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

98) IOp(1/101. 103. 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 . 102. 97. 95. 96.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.

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 .

SCAN THE LST PATH.LT. L103. L109. L114) N NSTEP = N IOp(1/7) L103. L105. DISPLACEMENT . L109) = Min(40. L105. L113. L113. L114: CONVERGENCE ON THE FIRST DERIVATIVE AND ESTIMATED DISPLACEMENT FOR THE OPTIMIZATION RMS FIRST DERIVATIVE .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. L107. RMS EST. L112. IOp(1/8) L103.NVar+20) (L113. CONFV. L112. L117: Convergence on electric field/charges -1 0 N Default value for optimizations: 10**-7. Default value for single-points: 10**-5 in L116. L109. L112) = Min(20. CONVX=4*CONVF -1 0 N ConvF = 1/600 HARTREE/BOHR OR RADIAN CONVF = 0.0003 HARTREE/BOHR OR RADIAN CONVF = N*10**-6 L116. 0 NSTEP = Max(20. L105. L114: MAXIMUM NUMBER OF STEPS (OR NUMBER OF STEPS FOR AN LST SCAN).LT.NVAR+10) (L102. 10**-N. L109.NVAR+10) (L103. IOp(1/6) L102. 1 . L112: MAXIMUM STEP SIZE ALLOWED DURING OPT. 10**-7 in L117.

Default is III for Tomasi (interlocking spheres) and IV for general for the grad RHO or Vne trajectories. Apprx.).3 Bohr or Radian (L113.0 DXMAXT = 0. 1000 2000 3000 4000 N0000 10000 20000 Apprx.Do Self-Polarization and Compensation. = 0. Use only the Center of NuclearCharge Use Interlocking Spheres Order of Adam's-Bashforth-Moulton (ABM) predictor-corrector method to use in solving diff. default Yes). But No Compensation.3 Bohr or Radian (L103. L117: Whether to delete points which are too close together: 0 2 No . L121: Time step.01 * N L117: General control. Default is 5.Do-Self Polarization.1 BOHR OR RADIAN (L103. Default is 4. = 0. Apprx. Number of small steps per ABM step to be used in starting ABM and when "slow down" is needed Which approximation to make. I . IV .0001 fs. NO. = 0. Read or CalcFC). Default is Yes for minima. Yes. eqns. Default is 4 GradVne GradRho Don't Use Basins. in ABM. max is 9.2 Bohr or Radian (L105). Use MOs. Use density. II . III . L114). N DXMAXT = 0. Whether to scale or search the sphere when reducing the step size to the trust radius (Default Scale.Do III and Allow Surface To "Relax" in Solution if no spheres Whether to evaluate densities using orbitals or density matrix. Estm or UnitFC). L107: WHETHER TO MAINTAIN SYMMETRY ALONG THE SEARCH PATH. Search. scale for transition states. default 0. no for TS. Default is to use density. No. 0 1 2 00 10 20 Whether to update trust radius (DXMaxT. Which type of basin to use to partition the density isosurface.Don't Do Self-Polarization or "Compensation" Apprx. 0 1 2 3 4 N0 N00 N000 surface.1 IOp(1/9) L103: Use of Trust radius. 0 1 YES. N*0. search for minima.

Bohr. L113.1 -N 0 N Yes.NVAR) (8F10. 3 . Estimate force constants using valence force field. L117: Scaling Factor for Determining Overlaps of VDW atoms -1 0 N Turn off scaling Default is 1.J). BOMB THE JOB IF THE SECOND SECOND DERIVATIVE MATRIX HAS THE WRONG NUMBER OF NEGATIVE EIGENVALUES.010 1. L105. Second derivative matrix calculated analytically.J).000 + N*(0.I). 0 1 2 3 4 5 6 7 8 9 10 Use defaults (not valid for L109).J. 0 1 2 DEFAULT (TEST for z-matrix or cartesian TS but not for LST/QST or for minimum).6) (L103 only). Read ((FC(I.8) from input stream. using a default criteria (0. 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 6F12.001) Step size for ABM method in Trudge for isodensity method. End with a blank card. IOp(1/11) L103: TEST OF CURVATURE. using a (10**-N Angstroms) criteria.05 Angstroms) Yes. Use semiempirical force constants.F20.0**-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. (not valid for L109). Use unit matrix (default for L105. Read I. TEST.J=1. only recognized by 103). Use unit matrix throughout. followed by a blank line. How close to get to the isosurface in search. L109.0) (L103 only). Approx 1. and radians). Read from checkpoint file in internal coordinates. (5I3. DON'T TEST.0D-6 (N=20) 2. coordinates.I=1. L112.FC(I. L114: Input of initial Hessian: All values must be in atomic units (Hartree.

. D2Corr (New if energy rises. only in L103 and L115). 7 for L103 TS. BFGS (not in L103) BFGS. FMAXT = 0.0 N 0. number of bad steps to allow before attempting a linear minimization (i.e.L114. IOp(1/15) L103. Allow N -.1 * N 4 .L115: Type of Hessian Update: 0 1 2 3 4 5 6 7 8 9 Default (9 for L103 minimization. D2Corr (Old. only in L103 and L115).L109: ABORT IF DERIVATIVES TOO LARGE -1 or 0 N No force test at all. safeguarding positive definateness (not inL103 or L115) D2Corr (New. use initial Hessian).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. D2Corr (No update.05 (N=2) 0.L113. 1 for minima). otherwise BFGS). no quadratic step). Powell (not in L103).linear only starts with the N+1st. L121: Multi-time step parameter (NDtrC. D2Corr (BFGS) D2Corr (Bofill Powell+MS for transition states). 0 N Default (0 for TS. D2Corr and L115.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. Powell for L113 and L114).

0001 EIGMIN = 1.L114: MAXIMUM ALLOWABLE MAGNITUDE OF THE EIGENVALUESOF THE SECOND DERIVATIVE MATRIX. 0 N EIGMAX = 25. will occur.L113. Skip MM atoms in internal coordinate definitions and do microiterations the old way. Include MM atoms in internal coordinate definitions (no microiterations).IOp(1/16) L103. Microiterations for pure MM. Do full optimization in pruned distance matrix coords. done in L402. No optimization Do optimization in cartesian coordinates.0 HARTREE / BOHR**2 OR RADIAN**2 EIGMAX = 0. AND PROCESSING CONTINUES. SIMMILAR TO IOp(16) 0 N EIGMIN = 0. / N IOp(1/18) L103: Coordinate system.L113. Do full optimization in redundant internal coord. 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. IF THE LIMIT IS EXCEEDED. 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. Do full optimization in redundant internal coords with large molecular tools. Default (2000000). in L103.1 * N IOp(1/17) L103. Skip MM atoms in internal coordinate definitions and do microiterations the new way.L114: MINIMUM ALLOWABLE MAGNITUDE OF THE EIGENVALUES OF THE SECOND DERIVATIVE MATRIX. IOp(1/19) L103: SEARCH SELECTION 5 . in L120. Do optimization in Z-matrix coordinates. Read the AddRedundant input section for each structure. THE SIZE OF THE EIGENVALUE IS REDUCED TO THE MAXIMUM.

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

e. Normal accuracy for HF (energy and gradient both 1. round angles within 0.d-5.2 3 4 5 6 7 BACKWARD DIRECTION AND GENERATE S.6 FORWARD DIRECTION AND READ S. IOp(1/25) Wether SCRF is used with numerical polarizability: 0 1 No. gradient 1.6 IOp(1/23) L112: Derivative availability. Gradient 10**-(MM).d-4) Fine grid accuracy for DFT (Energy 1. Energy + Forces. 0 1 2 Energy only. VECTOR 8F10. VECTOR BOTH DIRECTIONS AND GENERATE S. VECTOR 8F10.d-6) IOp(1/27) = IJKL (i. No. gradient 1. 0 1 2 Default (Yes). IOp(1/26) Accuracy of function being optimized: -NNMM Energy 10**-(NN).d-7. VECTOR FORWARD DIRECTION AND READ S. Standard grid accuracy for DFT (Energy 1. 1000*I+100*J+10*K+L) 7 . VECTOR 8F10. the field in /Gen/ must be cleared each time.6 BACKWARD DIRECTION AND READ S.001 degree. Yes. -1 0 1 2 3 Read in values Default (same as 1). Yes. VECTOR 8F10.6 BOTH DIRECTIONS AND READ S.d-7). Energy + Forces + Force constants IOp(1/24) Whether to round tetrahedral angles.

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

It also suppresses error termination on large gradients. Default.3000 4000 5000 00000 10000 20000 100000 200000 Purge flags except the frozen variables. but retain the redudant internal coordinate definitions. 0 1 DON'T PUNCH. 9 . Do not retain symbolic constants. Generate a symbolic z-matrix using all Cartesians if none is present on the checkpoint file (a Same as one. (2+3) Purge all flags but keep the coordinate definition. hack to make IRCs work with Cartesian input). PUNCH. Rebuild the coordinate system. 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. 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). 0 1 YES NO IOp(1/32) TITLE CARD PUNCH CONTROL. Mark Z-matrix constants as frozen variables rather than wired-in constants. same as 10000.

0. 0 1 N>1 . free-format. L110. L110. 0. Read stepsize (up to 2 for L106. In L106: differentiate wrt nuclear coordinates. In L106: differentiate wrt electric field.005 A in L109. L111. L103. 0 N -1 -N>1 Use internal default (0.001 Angstroms in L106. L109. L109. In L110 and L111: differentiate energy N times. and L105). SUPPRESS CHECKPOINTING.0001*N (angstroms in L106. N0 000 100 200 Continuation of run. au in L111). Use step-size of 0. FIRST POINT OF A RESTART. WAVEFUNCTION. DELETE THE D2E FILE AND THE BUCKETS AND TRUNCATE THE READ/WRITE FILES. In L106: differentiate Nth derivatives once. Path step size in L115. L109. and L112 but not L102. THIS IS THE LAST POINT AT WHICH ANALYTIC SECOND DERIVATIVES WILL BE DONE. electric field au in L111). L108. L111). GET GEOMETRY.01 Angstrom in L110. In L103: Initial entry of guided optimization using N levels.0001*N atomic units everywhere. Initial entry. FROM THE CHECKPOINT FILE. IOp(1/38) Entry control option (currently only by L106. 10 .001 Use step-size of 0. L107. IOp(1/36) CHECKPOINT. IOp(1/39) Step size control for numerical differentiation. (L106. IOp(1/37) D2E CLEANUP (obsolete) 0 1 NO CLEANUP. ET. 1 for others). In L106: differentiate wrt field and nuclear. L110. 0. 0 1 NORMAL CHECKPOINT OF OPTIMIZATION.0 1 NORMAL OPTIMIZATION.

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

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

0 N Default: 50000 N. Default: 1 3/2/1 dont/QST3/QST2 optimize TS structure (for QST input). IOp(1/48) IRCMax control.IOp(1/47) Number of redundant internal coordinates to allow for. 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 . The default is MK charges. No electrostatics included in the model systems Do full square for testing. Use atomic charge type N-1 during microiterations. Default: 1 2/1 dont/do optimize product structure. IOp(1/53) L120: Action of each invocation of L120: 0 1 2 there. 3 Restore point M from data on the rwf. Default: 1 2/1 unimolecular/bimolecular reaction. 1 20 Do IRCMax Include zero-point energy. IOp(1/49) Options to IRC path relaxation (IJKL) L K J I 2/1 dont/do optimize reactant structure. Default: unimolecular IOp(1/52) L101 and L120: Type of ONIOM calculation: 0/1 2 3 00 10 20 100 N000 One layer. normal calculation Two layers Three layers Default (20) Include electrostatics in model systems using MM charges.

L115: IRC optimization. Use displacements to find the next geometry. gradients. gradient. NN = MaxLev**2 + 1 (currently 17) to restore real system. IOp(1/57) Whether to produce connectivity: 14 . hessian) of point NN of the ONIOM Next point to do is MM. 0 1 Default. Amber allowing any symbol. Save necessary information (some rwf's.4 5 6 7 NN0 MM000 Integrate energy Integrate energy and gradient Integrate energy. Yes. and hessian Restore point MM from RWF but do not create a new model system. grid. for use with parameters in input stream. IOp(1/55) L103: Options for GDIIS: ICos*1000+IChkC*100+IMix*10+Method form. energy. Dreiding/UFF. use gradients to find the next geometry. 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(1/56) Set of atom type names to parse: 0 1 2 3 Accept any. Amber.

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

2=Gradient. If IOp(67)=3. UFF. Lowest 2 digits then have no meaning. MM2 (NYI). MM3 (NYI). Use soft and hard-wired.2*NFrag -. Use only hard-wired. Use only soft. With soft parameters. corresponding to the 'classes' in FncInf. hard-wired has priority. then the default is to apply soft parameters with higher priority. 0 1 10 100 1000 10000 Do all (default) Non-bonded Stretching Bending Torsion Out-of-plane 16 . Use the first when there are equivalent matches. Use the last when there are equivalent matches. IOp(1/65) Control of which terms are included in MM. Use soft and hard-wired.IOp(1/63) Step in counterpoise calculation: MNN NN = 0 1-NFrag M = order of derivatives (1=Energy. MMFF (NYI). Quartic fitting field (NYI).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. Supermolecule Fragments with ghost atoms NFrag+1 . abort when different parameters match to the same degree. AMBER. Read modifications to parameter set. Dreiding. Do not read modifications to parameter set.

reading from input if requested by IOp(64). 0 1 2 00 10 20 000 100 200 Default (2. Generate here. Copy from chk file. Default (100) Do for atoms which have charge specified or defaulted to 0. or to use the values already there. over-written the values in AtChMM.IOp(1/66) Whether to generate QEQ charges. 1==> 221) 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. else 1. Do for all atoms regardless of initial charge. Pick up non-standard parameters from chk file. 17 . Default (10) Do for atoms which were not explicitly typed. 0 1 2 3 Default: 2 if reading geom from chk file. Yes. IOp(1/67) Source of MM parameters. IOp(1/72) L103: Algorithm choice for microiterations. Don't do QEq. Do for all atoms regardless of typing.

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). 4 if geometry read from chk) Use most abundant isotopes. Average energy (in microhartree) to be maintained during Simulation. 0 1 2 K0 L00 M000 Do not integrate.M. IOp(1/87) ONIOM integration of density. Read isotopes from chk. 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. The temperature and pressure are read first. Read isotopes from rwf. in Kelvin. Read isotopes from input. Integrate densities specified by following digits: Density to use from gridpoint 1 Density to use from gridpoint 2 etc. IOp(1/90) To read in the velocity in cartesian coordinates Nuclear Kinetic Energy Thermostat Option.2 Print eigenvectors as well. for backwards compatibility.L. K. 21 . Integrate current densities.

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

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(1/108) Gradient magnitude for DVV stopping criteria (Crit1) 0 N Default (N=15) N*0.04 (N=400) v0 is set to N*0.003 (N=30) Error=N*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.

0 N Default (1 unless specified by IOp in overlay 7 or read in). IOp(1/114) Scale factor for harmonic frequencies for use in thermochemistry and harmonic vibration-rotation analysis. 0 N Default (1 atomosphere. N/1000000. 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 . unless read in). IOp(1/113) Pressure for thermochemistry. N/1000 degrees. 0 N Default (standard temperature.IOp(1/110) Scaling of rigid fragment steps during microiterations. 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. N/1000 atmospheres.

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

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

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

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

UGBS basis set.T. He 6-311+G(2df) on Li .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. whether intended for use in expanding AOs (IOp(5)) or in expanding the density (IOp(82)).Overlay 3 IOp(3/5) TYPE OF BASIS SET.2p) on H . 29 .Ne 6-311+g(3d2f) on Na . Type selected by IOp(6).Ar CBS basis #2 -. 0 1 2 3 4 MINIMAL STO-2G TO STO-6G EXTENDED 4-31G. USE IOp(8) TO SELECT 5D/6D. rest LANL1MB. (VARIOUS IMPLEMENTATIONS MAY OMIT SECOND ROW ATOMS.) SEE IOp(6) FOR DETERMINATION OF THE NUMBER OF GAUSSIANS IN THE INNER SHELL.p) on H . EPR-III basis sets.6-31G. G3MP2large basis set.Si 6-31+G(df. G3large basis set.6-31+G(d. B-Ne 2MWB. 6 7 8 9 10 11 12 13 14 15 16 LANL ECP basis sets. IOp(6) selects options. Literature citations in SDDPot. The same numbers are used for all basis sets. Stevens/Basch/Krauss/Jasien/Cundari ECP basis sets for H-Lu.Large APNO basis set CBS basis #6 -.Be 2SDF.6-31+g(d. 5 SPLIT VALENCE N-21G (OR NN-21G) BASIS FOR FIRST OR SECOND ROW ATOMS.Q. type selected by IOp(6) (=0-4 for V{D.9p)-->(631111. Dunning/Caltech basis sets. Type selected by IOp(6) for H-Ar. use daggers if any polarization CBS basis #3 -.Core correlation basis set Dunning cc basis sets. Cl CBS basis #5 -. S. MIDI! basis sets. GENERAL--SEE ROUTINE GenBas FOR INPUT INSTRUCTIONS. EPR-II basis sets. Literature citations in CEPPot. Ahlrichs SV basis sets. 17 18 19 20 21 22 23 24 25 26 Stuttgart/Dresden ECP basis sets. Ahlrichs TZV basis sets.5. CBS basis #1 -.Ar CBS basis #4 -.p) on P.5-31G.p) on H. Coreless: Li.6-311++G(2df.6}Z) and augmented if IOp(7)=10.52111) for second row.Ne 6-311++g(3d2f) on Na . MacLean-Chandler (12s. IOp(3/6)=5 for MTsmall basis set. IOp(6) specifies type.

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

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

7.95)/3 None. Store in RWF N. Na-Ar=(0.65*I-4. Uncontract the density basis Uncontract both basis sets. Write over the dipole length integrals (518). Do not split AO S=P=D Uncontract the AO basis. 32 . FORCE 10F. shells. Split S=P AO basis shells into separate S and P shells. the values for H-Ar can be determined by Slater's rules: H=1. Massage the data in Common /B/ and Common /Mol/.IOp(3/8) SELECTION OF PURE/CARTESIAN FUNCTIONS. For VSTO-nG. Do not split S=P AO shells. FORCE 6D. He=1. 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. Add ghost atoms to /B/ so that every shell is on a separate center. D. 0 -1 N Usual place (572). FORCE 7F.325*(IA-1). 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. Read in general basis data in addition to setting up a standard basis. F. Split S=P=D= AO shells into S=P.2. Li-Ne=0. 6D/7F 5D/7F 5D 5D 5D 5D 5D/7F IOp(3/9) Where 308 should store dipole velocity integrals.

01 LP-N1G SCALE=1.00 1.80 1.74 1.33 J.PHYS.00 0.40 2SP 3SP 0.12 1.59 12.00 3SP 0.67 7.90 2.03 1.00 1.25 2.00 1.68 4.67 6.61 11.99 0.00 3SP* 1.47 16.ATOM H HE LI BE B C N O F NE NA MG AL SI P S CL A 1S 1.98 1.00 2SP* 1.48 3. 2686 (1963) OUTER SHELL HAS BEEN SELECTED ON THE BASIS INNER SHELLS ARE BEST ATOM VALUES OF NUMEROUS OPTIMIZATION STUDIES ON VARIED SMALL MOLECULES.50 1.43 17.65 9.99 1.00 1.98 0.00 1.69 2.01 1.36 4.95 2.CHEM.98 0.83 5.66 8.72 1.68 5.90 4.70 1.31 5.70 1.00 SECOND ROW ATOMS ATOM P S CL 1S 1.04 0.98 1.00 2SP 1.00 1.79 6.55 2. N-31G (ALSO N-31G* AND N-31G**) STANDARD SCALE-FACTORS HYDROGEN H 1S 1.0 FOR LI-AR (INNER AND OUTER) STANDARD POLARIZATION EXPONENTS FOR N-31G* AND N-31G** BASES 33 . 38.64 10.15 FIRST ROW ATOMS ATOM B C N O F 1S 1.26 6.02 1.75 1.56 13.15 1.75 1.00 1.00 1.10 2.20 1S* 1.53 14.24 1.88 3.50 15.05 2.69 3.00 2SP 1.

0 1 2 3 9 REGULAR INTEGRAL FORMAT IS USED. blank terminated. ZINDO/S. IOp(3/14) Addition of electrostatic integrals to core hamiltonian.2 0.multiply moments by fudge factor for charged species.1 0. ZINDO/1. 0 -1x -6 No. MG AL-CL VALUE 0. RAFFENETTI '2' INTEGRAL FORMAT. SUITABLE FOR USE WITH THE OPEN SHELL (UHF) SCF. to account for sparkles.6 0. Read coefficients of field. but not yet accepted by them. starting with electric field.39 IOp(3/11) CONTROL OF TWO-ELECTRON INTEGRAL STORAGE FORMAT.ATOM H LI BE B C-NE VALUE 1. INDO/2. CAN ONLY BE USED WITH THE CLOSED SHELL SCF. CNDO/2. IOp(3/12) Flag for semi-empirical runs. IOp(3/13) Nuclear center whose Fermi contact terms are to be added to the core hamiltonian.8 STANDARD POLARIZATION EXPONENTS FOR STO-NG* BASIS. RAFFENETTI '1' INTEGRAL FORMAT IS USED. up through 34 elements (hexadecapoles) in free format.09 0. 34 . RAFFENETTI '3' INTEGRAL FORMAT. translation vectors and d functions properly: 1 2 3 MNDO/AM1.4 0. PROCEDURES. SUITABLE FOR USE WITH OPEN SHELL RHF SCF AND THE POST-SCF USE ILSW TO DECIDE BETWEEN RAFFENETTI 1 AND 2. The magnitude is specified by IOp(3/15). ATOM NA. SCRF calculation -.

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

IOp(3/18) PRINTING OF PSEUDOPOTENTIALS 0 1 2 PRINT ONLY WHEN INPUT IS FROM CARDS or if GFPrint was specified. 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.SHF Dresden/Stuttgart potentials .MWB (second set) Dresden/Stuttgart potentials .MDF Dresden/Stuttgart potentials . 36 . 0 N Default (Machine dependant. IOp(3/20) Size of buffers for integral file.MWB (third set) Pseudopotentials for all coreless basis.CHF).MHF (second set) Dresden/Stuttgart potentials . 55296 words on Cray).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. N integer words. 16384 integer words on VAX. No longer used.MWB (first set) Dresden/Stuttgart potentials . N integer words. Alternative potentials for coreless basis. 0 N Default (3200 integer words).12 13 14 15 16 17 18 19 20 Dresden/Stuttgart potentials .

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

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

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

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

12). followed by ((HDiag(J. Use STO-3G scale factors. 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. Default (unit overlap matrix). 1400000 Regular SCF with separate K. Read from Chk. Use the unit matrix for the overlap. 1500000 J as usual but NDDO for K. Default parameters (same as 5). Use Slater's rules scale factors. followed by ((Beta(J.J=1. for testing. 1100000 Do Harris functional scaling atomic densities for current charge and multiplicity.18) (Format 3D20. Do ZINDO/1. Read parameters from chk.I=1. Same as 1.3). Read parameters for atomic numbers 1-18 in the order Scale (D20. 1300000 Second-order XC (NYI).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.I=1. Original INDO/2 Beta and HDiag Parameters. Do ZINDO/S.I). 41 . 1000000 Do Harris functional. Do INDO/2.3). GNDDO/1 parametrization. Use the real overlap matrix.J=1. Geometric mean in NDDO.12). Read the integrals sequentially. or read-in parameters). 1200000 First-order XC. geometric mean for NDDO/1 Arithmetic mean in NDDO. Load all the integrals into memory. Do CNDO/2.18) Read parameters from rwf. IOp(3/42) How to form NDDO core hamiltonian in L317: 0 1 2 Default (same as 1).I).

Do not transform 2e integrals.no (ab|xx) and no NDDO on 2e and V ints only -. the perturbation is computed separately and stored in the third and fourth matrices in the core Hamiltonian rwf. no. Force units of Bohr for coordinates. IOp(3/46) Whether to abort the job if badbas detects an error: 0 1 2 Default (yes). Order of multipoles in SCRF for L303. yes IOp(3/47) Flags for use in PRISM and CalDFT throughout the program. . 0 1 2 3 4 5 6 keep all integrals. IOp(3/44) integral rejection using L318.T and S unchanged.0) integrals. Force use of only the OS path for all calculations. If negative. IOp(3/45) Transformation matrix in L318. 0 1 2 use S**-1/2.5 10 20 Read charges and DBFs from input stream in standard orientation Force units of Angstroms for coordinates. just orthogonalize functions on the same center. neglect four center transformed integrals. -1 Bit flags: 0 1 42 If bit 0 is set (use AllowP array) then read in a list of allowed paths. NDDO approximation -. Use unit matrix (for debugging). neglect four center and 3 center (ab|ac) integrals. only 1e. neglect four center and three center (0. Use expanded matrix logic for PBC exact exchange.

Trace input and output using Linda/subprocess. Forbid use of gather/scatter digestion even for small numbers of density matrices. currently only affects CalDFT. 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). Force square loops. currently only in PrismC. Make all atoms large in XC quadrature. Force use of FoFCou.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. Turn on use of precomputed XC weights. matrices only on Fujitsu and NEC. The default is diagonal only on vector machines. Do allocation for parallel XC but run sequentially. Force all near field in FMM. even if not doing FMM. replicated Fock matrices. Do allocation for parallel 2e integrals but run sequentially. Default is to use replicated Turn off Schwartz during FMM/NFx calculations. Turn off MP-based cutoffs in FF part of NFx. Do not symmetry reduce grid points on unique atoms. Make all shells large in XC quadrature. Turn off use of Sqrt(P) in density-based cutoffs. Reserved for more control of scatter/gather. Reverse normal choice of Scat20 vs. Reverse normal choice of diagonal/canonical sampling in Prism and PrmRaf. Forbid use of FoFCou. Force single matrix code in CPKS. No longer used. Turn off vFMM. 43 . Use Mura radial grid instead of Euler-2 grid. Do nuclear contribution in FoFCou even for non-PBC Do not use special Coulomb algorithm in FoFCou. Skip consistency checks for XC quadrature Do not do extra work to use cutoffs better. Turn on angular offsets in XC grid generation.

Convert to sparse storage under FoFCou for testing. Set by PsmSet to indicate whether the NAtoms test for defaulting FMM was passed. UseUAB. IOp(3/52) Turn off normal evaluation of ECP integrals. 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). Do not default to FMM. if 128 set. Do not save as many multipole expansions as possible in memory.0 Bohr). Old routines will be used. Force FMM on. Turn on FMM print. Apply symmetry to derivative distributions (NYI). ECP integrals are evaulated in L302. Set up for parallel FMM but run loops sequentially. Uncontract all shell pairs.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. Split primitives for better boxification. Default box length. IOp(3/53) Accuracy in ECP integral evaluation: 0 -1 N Default. Default UseUAB/Use 256. 0 1 Default: if needed. 44 . No Cutoffs 10**-N. so L302 does not do ECP ints. Box length MMMM/1000 Bohr. based on geometry (but minimum for molecules 3.

of core electrons for Stuttgart/Dresden ECP's. 10**(-N). cutoff 10**(-N) IOp(3/56) Cutoff for intermediate matrices during sparse operations: 0 N 100 times smaller than storage cutoff. default accuracy (10**-10). crude accuracy (5x10**-5) Yes. cutoff 5 x 10 ** (N+100) Yes. IOp(3/59) Threshold for throwing avay eigenvectors of S: 0 N Default (10**-4) 10**-N. IOp(3/58) Cholesky control options. No Yes. intermediate accuracy (5x10**-7) Yes. 45 . IOp(3/57) No.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.

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

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

mPW91PW91. Perdew 81 correlation. Perdew 81 + Perdew 86 correlation. Becke3 with Perdew 91 correlation. HCTH147. mPW1PBE. B97-2. Becke "Half and Half": 0. Default. same as 100. HCTH407. VWN 80 (LSD) correlation VWN 80 (LSD) + Perdew 86 correlation . Becke "Half and Half" with LYP/VWN correlation. B97-1. BA1PBE. mPW1LYP. LG1LYP. HCTH93. B98. No correlation. Becke 3 with Perdew 86 correlation.IOp(3/72) Solvent type flag (for SCRF setup). Becke3 using VWN/LYP for correlation.5 LSD Do only coulomb part. IOp(3/73) ONIOM system flag (for SCRF setup). 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. B1B95. Lee-Yang-Parr correlation.5 HF + 0. skip exchange-correlation. PBE1PBE mPW3PBE. B1LYP. BA3PBE. PBE3PBE. Vosko-Wilk-Nusair method 5 correlation.

Becke "Half and Half" 0. same as B1B95. respectively. B97-2 coefficients. B98 coefficients. 200 is Hartree-Fock-Slater.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. -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 O3LYP coefficients. Coefficients of 0. IOp(3/75) Number of radial and angular points in numerical integration for DFT: 0 IIIJJJ Use a special grid designed for efficiency (default).0 and 1. and 402 is BLYP. LG exchange. B97-1 coefficients. Note that Becke actually used Perdew correlation rather than LYP. mPW91PW91 coefficients. Becke 3 coefficients: aLSD + (1-a)HF + b(dBx) + VWN + c(LYP-VWN). HCTH coefficints.7) Becke 1988 exchange.8 b=0.0 for DFT and HF. JJJ angular points.5 HF + 0.81. III radial points. 205 is Local Spin Density. X-alpha exchange (alpha= 0.72 c=0. IOp(3/76) Mixing of HF and DFT. 49 .5 Xc + Corr Coefficients of 0 and 0 (no exchange). PW91 exchange Gill 96 exchange PW86 exchange mPW exchange PBE exchange BA exchange VSXC exchange So 100 is Hartree-Fock. with a=0. Hartree-Fock-Slater exchange (Alpha = 2/3).

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

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

Iterative. Use fits.. Use fits. 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. Do not split S=P density shells. forming Z = modified A^-1. IOp(3/88) Modification of internally stored density basis.. Solve iteratively with no preconditioning Solve iteratively with diagonal preconditioning. This is also done if requested Split S=P density basis shells into separate S and P shells. Split S=P=D=. IOp(3/89) Set up for density fitting. Use fits. 1 otherwise).. in IOp(3/10). Do not use density fits. Form A' over neutral distributions via multiplies by A. IOp(3/87) Discard density basis functions based on angular momentum: 0 N No. Massage the data in Common /B/ and Common /Mol/. Solve iteratively with symmetric block-diagonal preconditioning. .IOp(3/86) Discard basis functions based on angular momentum: 0 N No. solving iterative with direct products. 0 1 10 100 1000 2000 10000 20000 None. F. 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. D. with A formed to generate preconditioning. Discard density basis functions with angular momentum >= N. solving iterative with stored A. no formation of A.. Discard basis functions with angular momentum >= N. . shells. density shells into S=P. Form A' over neutral distributions via direct products. Do not split density S=P=D. Form inverse matrix once...

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

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

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

56 .2 3 Do not reuse. to compensate for the loss of some symmetry with FMM. Read from chk file. 0 N Default (60) N atoms for the C1 case. 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.

Read intermediate SCF results which are on the checkpoint file.2. in which case Read guess from the checkpoint file. Renormalize and orthogonalize intermediate SCF results which are on the RWF.3. Read generalized density specified by IOp(38) from the chk file & generate natural orbitals from Huckel is used.Overlay 4 IOp(4/5) Type of guess: 0 1 2 3 4 5 6 7 8 it. 57 .3. Re-use Fock matrices instead of orbitals. Read generalized density specified by IOp(38) from the rwf file & generate natural orbitals from Default. Check MOs for othornormality. Projected ZDO guess.4 above. chosen via IOp(11). IOp(4/6) FORCED PROJECTION WHEN GUESS IS READ FROM CARDS (401). Default (1 for PBC without alter. Note that variable IGuess here has 4.2. otherwise 2). Suppress orthogonalization. Re-use orbitals not Fock matrices. EVEN WHEN BASES ARE IDENTICAL. Guess from model Hamiltonian. on the checkpoint file. 9 it. This uses the Harris functional unless atoms heavier than Xe are present. Default orthogonalization (perform) Schmidt orthogonalize guess orbitals. Don't check MOs for othornormality.1 corresponding to 1. 0 1 2 00 10 20 000 100 200 FORCE PROJECTED GUESS.5 * V. IGuess values of 10-14 are generated internally and are the sparse versions of 0 and 5-8. Default MO checking (yes). Renormalize and orthogonalize the coefficients which are currently on the read-write files. SUPPRESS PROJECTION. EVEN WHEN BASES ARE IDENTICAL. FORCE PROJECTED GUESS. Huckel guess (only valid for NDDO-type methods). 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.

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.

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

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

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

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

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

r**-2 Coulomb. Turn on non-bonded terms. 0 1 66 Default (211 if UFF. No variable thresholds.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 5x10-5. IOp(4/67) Whether to use QEq to assign MM charges. Yes. 1==> 221) Do QEq. 0 1 N N N<-100 Default: Yes. Eps = N / 1000.0. 0 N Default (no cutoff). r**-1 Coulomb. 0 N Eps = 1. IOp(4/65) Tighten the zero thresholds as the SCF calculation proceeds. 2 otherwise. Turn on inversions/improper torsions Turn on torsions. IOp(4/64) Cutoff for MM non-bonded term. 10**-N. initial threshold 10**(-N) Yes. . Turn on angle bending. Turn on bond stretches. Turn on all terms. initial threshold 5 x 10 ** (N+100) IOp(4/66) Dielectric constant to be used in MM calculations. 10**-N.

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

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

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

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.

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

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

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

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

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

1 2 3 4 5 10 20 30 40 1xx Nxxx Mxxxx No. Optimize only 2nd and higher naturals. Freeze all closed. with "FON" version of Fermi broadening. 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. IOp(5/25) Rotation application (L506). IOp(5/24) Orbital freezing (L506). Yes. Orbital mixing control in L506. with Fermi broadening as well. 0 1 2 Read from input stream. Yes. deciding on the fly between the two forms. weight of energy DIIS in method 4.2) . Yes. mixture of energy and Use energy DIIS when commutator gives huge coefficients. high spin and first natural orbitals. Read from rwf. Yes. 3 78 3rd root of CAS(2. Apply rotations sequentially. so that pair and hamiltonian information can be reused (L506. otherwise. with "pFON" version of Fermi broadening. L509). 0 1 Default (exponentiate rotation angles). 0 1 Optimize all orbitals. IOp(5/26) Type of calculation (L504). used (DIIS error/10^-N) for Use print level M in DIIS. commutator. Read from chk. IOp(5/23) Flag for later points of an optimization. Switch from energy to commutator when error is 10^(-N) in method 3.

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

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

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

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

If negative reads the vector from rwf. If positive reads vector from input 4f20. IOp(5/47) In L510. 10 Include the CSF contribution to the orbs for the DerCpl.7: the displacement in geometry in internals in Angs. 1 to set up for CAS-MP2 or 2 to do spin-orbit calculation. 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. 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.8 f10.

Speeds up convergence in CP-MCSCF Do not Canonicalize (turn this on to maintain compatability with previous versions of code. Square storage (only in Fock formation if CG). 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. also none.N Control flag for CalDFT is N. Diagonalization Conjugate gradient. L510: configuration cutoff for mp2 0 i 84 . 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. or 22 if sparse is set in ILSW). 0 1 Yes canonicalize. N words. 0 1 2 10 20 Default (11.1 Float(1/i) . Linear storage (only in Fock formation if diagonalization).) IOp(5/52) Amount of memory to allocate to stashing integrals. -1 0 N None Default.

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

instead of a State Averaged second derivatives calculation. only used with IOp(97)=11 IOp(5/60-62) Override standard values of IRadAn.8 0.0-1. 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.8 0.3/0.2/0.5 crash new SA 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. and IRanGd. IOp(5/63) Whether to do FMM.0/0. IOp(5/64) Over-ride default value of FMFlags 0 N No.6 0.4/0.1 Egap <limit <limit >limit >limit >limit Note: trajectory option set in overlay 10: IOp(10/99)=n -.7 0.3/0.0-1.0001 (checks Offdiagonal element).2/0. Turn off FMM here regardless. 1: 2: Normal computation when (0.a region of (0. use N. IRanWt.5/0.8 0. 0 1 Use global default.Threshold for a determining an adiabatic hop threshold=n*0. 86 .9 0. for a hypothetical example with nroot=2: cycle 4 5 6 7 8 old SA 0. regardless of weather that continues to decrease after the switch point.7 0.0).1/0.2/0. Yes.

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

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

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

separate Coulomb and NFx exchange for PBC). Typically 10^-10 for molecules and 10^-12 for periodic systems. set in FoFDir/FoFCou/CalDSu based on accuracy part of IOp(5). Retain 3 and 4 point extrap. IOp(5/92) Whether to use FoFDir or FoFCou for exact exchange: 0 90 Default: normal processing based on FMM for non-PBC. if DIIS is on. IOp(5/90) Whether to generate sparse guess here. do preliminary AM1 calculation and compare with guess from previous step in geometry optimization. Yes. . this and ZFormV should be moved to L302. 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.200000 off. IOp(5/88) L510: controls the amount of printing. The energy is only checked after FON has been turned IOp(5/87) Accuracy criterion in Fock matrix formation: 0 N Default. 10**-N. do preliminary AM1 calculation. IOp(5/91) Control option for chebyshev sparse control. 1 2 Yes.

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: .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). separate FoFDir for exchange. threshold for a hop up (after a hop down) = 0. threshold =0.25 in GS) and if Zyy=125. Value = xxxzyy.75 in ES (or 0.01) or the lower state (=(1-x)*0. value=xxx. where: Zyy yy Z xxx Threshold for a hop down determined by probability of being on the upper state (=x*0.1 fempto seconds If IOp(97)=23. separate FoFCou/NFx for exchange. N iterations.1 2 FoFCou for Coulomb. FoFCou for Coulomb. L510: has different meaning depending on if you are using IOp(97)=22 or IOp(97=23) If IOp(97)=22. wait time is xxx*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. variable=IWait.

The CAS active space is subdivided into three RAS active subspaces. 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. Ras1. In the reference space.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. RAS control in L510. default xxx=4 ie deltaE<0. Ras1 orbitals are doubly occupied. Ras2 and Ras3. and for pass 0 grid in L502. default yyy=5 ie deltaE>0. 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. 0 N Use default. Use grid N.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 L509.01 (checks energy gap). where ww 92 Number of Ras1 orbitals . Ras3 orbitals are empty.04 switch timdep on. variable=mHTest IOp(5/94) PCM/ONIOM calculation 0 1 2 Standard PCM calculation PCM/ONIOM calcn. on the real system PCM/ONIOM calcn. and the maximum number of electrons in the Ras3 space: zzyyxxww.

Yes. 0 Default: 32. Needs option 80 and option 99.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. if available). 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.2. for detailed control of the propagation conditions see options 92-96. 22 23 Propagation of the wavefunction. 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). No. Yes.12.3d IOp(5/100) Tight convergence during CGDMS: 0 1 2 Default (No). 0 1 2 10 11 21 Default (Yes.8 for 1. Hopping threshold during trajectories with L510. 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. IOp(5/99) Grid for numerical k-integration in FT-LT method. hop is based on the probability of being in a specific state Propagation of the wavefunction but no hopping. 93 . Yes. 0 1 2 Default (No). No.

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 . N. IOp(5/102) Maximum number of configurations for CAS-MP2: 0 N Default (1000).IOp(5/101) SDif test on numerical accuracy of PBC diagonalization.

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

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

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

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

This precludes calculations of atomic property derivatives with respect to nuclear displacements.s. 10. 99 ... using alpha and beta separately for UHF. L605. Use spin density. default 6). 0: Print the atomic occupancies of localized (spin)orbitals (default). IOp(6/31) Whether to determine natural orbitals from densities: 0 1 2 3 4 5 No. Store only alpha NOs.. 0 1 Make this a scratch file. Store only beta NOs.. Transition denstiy between state IOp(29) and state N. Yes. Localized (spin)orbitals with atomic occupancies less than 0. attractor interaction lines. ring points.**(-ITlLoc) is the convergence criterion for (spin)orbital localization (1. Compute energies of electrostatic interactions between AIMs (IDoPot). Compute atomic overlap matrices (IDoAOM).. Yes. IOp(6/32) CONTROL PARAMETERS FOR COVBON in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 10000*MItLoc+1000*ITlLoc+100*IDcInt+IPrLoc. L606: naming of RPAC interface file. and cage points. Compute charges of AIMs (IDoAtC). 1: Do not print the atomic occupancies.IOp(6/30) 2nd excited state for transition density: 0 N Transition denstiy between state IOp(29) and g.. using total density.9.9.11' IOp(6/35) WHAT TO DO: 0 1 2 4 10 100 Determine attractors.99.01*IDcInt are interpreted as lone pair MOs rather than bond MOs (1. Compute kinetic energies and multipole moments of AIMs (IDoPrp). Name this file 'rpac. default 9). where MItLoc ITlLoc IDcInt IPrLoc MItLoc*NOrb*(NOrb-1)/2 is the maximum number of iterations in localization of (spin)orbitals (1. default 10). Determine zero-flux surfaces (IDoZrF).

.200 400 1000 Compute other atomic matrix elements (IDoAMa).. Maximum value of the control sum (1. default 2). Use numerical instead of analtyic integration and use reduced cutoffs. RNGPNT. where MxBpIt SBpMax NGrd LookUp Maximum number of iterations in trial path determination (1. No other properties can be calculated.999. IOp(6/36) CONTROL PARAMETERS FOR NEGLECT OF ORBITALS AND PRIMITIVES in L609: 10000*INoZer+100*IPrNDe+IPrNAt. AND CAGPNT in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 1000000*MxBpIt+100000*SBpMax+1000*NGrd+LookUp. Include zero-flux surface relaxation terms in all atomic matrix elements (IDoSRe) Compute derivatives of atomic properties with respect to electric field (IDoSeP).. and IEpsIn=100. default 7). Compute atomic orbitals in molecule (IDoAOs). (IHwAug=1).99.9. augment valence electron densities with nonrelativistic core contributions Abort if pseudopotentials have been used (IHwAug=3). augment valence electron densities with relativistic core contributions.. 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).. default 20). If necessary. IOp(6/38) CONTROL PARAMETERS FOR ZRFLUX AND OIGAPI in L609 (NOT TO BE CHANGED UNDER MOST CIRCUMSTANCES): 100 . which is a default anyway (IHwAug=0). Use numerical instead of analtyic integration.**(-IPrNAt) in integrations over atomic basins (099. where INoZer IPrNDe IPrNAt 0: Ignore (spin)orbitals with zero occupancies (default). zero occupancies. If necessary.. IOp(6/37) CONTROL PARAMETERS FOR ATINLI. Neglect primitive contributions below 10..99. Note that analytical force constants have to be available! Compute localized orbitals and bond orders (IDoLoc). default 100). default 10). Number of grid points in critical point search (1. This option sets IPrNDe=5. 1: Do not ignore (spin)orbitals with Neglect primitive contributions below 10. 99. Length of Fourier expansion for the trial path (1. IPrNAt=5..**(-IPrNDe) in evaluations of electron density and its derivatives (0 default 7). Reduce accuracy so atomic charges can be computed more rapidly (IQuick).

3 Read the deletion energy produced by a previous run with IOp(40)=2 and print it. This link must have been invoked with IOp(40) = 0 or 1 prior to invoking it with IOp(40)=2... IEpsSf IEpsIn INTrig IEpsSf is the safety factor used for patches with surface faults in the adaptive integration subroutine (1... 10... default 2). 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).9.. 10. default 2). 10*INTrig is the number of sine and cosine functions in the trial function for surface sheets 10. default 2).don't read input. 10*INInGr is the initial number of grid points in theta and phi in the adaptive integration subroutine (1... default 2).9... default 6). 101 . Read input data to control NBO analysis. default 5).9.. Delete selected elements of NBO Fock matrix and form a new density.. N layers.9. IOp(6/41) Number of layers in esp charge fit.**(-5-IToler) is the tolerance for the intersection search (1.9. 0 1 2 Default NBO analysis .0001*IEpsIn is the target for integration error (1..9. default 2).99...*IPraIn is the cut-off for zero-flux surfaces (1. (1.. must be >=4..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. IRScal is the scaling factor in the nonlinear transformation used in the intersection search 10..**(-IGueDi) is the initial displacement from the critical point in the Runge-Kutta integrations (1. default 6)..9.. 0.. default (1.*IRtFSe is the safety factor used in the intersection search (1.. 0 N Default (4).9. whose energy can then be computed by one of the SCF links. IOp(6/40) Control of link 607.9. default 2). default 5).. 5+INInCh is the initial number of sampling points in the intersection search (1.9. IHowFa is the maximum distance in the Runge-Kutta integrations along gradient paths (1. default 2).9.

IOp(6/42) Density of points per unit area in esp fit. 0 N Default (1).10**-3 Read from input. same as 2. IOp(6/46) Threshold for molecular volume integration. 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(6/43) Increment between layers in MK charge fit.01*N.for tight accuracy. 0 N Default (0. IOp(6/44) Type of calculation in L604: 0 1 2 3 10 Default. points per unit area. 102 . 50 is recommended. 0 -1 N Default .4/Sqrt(#layers)) 0. N*10**-4. 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(6/47) Scale factor to apply to van der Waals radii for the box size during volume integration: 0 N Default. on non-SCF densities. Fine grid. Yes. IOp(6/48) Use of cutoffs 0 N Default (10**-6 accuracy for cubes. N*0. Yes. on non-SCF densities. 12 points/Bohr. all possible IPs and EAs.01 . 3 points/Bohr. 6 points/Bohr. N. IOp(6/52) Number of radial integration points in L609: 0 N Default (100). -N>4 Grid using 1000 / N points/Bohr. Coarse grid.for debugging. up to N IPs and EAs. 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. 103 . 1 digit better than desired acuracy for volumes). Medium grid. IOp(6/51) Whether to apply Extended Koopman's Theorem (EKT): 0 N -1 -2 Default (No). No.

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

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

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

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

108 .0 1 tensors. 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. 2 Default.

0/1 yes/no. 3=CBSQ=0. Default scale factor is #N (1=HF. 0 1 2 10 20 Default (12). with most common isotopes.91671. Print normal mode displacements in redunant internals. Read hindered rotor parameters from input. Increment the 1e derivative matrices with ecp terms. only harmonic thermochemistry. Do not form 1e derivative matrices. 109 . 2=CBS4=0. If M=3.91844) If M=1. If M=2.12. Print higher precision normal modes. No. Yes. IOp(7/9) Whether to rotate derivatives back to the z-matrix orientation. but neglect first derivatives in conversion of second derivatives to internal coordinates. IOp(7/8) Harmonic frequency calculation: 0 1 2 3 10 20 30 Nxx Mxxx No. IOp(7/7) USE OF INTERNAL COORDINATES. 0 1 2 YES NO Yes. with read-in isotopes. 1/1. Print both HP modes and internal displacements. Do not the ecp contribution to the SCF forces. Yes.Overlay 7 IOp(7/6) operation of link 705 (NYI). Form the ecp contribution to the SCF forces. do hindered rotor analysis.

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

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

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

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

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

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

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

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

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

Calculate 2 J . IOp(8/13) Control of output. IOp(8/15) Control of DRT output. if there is only 1 electron or 1 virtual spin-orbital. Also.0 or the smallest difference between occupied and virtual orbital energies is less than 0.g. IOp(8/14) Control of drt input. kill a frozen-core job if there there is significant core-valence mixing in the canonical orbitals 00 10 20 Default. Suppress such a test (CPHF may still be done for such a case). 119 . e.001. This is done when IOpCl=0 or when IOpCl=1 and the calculation is rohf or gvb. same as 10. IOp(8/11) MO coefficient. and number of electrons test. 0 1 2 Default. Kill the job if any mo coefficients are greater than 1000. 0 1 Output to Gaussian system buckets. Used to select output mode.K over deleted orbitals and add to Core-Hamiltonian. Use overlap with atomic core orbitals from Core Ham to choose core orbitals. Read 'old-style' drt input cards. 0 1 No. Kill the job if there is no correlation energy. Use overlap with atomic core orbitals from Harris to choose core orbitals. orbital energy.. 0 1 Take necessary input from Gaussian data structures.10x 20x 30x Use orbital energies to choose core orbitals. IOp(8/12) Calculation of frozen-Core contributions. same as 2 Just print a warning message. Output transformed integrals for DRT-CI calculation.

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

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

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

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

Ionization potential. (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. IOp(8/44) K-point specification for MP2 band correction. Yes. 0 1 post-SCF.1 2 00 10 000 100 0000 1000 Yes. Electron affinity. N -N Use N points for MP2. 0 N Use the k-point for which the hoco is highest and luco is lowest. Use N points and set up for band gap correction. IOp(8/45) Type of quasiparticle job: 0 -1 1 -N N Band gap.MOs are canonical HF orbitals. Default . Input orbitals are not canonical HF and pseudocanonical orbitals must be generated here for the 124 . 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. K-point at which to apply correction.

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

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

NO IOp(9/8) L902: Constraint on output wavefunction for stability calculations (see link 902). Term and method selection for debugging in 906. 11 for UCI. UMP42. fast and R where possible). Original routines. IOp(9/10) Test flag in link 902 Whether to do "fake" frozen-core (i. Use DD[1-3]R and UMP4xR (closed-shell). UMP42. UMP41. DD4RQ (closed-shell). UMP41R.2. L913: Whether to use fast routines: 000 1 2 10 20 30 40 000 100 200 Default (no Slava. UMP43. 42 for RQCI. default 1 in 907 and 10 in 919.3.2. Number of roots in 907 and 919.4) for first iteration Use DD[1-3]R and UMP4xR (closed-shell) on 1st iteration Original code for 2nd and later iterations. with a full transformation). and 31 for UQCI. Original code (DD1. FP OR CIS-MP2 WE ARE INTERESTED IN THE NTH EXCITED STATE IOp(9/9) Convergence criterion (on energy for L913. 10**-6 wfn. The defaults are 22 for RCI. Only active in L914. Default. 10**-8 energy.. L914 single point:n: 10**-4 wfn.e. 127 . Use DD1. Slava routines. UMp41U. follow /Orb/. L913 gradient: L914 gradient: N 10**-N. 0 No. wavefunction for L914). same as 1. DD4UQ Use DD1R. L914: State of interest: 0 N WE ARE NOT DOING GRADIENTS. UMP43. 10**-5 wfn. WITH EMP4(SDQ) OR EMP4(DQ) IF SINGLES ARE NOT AVAILABLE.3. 0 Default: L913 single point: 10**-7 energy.2 7 YES. 10**-6 wfn.

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

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

0 1 2 3 4 5 00 10 100 Default: CI using old extrapolation. 0. 0 1 Yes. QCISD using RLE. IOp(9/20) Whether to update the total energy with the MP2 energy in L901. IOp(9/22) Conversion factor in L919.0. IOp(9/21) Guess for eigenvector of y-matrix in link 902. FOR METHOD . Use scaled A as guess for Z. Use RLE. 1. -1 0 N Read in factor in format D20. NOTE THAT FOR PERTURBATION METHODS (METHOD=2.4. IOp(9/19) EXTRAPOLATION.DE = W(0)/A0. Use BFGS. IOp(9/23) Localization of orbitals in L919. No (used in HF second derivative calculations). Use old extrapolation for CI. ALWAYS.5) DE IS NOT REALLY NEEDED SINCE THE WAVE FUNCTION FORMED NEVER GETS USED.3. Default of 10**-8. 130 . Do not extrapolate. 1 2 W(0)/A0.10.GT.1->.->DE = 0. Use A as guess for Z. 10**-N. ALWAYS. Use DIIS. Reset RLE for Z iterations.

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

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

Default: energy and gradient.Converge on energy and gradient 3 . HF. the only dimension is IOp(39) L913: Type of convergence test 0 . CASSCF.Converge on energy only 2 .IOp(9/38) Minimum dimension for the QCISD extrapolation. 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. IOp(9/40) Reference wavefunction for MP2 in L906: 0 1 2 Default (HF). 1 . 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 BFGS extrapolation. 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). The maximum dimension is eight. the smallest dimension one and the default is three.Converge on gradient only Convergence on gradient is for extrapolated CI and QCISD procedures. FORMAT I5 ON LAST CARD BEFORE EOF 133 . For DIIS extrapolation. the previous job indicates wavefunction not just expansion vectors has converged).

a restart. then half the number of states at the second iteration.BB AB BA AA. then don't. 333. transition densities from ground. DEFAULT IS: 3 (RHF REFERENCE STATE) (UHF REFERENCE STATE) MATRIX AA. 222. AO basis. 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. In-core.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. 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. MO Mapped to 3. Force Davidson not to half the number of states at iteration 2. If this is Force Davidson to half the number of states at iteration 2.BB 10 100 AA.BB AB BA AB BA METHOD -|-> FORCE DAVIDSON IN A. and transitions densities among all excited 134 . or 20 as appropriate.O. or 30 as appropriate. Mapped to 2.O.

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(9/45) 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 .

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

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

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

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

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

IOp(10/17) Frozen-core: 0 1 2 3 Default (use AO 2PDM for Lagrangian only if orbitals are frozen in /Orb/). This may be acceptable in the electric field second order CPHF. reusing old expansion vectors for new B vectors can reduce accuracy. for debugging frozen-core with integrals over the full window. Ignore old vectors. This option is normally used to pass 1st order electric field results to the second invokation of 1002 during frequency calculations. 141 . C2. Delete vectors at end of each CPHF. Do C1. Pass vectors from 1st to 2nd order CPHF. 0 1 2 10 20 100 Default (yes if IOp(5). but use of electric field expansion vectors for nuclear coordinate CPHF can cause errors of up to 1 cm**-1 with current tolerances. 0 N Use standard machine tolerance (MDCutO) on maximum and rms. but delete at end of link. IOp(10/15) What to do with expansion vectors from the linear equations. IOp(10/16) Convergence in secondary linear equations (only for simultaneous solution). Save as 2.0). Note that because of numerical instabilities in the simultaneous solution method. S1. otherwise). Force 2nd order cphf for polarizability derivatives. Don't update dipole Update polarizability. Convergence is 10**(-N) for max and rms. which is used only for one term in polarizability derivatives and for which the accuracy requrirements are less stringent. and S2 off the AO 2PDM. but leave the full version of /Orb/ on the disk. Don't update polarizability. Use old vectors if available. =2 Save vectors at end.IOp(10/14) Whether to update dipole and polarizability derivatives.eq. Update dipole. Default (Use old vectors if available). Convert /Orb/ to full. 0 1 2 3 00 10 20 Default (=1 if IOp(8)=1 and electric field only and no derivatives are being computed.

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

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

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

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

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

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

but generate and write out the HF 2PDM here for debugging purposes.Overlay 11 IOp(11/5) IFWRT: DERIVATIVE INTEGRAL WRITE OPTION. I0 148 WHETHER OR NOT TO USE THE CONTENTS OF IRWFX. Yes. YES. YES. 0 1 2 NO. IOp(11/8) IFF1: WHETHER OR NOT TO COMPUTE F1 OVER AO'S. PRODUCE A D2E FILE. IOp(11/7) IFTPDM: WHETHER OR NOT TO CONTRACT INTEGRAL DERIVATIVES WITH A 'READ-IN' TWO-PARTICAL DENSITY-MATRIX. IOp(11/9) IDOUT: FIRST-DERIVATIVE OUTPUT 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. 0 1 DO NOT PRODUCE A D2E FILE. Yes. . 0 1 2 NO. 0 1 No. CONTAINS I2*100+I1*10+I0.

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

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

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

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

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

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

0 1 No. WFN file output WFN file output with magnetic orbital derivatives. if a new version was not generated in this step). but don't remove extra data (i.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. Restart a multi-step job. 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 with GIAO magnetic orbital derivatives. BASIS SET INFO ETC. 157 . PolyAtom output in working precision to Fortran unit 8. 1 2 3 4 5 THE RUN IS NOT AN OPTIMIZATION.e. Use natural orbitals in WFN file IOp(9999/7) Controls whether MOs are written to the polyatom integral tape in LANL style. 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. SAVE THE PERMANENT INFORMATION (MOS. SO BOTH THE PERMANENT AND RESTART FILES ARE IN THE CHECKPOINT FILE. Save data on the chk file. No GVB2P5 trans file. GVB2P5 trans file to unit 14.) ON THE CHECKPOINT FILE.E.. Yes. Archive data from the checkpoint file. DELETE THE RESTART INFORMATION IF THE RUN IS FINISHING NORMALLY (I. IF THE ERROR TERMINATION ILSW BIT IS NOT SET).

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

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

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

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.

scribd