH SINGLETS AND TRIPLETS GUESS ONLY TRIPLETS NOTE: A SINGLET GUESS MAY RESULT IN A TRIPLET ROOT IN EXTREME CASES (SMALL NUMBER OF ROOTS SOUGHT) IOp(9/49) INITIAL GUESS VECTORS 0 1 2 3 4 MAKE A GUESS BASED ON DIAGONAL ELEMENTS USE GUESS VECTORS ALREADY ON RWF USE GUESS VECTORS ALREADY ON CHK GENERATE GUESSES FROM CIS DENSITIES on CHK GENERATE GUESSES FROM CIS DENSITIES on RWF 135 .

IOp(9/72) Whether to computed frequency-dependant polarizabilities.IOp(9/60-62) Over-ride standard values of IRadAn. Use form N (see IOp(88) in overlay 5). IOp(9/70) 1 to force TDHF in L914. 0 1 No. Yes. IRanWt. No. no for stability). if doing excited states. 0 1 2 Default (No). No. use equilibrium. IOp(9/73) Whether to do non-equilibrium solvation in L914: 0 1 2 Default (Yes. IOp(9/74) Over-ride default choice of frequency dependence of the XC functional in L914: 0 N Use default value. Yes. IOp(9/71) Whether to do an extra iteration after Davidson convergence. Yes. IOp(9/75) Whether to save amplitudes and integrals in L906: 136 . and IRanGd.

0 N Use the default. 0 1 No. save them. Population Boys+Population. i.p')).5 (assuming CBS-4 calculations. NStart. 0 -N N Default . don't save (default). Save amplitudes. Boys. No localization. intended for visualization). Do 2nd order. Yes. IOp(9/84) Save CBS localized orbitals to RWF (this will overwrite the SCF orbitals. Minimal population. Localize core even if not needed. IOp(9/85) Flags for SAC-CI 137 . Use 10**(-N) IOp(9/83) Localization Method. 6-31+G(d'. Get the lowest energy value between CBS(N) and CBS(NVirt). Calculate the extrapolated value at N only. IOp(9/81) Minimum number of Pair Natural Orbitals (PNO) to start the extrapolations from.0 1 2 Save only if doing second derivatives (SqS12 set). -1 0 1 2 3 4 5 10 100 No localization. IOp(9/82) Convergence tolerance for CBS localization. Default (4).e. Save amplitudes and integrals.

138 . Yes.IOp(9/86) Whether L906 should generate data compressed to active atoms during mp2 frequencies with ONIOM: 0 1 2 Default (2). No.

Compute F1 and S1 derivative terms here. D E(CCSD/QCISD) / D R Calc. Use old Px/Wx digestion code. Don't process any derivative terms here. D E(BD) / D R Calc. D E(CCD) / D R Calc. D E(CIS) / D R Calc. D E(CID) / D R CALC. 0 1 2 00 10 20 30 100 1000 10000 20000 No. Z-Vector method. Compute dipole derivatives using only electric field CPHF and F(x) matrices. D E(MP3) / D R Calc. Use new Px/Wx code but save both S1 and F1 over MOs. Set up for GIAO MP2 calculation.Overlay 10 IOp(10/5) CALCULATION OF FIRST DERIVATIVES OF POST-SCF ENERGIES. Do DFT 3rd derivatives. Only implemented for closed-shell and UHF. 0 1 2 3 4 5 6 7 8 9 00 10 20 30 000 100 200 NO CALC. Test Z-Vector using full CPHF. No contributions to the force constants are done here). Default: use new Px/Wx digestion code if possible. 139 . D E(CISD) / D R Calc. Yes. Use new Px/Wx code and don't save S1 but do save F1.just set up here unless doing HF 2nd derivatives simultaneously. IOp(10/6) Calculation of the second derivatives of the SCF energy. Default derivative processing .e. Available for RHF and UHF. do D2 E(SCF) / D R(I) D R(J) Setup For MP2 2nd Derivatives (i. D E(MP2) / D R CALC. D E(MP4) /D R Default CPHF usage (Z-vector unless HF D2E) Full 3*NAtoms CPHF. save as little data as possible. Do hyperpolarizabilities for second-harmonic generation. Partially coded but NYI for high-spin ROHF. Setup for external processing of W and Z.

Electric field Kth order. 1. Always use DIIS. IOp(10/11) Largest matrix for direct inversion in LinEq2.D-11 for Z-Vector CPHF. Expand each variable in a separate expansion space. IOp(10/9) Whether to compute Born-Oppenheimer corrections. IOp(10/10) Control of CPMCSCF during avoided crossing/conical intersection searches. 0 1 2 3 Default (same as 2).D-9. Yes. possibly reverting to the old (one variable at a time) method in Invert the A matrix directly. 0 -1 N Default invert directly if there is enough memory. 0 IJKL Default (1st order nuclear and electric field). Solve all equations together. IOp(10/8) Selection of linear equation solution method. The max element is tested against 10* this value. Magnetic Field Jth order. 0 N Default: 1. 140 . N.D-N. 0 1 No. the secondary solution.IOp(10/7) RMS CONVERGENCE ON C1(I. except 1. Nuclear Lth order. IOp(10/13) The nature of the perturbation(s). never invert directly.A) contributions. Nuclear magnetic moment Ith order.

which is used only for one term in polarizability derivatives and for which the accuracy requrirements are less stringent. S1. but use of electric field expansion vectors for nuclear coordinate CPHF can cause errors of up to 1 cm**-1 with current tolerances.eq.IOp(10/14) Whether to update dipole and polarizability derivatives. Don't update polarizability. Note that because of numerical instabilities in the simultaneous solution method. but leave the full version of /Orb/ on the disk. IOp(10/17) Frozen-core: 0 1 2 3 Default (use AO 2PDM for Lagrangian only if orbitals are frozen in /Orb/). and S2 off the AO 2PDM. Update dipole. Convergence is 10**(-N) for max and rms. Do C1. Save as 2. Default (Use old vectors if available). =2 Save vectors at end.0). 0 1 2 3 00 10 20 Default (=1 if IOp(8)=1 and electric field only and no derivatives are being computed. IOp(10/16) Convergence in secondary linear equations (only for simultaneous solution). Convert /Orb/ to full. C2. otherwise). Ignore old vectors. This may be acceptable in the electric field second order CPHF. Don't update dipole Update polarizability. reusing old expansion vectors for new B vectors can reduce accuracy. IOp(10/15) What to do with expansion vectors from the linear equations. 0 N Use standard machine tolerance (MDCutO) on maximum and rms. 0 1 2 10 20 100 Default (yes if IOp(5). 141 . Use old vectors if available. Pass vectors from 1st to 2nd order CPHF. but delete at end of link. Delete vectors at end of each CPHF. for debugging frozen-core with integrals over the full window. This option is normally used to pass 1st order electric field results to the second invokation of 1002 during frequency calculations. Force 2nd order cphf for polarizability derivatives.

Don't use integrals.e. 0 1 2 Default (yes). Force use of AO integrals. Yes. 0 1 2 3 CPHF is done correctly. Compute the 2e integrals when needed. Yes.e. force AO or direct.. uncoupled Hartree-Fock is used). The U-matrices are set to zero.IOp(10/18) Whether to do correct or approximate CPHF. independent of convergence criteria. Save magnetic MO deriviatives. Use option MN in control of 2e integral calculation. 0 1 2 10 Default (No). IOp(10/19) Whether overlap (S1) terms must be included. Simultaneous solution is implied. MO if possible. Disables use of symmetry to reduce the size of the CPHF problem here. even if MO ones are available. even if present. Read the 2e integral files. and full MO Fock matrix derivatives in permanent rwfs. No. Note that the appropriate rwf (588) must be present in any case. IOp(10/21) Whether to store Uai. No. Only a single set of products AX are computed. and hence the U-matrices are set equal to the B-matrices (i. IOp(10/20) How to handle 2e integral contributions: 0 1 2 3 4 MNx Default (decide on the fly). i. Spq. 142 . The A-matrix is neglected.

Force recomputation. -N 0 1 N All integrals done as Raffenetti if there are N or more matrices. less than N. Integrals with degree of contraction greater than or equal to N are done are regular integrals. All integrals are done as regular integrals. Yes. Hexadecapole. recover ints if available on rwf 610. No. Uniform electric field (dipole) only. IOp(10/31) Whether to use symmetry to reduce the number of CPHF equations: 0 1 2 Default (yes).do if possible in direct calculation. 143 . IOp(10/29) Use of rafinetti integrals during direct SCF. Octopole. all as regular if there are Default: let FoFDir decide. IOp(10/28) State for CPMCSCF: 0 N Default (ground state).IOp(10/22) Which multipole (electric field) perturbations to include? Only used if J part of IOp(13) is non-zero. IOp(10/30) In-core storage of 2e integrals: 0 1 2 Default . Quadrupole (electric field gradient. all 6 cartesian components. 0 1 2 3 4 Default. Dipole (uniform electric field). Nth excited state. Force in-core storage.

No. 0 1 2 3 No (Default) Yes No Do only optical rotational strength. Default (16 if doing magnetic CPHF). Use IGAIM method .the gauge used to calculate the angular momentum perturbed wavefunctions. 144 . IOp(10/47) Whether to do spin-spin coupling constants. IOp(10/46) Whether to calculate dipole and rotational strengths (VCD). Use single gauge origin . Yes. No. -1 0 1 2 4 8 16 None. IOp(10/45) Type of Gauge Transformations to perform to calculate the current distribution within the molecule. No. and hence the molecule's other magnetic properties.IOp(10/32) Whether to apply interchange in link 1004: 0 1 2 Default (No).the coordinates of which are read in (in Angstroms). Use single gauge origin . Yes. Use GIAOs. Whether to read D2E file in link 1003: 0 1 2 Default (No). 0 1 2 Default (No) Yes.gauge origin coincident with the nucleus of the integrated atomic regions. Use CSGT method.

0 N Default (1000). Don't compress. 145 . Update frequency-dependent property file if frequency-dep. read in frequencies. Compress. 0 1 2 3 Default (For nuclear. IOp(10/73) Maximum number of CPHF cycles. calculation is performed. unless both electric and magnetic properties are requested). Yes. Use FMM if turned on globally. but blank contributions for inactive atoms. lower cutoffs suitable only for CPHF/CPKS. No. IOp(10/60-62) Over-ride standard values of IRadAn. use more aggressive cutoffs in integrals and FMM unless doing NFx. Update regardless.IOp(10/48) Whether to operate only over perturbations involving active atoms. IRanWt. Use global cutoffs. with formalism for frequency-dependent XC response. N. Turn off FMM here regardless. Use local. compress if overlay 11 did). Do not update. IOp(10/72) Whether to do frequency-dependent properties: 0 1 2 3 4 00 10 20 Default (No. 0 1 2 10 20 Default: Use FMM if turned on globally. Yes. and IRanGd. IOp(10/63) Changing defaults. use more aggressive cutoffs in Xc integration. Yes. Don't compress.

Canonical MO derivatives.IOp(10/74) Whether to do non-equilibrium solvation. Yes. IOp(10/78) Whether to solve CPHF equations for MOD method. Solve with SimEqn. Print eigenvectors as well. 146 . IOp(10/75) Print during NMR. MOD orbital derivatives. 0 N Use global value for this job step. 0 1 2 Default (1). 0 1 2 Default (No). IOp(10/76) Over-ride general choice of exchange-correlation frequency dependence. Yes. Print tensors and eigenvalues. 0 1 2 Default: Only if frequency-dependent. IOp(10/77) Test CPHF results by checking the CPHF equations using the complete MO Fock and density derivatives. Default (20). No. No. 0 1 2 00 10 20 Default (1). Use type N (see IOp(88) in overlay 5). Solve using DiagD.

restart point NN. for testing restarts. MNN Stop at pass M (default 1). 147 .IOp(10/79) Stop the link at selected points.

Yes. CONTAINS I2*100+I1*10+I0. 0 1 2 NO. IOp(11/7) IFTPDM: WHETHER OR NOT TO CONTRACT INTEGRAL DERIVATIVES WITH A 'READ-IN' TWO-PARTICAL DENSITY-MATRIX. . I0 148 WHETHER OR NOT TO USE THE CONTENTS OF IRWFX. IOp(11/8) IFF1: WHETHER OR NOT TO COMPUTE F1 OVER AO'S. 0 1 2 NO. IOp(11/9) IDOUT: FIRST-DERIVATIVE OUTPUT OPTION.Overlay 11 IOp(11/5) IFWRT: DERIVATIVE INTEGRAL WRITE OPTION. IOp(11/6) IFHFFX: WHETHER OR NOT TO CONTRACT INTEGRAL DERIVATIVES WITH HARTREE-FOCK DENSITY MATRIX TERMS TO PRODUCE HARTREE-FOCK TWO-ELECTRON CONTRIBUTION TO THE FORCES. Yes. also contracted electric field density matrix derivatives to form the two-electron integral derivative contribution to the polarizability derivatives. but generate and write out the HF 2PDM here for debugging purposes. YES. YES. PRODUCE A D2E FILE. 0 1 DO NOT PRODUCE A D2E FILE. 0 1 No. Yes.

Lagrangian. Illegal here. IOp(11/10) Whether to compute Fock matrices. PROCESSING OF TPDM CONTRIBUTIONS. IOp(11/11) Control of integral derivative algorithm: 0 2 3 4 5 6 7 8 9 10 11 12 13 Default Illegal here. use IsAlg to decide. NONE. TAKE HF CONTRIBUTIONS FROM FX1 (A LA IFHFFX). TAKE HF CONTRIBUTIONS FROM F1 (A LA IFF1). Illegal here. FoFDir: Prism spdf. and SCF energy in L1110: 0 1 No. IF NOT THERE. Illegal here. YES. Illegal here.0 1 I1 0 1 2 3 I2 0 1 NO. Illegal here. Scalar Rys SPDF. NONE. IOp(11/12) Selection of 1PDM in L1102 and L1110: 0 N Usual SCF density. 149 . Illegal here. (forms the 1/2(F-H) term in link 1110). Illegal here. PROCESSING OF TWO-ELECTRON HARTREE-FOCK CONTRIBUTIONS. MERELY SET THE ARRAY TO ZEROES. ADD IN CONTENTS OF FX2. Use generalized density number N for both the one-electron integral derivatives and the corresponding 2PDM terms. Illegal here. Form 1/2(F+H) term in link 1110. Illegal here. Yes.

default. Force O2V2 method. Use Ix. Force old N**5 I/O algorithm. Old gOV3 I/O algorithm.0 load fxyz from rw-files if it exists. 0 IJK Default (1st order nuclear and electric field). Also compute dipole derivative integral contribution to the HF dipole derivatives.ne. Also compute HF contribution to the dipole moment. Formation of Ux*I*T terms. IOp(11/15) Controls output of derivatives to rw-files. Default Ix*T algorithm (1) Force new algorithm. Use (2g+O)V2 memory algorithm even if O2V2 memory is available. i3 .0 controls output of 'old' format. Nuclear Kth order.eq. i4*10000+i3*1000+i2*100+1i*10+i0 i0 . 150 .ne.ne. Electric field Jth order.0 forces out-of-core algorithm IOp(11/16) Mode of operation of L1102.ne.1 calculate nuclear contribution. N**4 I/O algorithm. Force old algorithm. i1 . IOp(11/14) The nature of the perturbation(s). Use L(x) and Ux*I. Formation of Fx*T*T terms: default is to choose based on available memory. Magnetic Field Ith order.0 calculate one-electron contribution. same as 1. i4 .IOp(11/13) Flags for L1112: 0 1 2 00 10 20 000 100 200 300 0000 1000 2000 Default for Ix==>Sx (same as 1). i2 . 0 1 10 Default: compute dipole derivative matrices only.

IOp(11/17)

Frozen-core in L1111: 0 1 2 3 10 Default (use AO 2PDM for Lagrangian only if orbitals are frozen in /Orb/). Do C1, C2, S1, and S2 off the AO 2PDM. Convert /Orb/ to full, for debugging frozen-core with integrals over the full window. Save as 2, but leave the full version of /Orb/ on the disk. Form the derivative integral contribution to the Lagrangian as well. This is stored on disk as RL(NBasis,NBasis,NAt3,IOpCl+1) in rwf 1001.

IOp(11/18)

Save AO 2PDM from L1111. 0 N No. Save the AO 2PDM on rwf N. It is (NTT,NTT) and includes factors (2-Delta(ij))(2-Delta(kl)). It doesn't include any normalization factor.

IOp(11/19)

Whether to delete MO integrals after 1112: 0 1 2 Default (Yes). Yes. No.

IOp(11/20)

How to handle 2e integral contributions in L1112: 0 1 2 3 Default (same as 1). Read the 2e integral files, MO if possible. Compute the 2e integrals when needed. This link must have been built with the non-dummied version of FoFDir and associated integral routines. Force use of AO integrals, even if MO ones are available. MNx Use option MN in control of 2e integral calculation.

IOp(11/21)

Size of buffers for integral derivative file. 0 N Default (Machine dependent; see DSet2E). N integer words.

151

IOp(11/22)

In-core option in 1112.

IOp(11/23)

Use of rafinetti integrals during direct term in L1112: -N 0 1 N All integrals done as Raffenetti if there are N or more matrices; all as regular if there are Default: let FoFDir decide. All integrals are done as regular integrals. Integrals with degree of contraction greater than or equal to N are done are regular integrals. less than N.

IOp(11/24)

Output of 1102: 00 1 10 Default (01). Contract with density matrix to form dipole derivative contributions. Store dipole derivative matrices on disk.

IOp(11/26)

PROGRAM ACCURACY OPTION. 0 1 DO INTEGRALS ECOMOMICALLY TO 10**(-10) ACCURACY. 'TEST' OPTION BYPASS CUTOFFS.

IOp(11/27)

INTEGRAL RETENTION PARAMETER. 0 N RETAIN INTEGRALS GE 10**(-10) IN THE D2E FILE (IF SELECTED) AND/OR 10**(-10) IN THE INTEGRAL HEAP IF IFF1=1 AND MODE=2. RETAIN INTEGRALS GE 10**(-N).

IOp(11/28)

Location or generation of MO 1 and 2 PDMs for L1111: -7 -6 -5 152 Compute QCISD 2PDM Compute CCD 2PDM Compute CIS 2PDM

-4 -3 -2 -1 0 N

Compute CISD 2PDM. Compute CID 2PDM. Compute MP2 2PDM. Compute HF DMs. Default (RWFs 626, 627, and 628). RWFS N (1PDM), N+1 (W), and N+2 (2PDM).

IOp(11/29)

What to do: 1 10 100 200 it. 1000 10000 20000 Suppress writing alpha, beta, and spin density rwfs. Form and sort the 2PDM derivatives rather than the 2PDM. Generate replicated 2PDM copies for testing. Transform 1PDM and Lagrangian from MO to AO. Transform 2PDM from MO to AO. Sort AO 2PDM into shell order. If back transformation has not been requested, the double-length AO 2PDM is expected in file 1001. The sorted 2PDM is left in file 602. Form the contribution of the 2PDM to the forces right here. Note that if the 2PDM is also to be left behind, it will be over 6d/10f and have the HGP d and f scale factors in

IOp(11/30)

What to compute using integrals or D2E file. 1 10 Energy. Gradient.

IOp(11/31)

Whether to use symmetry in Rys integral derivatives in L1110: 0 1 Yes. No.

IOp(11/32)

Whether to do 2PDM or just Lagrangian in L1111: 0 1 2 3 4 Compute Full Gradient Compute Full Gradient (Same as Default). Compute Density Only. Compute Density and W Only. Compute 2PDM only, no density or W. 153

0 1 Evaluate usual integrals. PRINT COMPUTED FIRST-DERIVATIVES. IOp(11/41) NDDO flag. Evaluate matrices in the NDDO approximation. Compute over the full list of atoms. PRINT F1 MATRICES. Force compressed form. Neglect 1e integrals with diatomic differential overlap. Neglect four center integrals. 0 1 2 154 Default: compressed. IOp(11/40) Neglect of integrals (only option 1 works in Overlay 10): 0 1 2 3 10 20 30 Keep all integrals. Compute over the full list of atoms.IOp(11/33) IPRINT IOp(33) 0 1 2 NO PRINTING. Compute over active atoms only. but blank contributions for inactive atoms. Neglect three center one-electron integrals. IOp(11/42) Compressed file formats. Neglect three center two-electron integrals as well. Neglect 2e integrals with diatomic differential overlap. . IOp(11/39) Compression of derivative matrices: 0 1 2 3 Default (2). Force expanded form. PRINT OPTION. Do only overlap and not other 1e integrals.

smallest possible number of passes. IOp(11/60-62) Over-ride standard values of IRadAn. Save gradients to disk. IOp(11/53) Convert forces over shells to field-dependent dipole and forces over atoms (for debugging): 0 1 No. 0 1 2 No. Yes. Do at least one pass. 0 1 Use global default. For Rys in L1110. Turn off FMM here regardless. Yes.3 Compressed Sx but separate H1 and F1. but using the out of core algorithms. IOp(11/45) Force NAt3 instead of NAt3+3 storage of matrices (for debugging): 0 1 No. 0 1 N Default. IOp(11/43) Batching in overlay 11. IRanWt. N is 0/1/2 for default/in-core/out-of-core. IOp(11/46) Whether to include orbital rotation gradient terms for SAC-CI. Convert 1PDM to canonical representation. needed for non-canonical methods. and IRanGd. 155 . IOp(11/63) Whether to do FMM. Do at least N passes.

IOp(11/71) Debugging option for DBF derivatives: 0 1 2 Normal processing. 0 1 2 Default (No). Ignore fitting density and just process real density in L1110. Yes. Print tensors and eigenvalues. Only works using 1C shell pairs for the density basis and only with cartesian functions. No. 156 . Copy fitting density over real density. Print eigenvectors as well.IOp(11/70) Whether to allow cavity to move in PCM derivatives. 0 1 2 Default (1). IOp(11/75) Print during NMR.

157 . No GVB2P5 trans file. IF THE ERROR TERMINATION ILSW BIT IS NOT SET). SAVE THE PERMANENT INFORMATION (MOS. recovering data from the checkpoint file and figuring out which job step to run next and whether it needs restart if an optimization or numerical frequency.e. Archive data from the checkpoint file.Overlay 9999 IOp(9999/5) CONTROLS HANDLING OF THE CHECKPOINT FILE: 0 THE RUN IS AN OPTIMIZATION OR FREQUENCY RUN. Do not write anything to the checkpoint file. if a new version was not generated in this step).E. Yes. but don't remove extra data (i. Use natural orbitals in WFN file IOp(9999/7) Controls whether MOs are written to the polyatom integral tape in LANL style. BASIS SET INFO ETC. WFN file output with GIAO magnetic orbital derivatives. Save data on the chk file. SO BOTH THE PERMANENT AND RESTART FILES ARE IN THE CHECKPOINT FILE. 0 1 No. PolyAtom output in working precision to Fortran unit 8.. IOp(9999/6) Controls output of Fortran unformatted files for other programs: 0 1 00 10 100 200 300 1000 No PolyAtom output. WFN file output WFN file output with magnetic orbital derivatives. DELETE THE RESTART INFORMATION IF THE RUN IS FINISHING NORMALLY (I. 1 2 3 4 5 THE RUN IS NOT AN OPTIMIZATION. GVB2P5 trans file to unit 14.) ON THE CHECKPOINT FILE. Restart a multi-step job.

A GAMESS input deck. formatted output to unit 7). IOp(9999/10) Controls punching of assorted information (i. 0 1 2 4 8 16 32 64 128 256 512 1024 2048 Nothing.e. except for archiving from the chk file. No. An input deck for HONDO. and isotopes during multi-step energy calculations: 0 1 2 Default (same as 1).3D20. in format suitable for guess=cards. This is independant of normal archiving to the main file.12). These are in the Z-matrix The archive entry. 0 1 2 Archive all as is. Old format. use defaults. A WFN file for PROAIMS. Archive all. Yes. Atomic numbers and coordinates in format (I3.12).IOp(9999/8) Reading temperature. pressure. Title. Use natural orbitals in WFN file. The natural orbitals generated by link 601. Both.3D20. New format. Don't archive. The molecular orbitals. but rotates to z-matrix orientation first.. 158 . orientation. IOp(9999/9) Controls archiving of dipole moment and other electic field derivatives. Output hyperfine tensors as input to Pickett's program (sent to the output file). in the standard orientation. Read a list of atoms to use in the Pickett input. Derivatives (forces and force constants) in format (2X. IOp(9999/11) Which type of database to update: 0 1 2 3 Default (3).

No. 159 . IOp(9999/18) How many virtual orbitals to include in the WFN file. No. IOp(9999/14) Whether to attempt to express the final optimized structure in terms of the input z-matrix: 0 1 2 3 Yes if there are 20 or fewer atoms.IOp(9999/12) Flag for coordinate optimization: 0 1 No. remove /ZMat/ and /ZSubst/ from the rwf and chk files. IOp(9999/13) Whether this is the end of the job step: 0 1 2 3 Default (Yes). Yes. Yes. and update rwfs. IOp(9999/15) Act as though in multi-step job type IOp(15). Go back to Link 1. Yes. IOp(9999/17) Treat as MSJDon=IOp(17) step in a multi-step job. Yes. IOp(9999/16) Treat the job as type (Info(7)) given by IOp(16).

Include N virtual orbitals. DEBUG PRINT.0 -1 N Default (None). 160 . IOp(9999/33) CONTROLS DEBUG PRINT: 0 1 NO DEBUG PRINT. Include all virtual orbitals.

Sign up to vote on this title

UsefulNot useful- Gaussian03help
- Gaussian 03 Manual
- Ansys tutorial for Design Optimization
- Actix Radioplan ACP Cookbook Atoll 3.12
- DMO Assignment 11135
- Aircraft Conceptual Design Optimization
- FSAE_RaceCar_SizeOpti
- 06104172
- Optimal Design and Operation of Batch Reactors
- Radibratovic Branislav 200905 Phd
- Biologically Inspired Optimization Methods
- Troublesome Pets
- 1-s2.0-S0025556414000510-main
- Intro Formulation
- 2 Pengenalan PL & Solusi Grafis
- 10.1.1.80.209[1]
- Index
- TN01 - Excel Solver Technical Note
- WEG Optimization of a Synchronous Generator Focusing on Biomass Cogeneration Market Sugarcane Industry Technical Article English
- Gigs 0541
- SON Inheritance ETSI
- 5 Exercise
- Welding,Design, Reliability and Optimization-Leo
- 01709202
- Crazy for Chemistry
- Chapter3 Towards Simplex
- Science+Focus+3+Coursebook+Answers
- W RFOptimizationGuide
- Eschenbach, E.a. - Multiobjective Operations of Reservoir Systems
- Topic 3.docx
- Gaussian03 Manual

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.