Overlay 1
IOp(1/5)
L103: Mode of optimization.
0  Find local minimum. 
1  Find a saddle point. 
N  Find stationary point on the energy surface with N negative eigenvalues of the 2nd deriv. matrix. 
L107: Mode of search.
0  Locate the maximum in the LST path. 
1  Scan the LST path. 
N000  Which approximation to make. Default is III for Tomasi (interlocking spheres) and IV for general surface. 
1000  Apprx. I: Don’t Do SelfPolarization or “Compensation” 
2000  Apprx. II: Do SelfPolarization, But No Compensation. 
3000  Apprx. III: Do SelfPolarization and Compensation. 
4000  Apprx. IV: Do III, and Allow Surface To “Relax” in Solution if no spheres… 
N0000  Whether to evaluate densities using orbitals or density matrix. Default is to use density. 
10000  Use MOs. 
20000  Use density. 
L121: Time step, N*0.0001 fs, default 0.1.
IOp(1/6)
L102, L103, L105, L107, L109, L113, L114: Maximum number of steps (or number of steps for an LST scan).
0  NSTEP = Max(20,NVAR+10) (L103) 
= Min(20,NVAR+10) (L102, L105, L109)  
= Min(40,NVAR+20) (L113, L114)  
N  NSTEP = N 
IOp(1/7)
L103, L105, L109, L112, L113, L114: Convergence on the first derivative and estimated displacement for the optimization, RMS first derivative < ConvF, RMS Est. Displacement < ConvX = 4*ConvF. For L123, similar convergence control for the GS integrator.
1  ConvF = 1/600 Hartree/Bohr or Radian 
0  ConvF = 0.0003 Hartree/Bohr or Radian 
N  ConvF = N*10^{6} 
L116, L117: Convergence on electric field/charges.
1  Default value for optimizations: 10^{7}. 
0  Default value for singlepoints: 10^{5} in L116, 10^{7} in L117. 
N  10^{N}. 
L123: Regular force/displacement convergence for GS2. For EulerPC, N/1000000 displacement conv.
IOp(1/8)
L103, L109, L112: Maximum step size allowed during Opt.
0  DXMAXT = 0.1 Bohr or Radian (L103, Estm or UnitFC). 
= 0.3 Bohr or Radian (L103, Read or CalcFC).  
= 0.2 Bohr or Radian (L105).  
= 0.3 Bohr or Radian(L113, L114).  
N  DXMAXT = 0.01 * N 
L117: General control.
0  Which type of basin to use to partition the density isosurface. Default is 4. 
1  GradVne. 
2  GradRho. 
3  Don’t Use Basins, Use only the Center of Nuclear Charge. 
4  Use Interlocking Spheres. 
N0  Order of Adam’sBashforthMoulton (ABM) predictorcorrector method to use in solving diff. eqns. for the grad RHO or Vne trajectories. Default is 4; max is 9. 
N00  Number of small steps per ABM step to be used in starting ABM and when “slow down” is needed in ABM. Default is 5. 
N000  Which approximation to make. Default is III for Tomasi (interlocking spheres) and IV for general surface. 
1000  Apprx. I – Don’t Do SelfPolarization or “Compensation” 
2000  Apprx. II – Do SelfPolarization, But No Compensation. 
3000  Apprx. III – Do SelfPolarization and Compensation. 
4000  Apprx. IV – Do III, and Allow Surface To “Relax” in Solution if no spheres… 
N0000  Whether to evaluate densities using orbitals or density matrix. Default is to use density. 
10000  Use MOs. 
20000  Use density. 
L121: Time step, N*0.0001 fs, if N>0, N*0.0001 au if N<0.
IOp(1/9)
L103: Use of Trust radius.
0  Whether to update trust radius (DXMaxT, default Yes). Default is Yes for minima, no for TS. 
1  No. 
2  Yes. 
00  Whether to scale or search the sphere when reducing the step size to the trust radius. Default search for minima, scale for transition states. 
10  Scale. 
20  Search. 
L106: Whether to use symmetry to reduce the number of L110 displacements.
0  Yes. 
1  No. 
10  Do not use symmetry to skip steps back. 
100  Do not use symmetry to skip equivalent atoms 
L107: Whether to maintain symmetry along the search path.
0  Yes. 
1  No. 
L117: Whether to delete points which are too close together and how close to get to the iso surface in search.
0  No; Approx. 1.0 D6 (N=20) 
1  Yes, using a default criterion (0.05 Angstroms). 
N  Yes, using a (10^{N} Angstroms) criteria: 2.0^{N}. 
N  2.0^{N} 
L121: Whether to read in initial velocities.
0  Default (same as 1). 
1  Generate random initial velocity. 
2  Read in initial Cartesian velocity (Bohr/sec). 
3  Read in initial MW Cartesian velocity (sqrt(amu)*Bohr/sec). 
IOp(1/10)
L103, L105, L109, L112, L113, L114: Input of initial Hessian. All values must be in atomic units (Hartree, Bohr, and radians).
0  Use defaults (not valid for L109). 
1  Read ((FC(I,J),J=1,I),I=1,NVAR) (8F10.6) (L103 only). 
2  Read I,J,FC(I,J) (5I3,F20.0) (L103 only). End with a blank card. 
3  Read from checkpoint file in internal coordinates. 
4  Second derivative matrix calculated analytically. (Not valid for L109). 
5  Read Cartesian forces and force constants from the checkpoint file are converted to internal coordinates. 
6  Read Cartesian forces followed by Cartesian force constants (both in format 6F12.8) from input stream, followed by a blank line. 
7  Use semiempirical force constants. 
8  Use unit matrix (default for L105; only recognized by 103). 
9  Estimate force constants using valence force field. 
10  Use unit matrix throughout. 
IOp(1/11)
L103: Test of curvature. Bomb the job if the second derivative matrix has the wrong # of negative eigenvalues.
0  Default (Test for zmatrix or Cartesian TS but not for LST/QST or for minimum). 
1  Don’t test. 
2  Test. 
L117: Step size for ABM method in Trudge for isodensity method.
0  0.05 (N=2). 
N  0.1/N. 
IOp(1/12)
L103: Optimization control parameters.
0  Use default values. 
1  Read in new values for all parameters (see INITBS). 
IOp(1/13)
L103, L113, L114, L115, L123: Type of Hessian Update.
0  Default (9 for L103 minimization, 7 for L103 TS, D2Corr and L115, Powell for L113 and L114, Bofill in L123). 
1  Powell (not in L103). 
2  BFGS (not in L103). 
3  BFGS, safeguarding positive definiteness (not in L103 or L115). 
4  D2Corr (New, only in L103 and L115). 
5  D2Corr (Old, only in L103 and L115). 
6  D2Corr (BFGS). 
7  D2Corr (Bofill Powell+MS for transition states). 
8  D2Corr (No update, use initial Hessian). 
9  D2Corr (New if energy rises, otherwise BFGS). 
L118: Integration method.
L121: Multitime step parameter (NDtrC,NDtrP).
0  No multitime stepping. 
NN  Iterate density constraints NN times per step. 
MM00  Do gradient once every MM steps. 
L123: Hessian update.
0  Default (Bofill). 
1  MurtaghSargent (SR1) update. 
2  PowellsymmetricBroyden (PSB) update. 
3  Bofill’s update. 
4  Sqrt(Bofill) update. 
5  No update (keep old Hessian). 
IOp(1/14)
L103: Maximum number of bad steps to allow before doing a linear minimization (i.e., no quadratic step).
1  0. 
0  Default (0 for TS, 1 for minima). 
N  Allow N — linear only starts with the N+1st. 
IOp(1/15)
L103, L109: Abort if derivatives too large.
1 or 0  No force test at all. 
N  FMAXT = 0.1 * N. 
IOp(1/16)
L103, L113, L114: Maximum allowable magnitude of the eigenvalues of the second derivative matrix. If the limit is exceeded, the size of the eigenvalue is reduced to the maximum, and processing continues.
0  EIGMAX = 25.0 Hartree / Bohr^{2} or Radian^{2} 
N  EIGMAX = 0.1 * N 
IOp(1/17)
L103, L113, L114: Minimum allowable magnitude of the eigenvalues of the second derivative matrix. Similar to IOp(16).
0  EIGMIN = 0.0001. 
N  EIGMIN = 1. / N. 
IOp(1/18)
L103: Coordinate system.
0  Proceed normally. 
1  Second derivatives will be computed as directed on the variable definition cards. No optimization will occur. 
10  Do optimization in Cartesian coordinates. 
20  Do full optimization in redundant internal coordinates. 
30  Do full optimization in pruned distance matrix coordinates. 
40  Do optimization in zmatrix coordinates. 
50  Do full optimization in redundant internal coordinates with large molecular tools. 
100  Read the AddRedundant input section for each structure. 
1000  Do not define Hbonds (default). 
2000  Define Hbonds with no related coordinates. 
3000  Define Hbonds and related coordinates. 
10000  Reduce the number of redundant internals. 
20000  Define all redundant internals. 
100000  Old definition of redundant internals. 
0000000  Default (2000000). 
1000000  Skip MM atoms in internal coordinate definitions and do microiterations the old way, in L103. 
2000000  Include MM atoms in internal coordinate definitions (no microiterations). 
3000000  Skip MM atoms in internal coordinate definitions and do microiterations the new way, in L120. 
4000000  Microiterations for pure MM, done in L402. 
IOp(1/19)
L103: Search selection.
0  Default (same as 6). 
2  Linear and steepest descent. 
3  Steepest descent and linear only when essential. 
4  Quadratic if curvature is correct; RFO if not. Linear as usual. 
5  Quadratic if curvature is correct; RFO if not. No linear search. 
6  RFO and linear. 
7  RFO without linear. 
8  NewtonRaphson and linear. 
9  NewtonRaphson only. 
10  GDIIS and linear. 
11  GDIIS only. 
15  GEDIIS. 
L113, L114: Search Selection.
0  PRFO or RFO step only (Default). 
1  PRFO or RFO step for “wrong” Hessian otherwise Newton Raphson. 
IOp(1/20)
L101, L106, L108, L109, L110: Input units.
0  Angstroms degrees. 
1  Bohrs degrees. 
2  Angstroms Radians. 
3  Bohrs Radians. 
IOp(1/21)
L103, L113, L114: Expert switch.
0  Normal mode. 
1  Expert mode. Certain cutoffs used to control the optimization will be relaxed. These include FMAXT, DXMAXT, EIGMAX and EIGMIN. 
IOp(1/22)
L107: Whether to reorder coordinates for maximum coincidence.
0  Yes. 
1  Assume reactant order equals product order. 
2  Read in a reordering vector from the input. 
L115/L123: Kind of search.
0  Both directions and generate search vector. 
1  Forward direction and generates vector. 
2  Backward direction and generates vector. 
3  Both directions and generates vector. 
4  Forward direction and reads vector 8F10.6. 
5  Forward direction and reads vector 8F10.6. 
6  Backward direction and reads vector 8F10.6. 
7  Both directions and reads vector 8F10.6. 
IOp(1/23)
No longer used.
IOp(1/24)
Whether to round tetrahedral angles.
0  Default (Yes). 
1  Yes, round angles within 0.001 degree. 
2  No. 
IOp(1/25)
Whether SCRF is used with numerical polarizability.
0  No. 
1  Yes, the field in /Gen/ must be cleared each time. 
IOp(1/26)
Accuracy of function being optimized:
NNMM  Energy 10^{NN}, Gradient 10^{MM}. 
1  Read in values. 
0  Default (same as 1). 
1  Normal accuracy for HF (energy and gradient both 1.d7). 
2  Accuracy for DFT with SG1 grid (Energy 1.d5, gradient 1.d4). 
3  Fine grid accuracy for DFT (Energy 1.d7, gradient 1.d6). 
4  Ultrafine accuracy (E 1.d7, G 1.d6). 
5  Superfine accuracy (E 1.d7, G 1.d7). 
6  UltraFine+PCM accuracy (E 5.d5, G 5.d6) 
IOp(1/27)
= IJKL (i.e. 1000*I+100*J+10*K+L).
Transition state searching using QST and redundant internal coordinates
L = 0,1  Input one structure, either initial guess of the minimizing structure or transition structure without QST. 
L = 2  Input 2 structures. The first one is the reactant, the second one is the product. The union of the two redundant coordinates is taken as the redundant coordinates for the TS. The values of the TS coordinate are estimated by interpolating the structure of R and P. R and Pare used to guide the QST optimization of the TS. 
L = 3  Input 3 structures. The first one is the reactant the second one is the product. The third one is the initial guess of the transition structure. R and P are used to guide the QST optimization of the TS. 
K = 19  Interpolation of initial guess of TS between R and P (TS = 0.1*J*R + 0.1*(10J)*P, default J=5). 
J = 1  LST constraint in internals. 
J = 2  QST constraint in internals. 
J = 3  LST constraint in distance matrix space. 
J = 4  QST constraint in distance matrix space. 
I = 09  Control parameters for climbing phase of QST (e.g. QSTRad = 0.01*I, default QSTrad = 0.05). 
IOp(1/28)
L103: Number of translations and rotations to remove during redundant coordinate transformations.
2  0. 
1  Normal (6 or 5 for linear molecules). 
0  Default, same as 1. 
N  N. 
IOp(1/29)
L101: Specification of nuclear centers.
0  By zmatrix. 
1  By direct coordinate input (must set IOp(29) in L202). 
2  Get zmatrix and variables from the checkpoint file. 
3  Get Cartesian coordinates only from the checkpoint file. 
4  By model builder, model A. 
5  By model builder, model B. 
6  Get zmatrix from the checkpoint file, but read new values for some variables from the input stream. 
7  Get all input (title, charge and multiplicity, structure) from the checkpoint file. 
10  Print details of the modelbuilding process. 
000  Default (same as 100). 
100  Do not abort job if model builder generates a zmatrix with too many variables. 
200  Abort job if model builder generates a zmatrix with too many variables. 
1000  Read optimization flags in format 50L1 after the zmatrix. 
2000  Set all optimization flags to optimize. 
3000  Purge flags except the frozen variables. 
4000  Rebuild the coordinate system. 
5000  (2+3) Purge all flags but keep the coordinate definition. 
6000  Generate new redundant coordinates, reading an input section selecting frozen and optimized atoms. 
7000  Mark all internal coordinates as frozen before handling ModRed input 
00000  Default, same as 10000. 
10000  Mark zmatrix constants as frozen variables rather than wiredin constants. 
20000  Do not retain symbolic constants. 
100000  Generate a symbolic zmatrix using all Cartesians if none is present on the checkpoint file (a hack to make IRCs work with Cartesian input). 
200000  Same as 1, but retain the redundant internal coordinate definitions. 
1000000  Get input type or chk file name to read from input stream; title and charge/multiplicity for each structure read from input. 
2000000  Read input type for each structure from input stream; title and charge/multiplicity are those of last chk file read. 
9000000  Same as 0000000. 
00000000  Default (read one set of charge/multiplicity pairs unless both NFrag and ONIOM are set). 
10000000  Read ONIOM charge/multiplicity pairs if reading any. Fragment values will be defaulted from the supermolecule. 
20000000  Read fragment charge/multiplicity pairs if reading any. ONIOM model system values will be defaulted from the real system. 
30000000  Read two lines, with ONIOM followed by fragment values. 
IOp(1/30)
L103: Are the readwrite 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. It also suppresses error termination on large gradients.
0  Yes. 
1  No, print internal coordinate step but don’t set up for microiterations and don’t update the RWF. 
2  Set up for step but don’t update coordinates; for QM/MM iterative frequencies. 
IOp(1/32)
Title card punch control.
0  Don’t punch. 
1  Punch. 
IOp(1/33)
L101, L102, L103, L106, L109, L110, L113, L114: Debug print.
0  Off. 
1  On. 
IOp(1/34)
L101, L102, L103: Debug + Dump print.
0  Off. 
1  On. 
IOp(1/35)
L102L112: Restart.
0  Normal optimization. 
1  First point of a restart. Get geometry, wavefunction, etc. from the checkpoint file. 
IOp(1/36)
Checkpoint.
0  Normal checkpoint of optimization. 
1  Suppress checkpointing. 
IOp(1/38)
Entry control option (currently only by L103, L106, L107, L108, L109, L110, L111, and L112 but not L102 and L105).
0  Continuation of run. 
1  Initial entry. 
N>1  In L103: Initial entry of guided optimization using N levels. 
N0  In L106: differentiate N^{th} derivatives once. In L110 and L111: differentiate energy N times. 
000  In L106: differentiate with respect to nuclear coordinates. 
100  In L106: differentiate with respect to electric field. 
200  In L106: differentiate with respect to field and nuclear. 
1000  In L106: Save original forces and force constants. 
00000  In L106: Assume all quantities available at the central point will also be computed at displaced points. 
10000  In L106: No analytic nuclear coordinate derivatives will be done at displaced points, even though they were done at the central point. 
000000  L106 control of number of diff. steps. 
100000  Do 2point differentiation (one step each way). 
200000  Do 4point differentiation (two steps each way). 
0000000  Default (1). 
1000000  Differentiate with respect to translation vectors for PBC for elasticity. 
2000000  Do not differentiate with respect to translation vectors. 
IOp(1/39)
L106, L109, L110, L111: Step size control for numerical differentiation.
L115: Path step size.
0  Use internal default (0.001 Angstroms and 0.001/3 au Efield in L106, 0.005 Å in L109, 0.01 Angstrom in L110, 0.001 au in L111). 
N  Use stepsize of 0.0001*N (angstroms in L106, L109, L110, electric field au in L111). In L106, the electric field step will be 3x smaller. 
1  Read step size (up to 2 for L106, 1 for others), freeformat. 
N>1  Use stepsize of 0.0001*N atomic units everywhere. 
L123: Stepsize.
0  Default (0.1 Bohr, except 0.075 Bohr for EulerPC, or original value if restart. DVV default is 0.25 fs). 
N<0  Supplied step size is in units of sqrt(amu)*Bohr. 
N>0  Supplied step size is in units of Bohr. 
IOp(1/40)
L113, L114: Hessian recalculation.
1  Pick up analytic second derivatives every time. 
0  Just update. The default, except for CalcAll. 
N  Recalculate the Hessian every N steps. 
L116: Whether to read initial Efield.
0  Start with 0.0. 
1  Read from checkpoint file. 
2  Read from input stream. 
L123: Truncation error threshold for the modified BS integration routine (EPS).
0  Default (10^{8} Bohr). 
N  10^{N} Bohr. 
IOp(1/41)
Take previous geometry from checkpoint file:
N > 0  N^{th} point of geometry optimization (zmatrix only). Converted to (N+1) if no zmatrix. 
N < 0  N^{th} geometry on trajectory file. 
IOp(1/42)
L103, L115, L118, L123: Number of points along the reaction path in each direction, or maximum number of points in one trajectory. Default 10 for IRC, 100 for trajectory.
L117: Cutoff to be used in evaluating densities.
0  1.0D10. 
N  1.0DN. 
IOp(1/43)
L116: Extent of Reaction Field.
0  Dipole. 
1  Quadrupole. 
2  Octapole. 
3  Hexadecapole. 
L117: How to define Radii.
0  Default is 11. 
1  Use internally stored Radii, centers will be on atoms. 
2  Readin centers and radii on cards. 
10  Force MerzKollman radii (Default). 
20  Force CHELP (Francl) recommended radii. 
30  Force CHELPG (Breneman) recommended radii. 
100  Read in replacement radii for selected atom types as pairs (IAn,Rad) or (Symbol,Rad), terminated by a blank line. 
200  Read in replacement radii for selected atoms as pairs (I,Rad), terminated by a blank line. 
Initial radius of spheres to be placed around attractors to “capture” the gradient trajectories. The final radius is then automatically optimized separately for each atom.
0  0.1 
NM  N.M=NM/10 
IOp(1/44)
L115, L123, and L125: IRC Type.
0  Optimization coordinates (Default 3 in L15 and L123, 2 in L125). 
1  Cartesian. 
2  Internal (NYI in L123). 
3  Massweighted Cartesian (NYI in L125). 
M0  Initial interpolation in L125 (default 2). 
L117: Maximum distance between a nucleus and its portion of the isosurface – used in Trudge only to eliminate, from the outset, points which clearly lie in another basin. This parameter should be chosen with the parameter Cont in mind:
0  10.0 au. 
NM  N.M au = NM/10. 
L121: Seed for random number generator (ISeed).
1  Use system time initialize iseed (Note each run will give different results). 
0  Use default seed value (ISeed = 398465). 
N  Set random number seed to N. 
IOp(1/45)
L123: Options.
00  Which IRC integrator to use. Default = 3, except 6 for ONIOM QuadMac/Micro or 1 for GradientOnly. 
01  Euler. 
02  LQA. 
03  HPC. 
04  GS2. 
05  DVV. 
06  Eulerbased PC. 
10  Coordinate driving schemes. 
0xx  Is the integration being done on an empirical surface? Default=2. 
1xx  Yes, this is an empirical surface. The energies and derivatives required for the IRC integration are NOT evaluated in this link. Instead it is assumed that an external program communicates the appropriate values with Link 402, etc. Also, the force constant matrix, when needed, is simply diagonalized, i.e. translation and rotation projections are NOT used. Also, all atomic masses are set to 1. 
2xx  No. 
0xxx  Order of magnitude for the step size relative to the integer value given with the StepSize=N option on the route line — IOp(1/39). Default=2. 
Nxxx  Integration step size is taken as IOp(1/39)*10^{N}. 
0xxxx  Whether or not energies reported in the final summary table should be given relative to the TS energy. Default=1. 
1xxxx  Yes. 
2xxxx  No. 
0xxxxx  Whether or not statistics over coordinates should be converted to Angstoms/Degrees when reported in the summary table. Default=1. 
1xxxxx  Yes. 
2xxxxx  No. 
0xxxxxx  Should a URVA input file be written? Default=2. 
1xxxxxx  Yes. 
2xxxxxx  No. 
0xxxxxxx  Should IRC data besaved to the PES data structure file? Default=1. 
1xxxxxxx  Yes. 
2xxxxxxx  No. 
0xxxxxxxx  When secondorder methods are employed, should the NewtonRaphson step test be carried out? Default=1. 
1xxxxxxxx  Yes. 
2xxxxxxxx  No. 
IOp(1/46)
Order of multipoles in numerical SCRF.
0  Dipole. 
1  Quadrupole. 
2  Octapole. 
3  Hexadecapole. 
IOp(1/47)
Number of redundant internal coordinates to allow for.
0  Default: Max(50000,MCV/(100*NStruc)) 
N  N. 
IOp(1/48)
IRCMax control.
1  Do IRCMax. 
20  Include zeropoint energy. 
IOp(1/49)
Options to IRC path relaxation (IJKL).
L  2/1 don’t/do optimize reactant structure. Default: 1. 
K  2/1 don’t/do optimize product structure. Default: 1. 
J  3/2/1 don’t/QST3/QST2 optimize TS structure (for QST input). Default: 1. 
I  2/1 unimolecular/bimolecular reaction. Default: unimolecular. 
IOp(1/52)
L101 and L120: Type of ONIOM calculation.
0/1  One layer, normal calculation.  
2  Two layers.  
3  Three layers.  
00  Default (20).  
10  Include electrostatics in model systems using MM charges (in case of threelayer ONIOM, this includes the charges in both the small model and the intermediate model system).  
20  No electrostatics included in the model systems.  
30  As 10, but exclude the MM charges in the calculations on the smallest model system in case of a threelayer calculation.  
40  As 10, but exclude the MM charges in the calculations on the intermediate model system in case of a threelayer calculation.  
100  Do full square for testing.  
N000  Use atomic charge type N1 during microiterations. The default is MK charges.  
M0000  Type of link atoms for the MM calculation in QM/MM.  


L00000  Whether to read additional charges with electronic embedding.  


K000000  Whether to create new entries in Common/Mol for the link atoms.  

IOp(1/53)
L120: Action of each invocation of L120.
0  Do nothing.  
1  Set up point MM on RWF from initial data.  
2  Set up point MM on RWF from initial data and restore point MM on checkpoint file if ONIOM data is present there.  
3  Restore point M from data on the RWF.  
4  Integrate energy.  
5  Integrate energy and gradient.  
6  Integrate energy, gradient, and Hessian.  
7  Restore point MM from RWF but do not create a new model system.  
NN0  Save necessary information (some RWF’s, energy, gradients, Hessian) of point NN of the ONIOM grid. NN = MaxLev^{2} + 1 (currently 17) to restore real system.  
MM000  Next point to do is MM.  

S M L system size
IOp(1/54)
Whether to recover initial energy during IRCMax from checkpoint file.
0  No. 
1  Yes. 
IOp(1/55)
L103: Options for GDIIS. ICos*1000+IChkC*100+IMix*10+Method form.
L115: IRC optimization.
0  Default, use gradients to find the next geometry. 
1  Use displacements to find the next geometry. 
IOp(1/56)
Set of atom type names to parse.
0  Accept any. 
1  Dreiding/UFF. 
2  Amber. 
3  Amber allowing any symbol, for use with parameters in input stream. 
IOp(1/57)
Whether to produce connectivity.
0  Default (4 if reading geometry from checkpoint file and connectivity is there, otherwise 3). 
1  No. 
2  Yes, read from input stream. 
3  Yes, generate connectivity. 
4  Yes, read from checkpoint file. 
5  Yes, read from RWF file. 
10  Read modifications. 
100  Connectivity input is in terms of zmatrix entries, including dummy atoms. 
IOp(1/58)
L115: IRCMax control.
1  Do IRC rather than IRCMax. 
10  Store compound energy; default for IRCMax. 
20  Zeropoint energies are available during IRCMax. 
IOp(1/59)
L103: Update of coordinates.
0  Default (1). 
1  New versions (RedCar/RedQ2X); fall back to ORedCr/RedQX1 if RedCar fails. 
2  Old version (ORedCr/RedQX). 
3  Old version (ORedCr/RedQX1). 
4  New version (ORedCr/RedQ2X) but fall back to ORedCr/RedQX if RedCar fails. 
MMMM0  IAprBG in Red2BG. 
10  Reuse eigenvectors of G only if exact. 
20  Reuse eigenvectors of G if they are linearly independent. 
30  Test old eigenvectors of G but don’t reuse them. 
40  Don’t look at old eigenvectors. 
50  Reuse eigenvectors of G if the RMS of the elements of the new G in the old null space is less than the threshold. 
NN00  RMS < 10^{N}, default 4. 
M0000  Default (1). 
10000  Form Ginverse from the B eigenvalues/vectors. 
20000  Form Ginverse directly from G. 
30000  Do G via diagonalization of G (NYI). 
40000  Do G via SVD on B, returning only the eigenvectors with nontrivial eigenvalues. 
IOp(1/60)
Interpret extra integer and fp values in zmatrix as scan information.
0  Default (No). 
1  Yes. 
2  No. 
IOp(1/61)
How ONIOM should leave the RWF at the end of each geometry.
0  Default (1). 
1  Normal: leave the RWF set up for the lowlevel calculation on the real system. 
2  MOMM: leave the RWF set up for the real system, but with NBasis and NBsUse for the highlevel calculation on the model system. Useful for treating the full system as having electrons only on the QM atoms. This is really a hack for two layer QM:MM ONIOM ADMP and should probably be generalized to behave like an ONIOMPCMA case. 
3  Lowest level is MO, but normal setup at end. 
00  Default (10, but 20 if doing EE microiterations). 
10  Leave the charges file (605) from the best calculation that produced one. 
20  Leave 605 in its normal state (present if from real, lowlevel). 
IOp(1/62)
Counterpoise control.
NN  NN fragments. 
1NN  Force use of new ghost atoms (default). 
2NN  Force use of old ghost atoms. 
1xxx  Counterpoise (default). 
2xxx  Fragment guess. 
IOp(1/63)
Step in counterpoise calculation.
LMM  L: order of derivatives: 1 = Energy, 2 = Gradient,… 
MM: 0 = Supermolecule, 1NFrag = Fragments with ghost atoms,  
NFrag+12*NFrag = lone fragments. 
IOp(1/64)
Molecular mechanics force field selection.
0  None. 
1  Dreiding. 
2  UFF. 
3  AMBER. 
000  Use only hardwired. 
100  Use soft and hardwired, hardwired has priority. 
200  Use soft and hardwired, soft has priority. 
300  Use only soft. Lowest 2 digits then have no meaning. 
0000  Do not read modifications to parameter set. 
1000  Read modifications to parameter set. 
00000  With soft parameters, abort when different parameters match to the same degree. 
10000  Use the first when there are equivalent matches. 
20000  Use the last when there are equivalent matches. 
If IOp(67) = 3, then the default is to apply soft parameters with higher priority.
IOp(1/65)
Control of which terms are included in MM, corresponding to the ‘classes’ in FncInf.
0  Do all (default). 
1  Nonbonded. 
10  Stretching. 
100  Bending. 
1000  Torsion. 
10000  Outofplane. 
100000  Stretchbend. 
IOp(1/66)
Whether to generate QEQ charges, overwrite the values in AtChMM, or to use the values already there.
0  Default (2, 1Þ 221). 
1  Do QEq here in GenChg. 
2  Don’t do QEq. 
3  Generate charges here using QEqN. 
00  Default (20). 
10  Do for atoms which were not explicitly typed. 
20  Do for all atoms regardless of typing. 
000  Default (200). 
100  Do for atoms which have charge specified or defaulted to 0. 
200  Do for all atoms regardless of initial charge. 
MMMMM000  IType passed to QEq. 
IOp(1/67)
Source of MM parameters.
0  Default: 2 if reading geometry from checkpoint file, else 1. 
1  Generate here, reading from input if requested by IOp(64). 
2  Copy from checkpoint file. 
3  Pick up nonstandard parameters from checkpoint file. 
IOp(1/70)
L118: Type of sampling (Nact).
0  Default (same as 3). 
1  Orthant sampling. 
2  Classical microcanonical normal mode sampling. 
3  Fixed normal mode energy. 
4  Local mode sampling (thermal sampling based on RTemp). 
Currently 0, 2, 3 and 4 are working.  
10  Read in Hessian from checkpoint for initial sampling. 
L124: SCRF flag.
IJKLLMM  See comments in overlay 3. 
00000000  Default (same as 2). 
10000000  Do the 1st iteration in gasphase. 
20000000  Do the 1st iteration in solution. 
IOp(1/71)
L103: How often to calculate analyic 2nd derivatives:
0  Default (every cycle if IOp(10)=4). 
N  Every Nth geometry, starting with the initial one. 
L115: Whether to print out input files for each structure along an IRC.
0  No. 
1  Yes. 
L118: Hessian update.
1  Gradient only. 
2  Gradient+Hessian, but never calculate full H (only updates). 
0  Full Hessian at every step. 
NN  Try to do NN updates between full Hessians. 
000  Default updating (same as 300). 
100  SR1 Hessian updating algorithm. 
200  PSB Hessian updating algorithm. 
300  Bofill’s Hessian updating algorithm. 
400  Sqrt(Bofill) Hessian updating algorithm. 
500  No update. 
0000  Default (same as 1000). 
1000  Reintegrated updated steps. 
2000  Suppress reintegration. 
L123: Hessian calculation.
1  Gradient only. 
0  Either only update or CalcAll as determined by IOp(10). 
N  Recalculate analytic Hessian every N^{th} calculation. 
IOp(1/72)
L121: Lagrangian constrain method for ADMP (ICType).
Half*Gamma*Tr[(P*PP)^{2}] + Lambda*[Tr(P)Ne] + Eta*Tr(P*PP) 
0  Default same as 7 if no massweighting (IOp(76) < 0). Same as 10 if massweighting (IOp(76) > 0). 
1  Use Lambda and Eta only. (Gamma=0). 
2  Use Lambda, Eta, Gamma. Gamma = .2. 
3  Use Lambda, Eta, Gamma. Gamma = 1. Constraints for scalar mass case. 
4  Use exact constraint Sum(ij)[Vij*(P^{2}P)ij]. 
57  Iterative Scheme same as 4. Different initial guesses. 7 is default for scalar mass case. Constraints for tensorial mass. 
811  Massweighting constraints. Documentation may be found in DVelV1. 10 is default. 
L124: Solvent type for external iteration PCM. (ISolv).
IOp(1/73)
L123: Whether to compute projected frequencies: 0/1/2 Default (No)/No/Yes.
L118 and L121: Initial Kinetic energy of the Nuclei (EStrtC).
0  Default (.1 Hartree). 
N>0  N*microHartree. 
N<0  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 Þ 5³ IAtTyp=6 (no scaling); all layers are scaled by at least as much as ones farther out. The default is 500.
M  Factor for charges one bond away from link atom.  
L0  Factor for charges two bonds away from link atom.  
K00  Factor for charges three bonds away from link atom. IJ etc.  
The actual factors used for each value of IAtTyp are:  

IOp(1/75)
ADMP control flag (ICntrl)
0  Standard ADMP. 
1  Read converged density at every step. 
2  Fix the nuclear coordinates. 
3  Test time reversibility (MaxStp must be even). 
00  Default (20). 
10  Read stopping parameters from input. 
20  Do not read stopping parameters. 
IOp(1/76)
Fictitious electron mass (EMass). Format of input: +/ XXXXZYYYY.
YYYY  Default (1000)  
(Note YYYY > 9999 makes no sense). 

Z  Massweighting option. If IOp(76) < 0, Z is meaningless.  
XXXX  If PBC: Mass of Box Coordinates (BoxMas) = XXXX*.0001 AMU BoxMas=0  
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  Default (0.0 Hartree). 
N>0  N*microHartree IWType is used to figure out how the initial velocity is computed (in gnvelp). 
If XXYYYY < 0: Initial velocity = 0.0 Hartree (i.e., currently same as N=0 above).
IOp(1/78)
L121: Sparse.
N  Sparse here with cutoff 10^{N}, full elsewhere. 
0  Use full matrices or sparse based on standard settings. 
1  Use sparse fixed form. 
IOp(1/79)
L115: IRCMax convergence.
L118, L121: stopping criteria.
0  Default, do not analyze pure rotation and vibration for polyatomic molecules. 
1  Do pure rotational and harmonic normal mode analysis for polyatomic molecule; EBK theory for diatomic vibrational analysis (require equilibrium information for each of the polyatomic molecules from saved checkpoint files and Morse parameters for diatomic molecule). 
2  Do pure rotational and harmonic normal mode analysis for polyatomic molecule; local harmonic vibrational analysis for diatomic molecule (require equilibrium information for each of the fragments from saved checkpoint files). 
3  Do pure rotational analysis and select the best normal mode analysis methods (harmonic and anharmonic) for polyatomic molecule; local harmonic vibrational analysis for diatomic molecule (require equilibrium information for each of the fragments from saved checkpoint files). 
00  Default, use default stopping criteria. 
10  Use user specified stopping criteria. 
IOp(1/80)
L106: 0/1 Cartesian/Normal mode/ Internal coordinate differentiation. 2 is NYI.
L118: = 1 to suppress the 5th order correction after surface hop has been made in Trajectory Surface Hopping calculations. Needs also IOp(10/80 = 1)
L121: Nuclear Kinetic Energy Thermostat Option (currently only velocity scaling is implemented).
0  No Thermostat. 
11XXXXX  Velocity scaling, but only for the first XXXXX simulation steps. (This option is useful, if thermostating is only required during equilibration.) 
1000000  Velocity scaling, all the way through the simulation. 
IOp(1/81)
Nuclear KE thermostat in ADMP — temperature is checked and scaled every IOp(81) steps.
IOp(1/82)
L121: Temperature for nuclear KE thermostat.
IOp(1/83)
Whether to read in frequencies for electric and magnetic perturbations.
0  Default (No). 
1  Yes unless geom=allcheck. 
2  No. 
3  Yes, even if geom=allcheck. 
00  Default (10). 
10  Ensure that the static case is done, by putting a zero frequency at the beginning. 
20  Do not put a zero frequency at the beginning. 
000  Default (100). 
100  Sort the frequencies into increasing order. 
200  Do not sort the frequencies (for debugging). 
IOp(1/84)
Differentiation of frequencydependent properties.
0  No. 
N  Mask for which properties on file 721 will be differentiated. 
IOp(1/85)
Band gap calculation in PBC ADMP.
0  Default (No). 
1  Diagonalize Fock matrix to get band gap, evolution, etc. 
2  No. 
IOp(1/86)
Printing for NMR for ONIOM.
0  Default (1). 
1  Print tensors and eigenvalues. 
2  Print eigenvectors as well. 
IOp(1/87)
ONIOM integration of density.
0  Do not integrate. 
1  Integrate current densities. 
2  Integrate densities specified by following digits: 
K0  Density to use from gridpoint 1 
L00  Density to use from gridpoint 2 
M000 etc.
Values for K, L, M, etc.:
0: SCF  
1: MP first order  
2: MP2  
3: MP3  
4: MP4  
5: CI oneparticle  
6: CI  
7: QCI/CC  
8: Correct to second order 
IOp(1/88)
Whether to read in atomic masses (isotopes).
0  Default (1 if geometry read from input, 4 if geometry read from checkpoint) 
1  Use most abundant isotopes. 
2  Read isotopes from input. The temperature and pressure are read first, for backwards compatibility. 
3  Read isotopes from RWF. 
4  Read isotopes from checkpoint. 
5  (Generated internally) Isotopes were read from chk file during guess=input. 
IOp(1/89)
Maximum allowed deviation from average nuclear KE during ADMP, in Kelvin.
IOp(1/90)
To read in the velocity in Cartesian coordinates.
IOp(1/91)
Nuclear Kinetic Energy Thermostat Option. Average energy (in microHartree) to be maintained during simulation, as required by IOp(80).
IOp(1/92)
Thermostat Option. Maximum allowed deviation from average nuclear KE specified in IOp(81). Also in microHartree.
IOp(1/93)
QM/MM TS vector guess.
000  Default (112) 
1  Retrieve from checkpoint file if available, otherwise diagonalize QM Hessian or read from input. 
2  Do not try to retrieve from checkpoint file. 
10  Diagonalize QM contribution to Hessian. 
20  Read from input. 
N00  How to deal with ‘suspicious RFO solutions’ (default is 1). 
1. Just take the step.  
2. Check if there is an eigenvector with wrong curvature. If there is, flip its eigenvalue.  
3. Check if there is an eigenvector with wrong curvature. If there is, take a small step into this direction, followed by a linear search. This should step out (or stay in) the wrong region, and fix the eigenvalue. 
IOp(1/94)
Davidson control for quadratic microiterations.
OP  Number of initial guess vectors (4).  
MN00  Iteration to scale down number of vectors (5).  
KL0000  Factor to scale down with; 1 for no scaling (2).  
J000000  Whether to do geometry steps when the CG is done (2).  


I0000000  Whether to reuse previous RFO solution or to regenerate guess (1).  


H00000000  Whether (1, default) or not (2) to add 0,…,0,1 guess vector. 
IOp(1/95)
RFO/Davidson control for quadratic microiterations.
MM  Convergence (7). 
LLL00  <0: Regular Davidson. 
0: Only check convergence on first vector, and iterate the lowest root only. Use all the intermediate vectors.  
>0: Only check convergence on first vector, and iterate the new vectors LLL times with the explicit last row/column. This is specifically appropriate for RFO. The last row/column of the Hessian comes after the diagonal elements. 
IOp(1/96)
Options for generating initial guess vectors for RFO/Davidson diagonalization in coupled QM/MM macro steps. Note that other RFO/Davidson diagonalization controls for coupled QM/MM macro steps are available in IOp(97). Format of input: GHIJKLLMM.
MM  Number of initial guess vectors to get from CG steps. The default is 0.  
LL  Number of initial guess vectors from the diagonal of the QM block (4). The default is 4.  
K  Add 0,…,0,1 guess vector?  
0  Default: K = 1;  
1  Yes  
2  No  
J  Add the gradient vector to the guesses?  


I  Prediagonalize a Hessian/RFO matrix without nonbonding contributions? Note that this control is only valid for IOp(98) > 3; otherwise, I is ignored.  


H  Scale factor for the size of the Davidson subspace in early iterations.  


G  Number of vectors to solve using Davidson diagonization.  

IOp(1/97)
RFO/Davidson control for coupled QM/MM macro step. Note that other RFO/Davidson diagonalization conrols for coupled QM/MM macro steps are available in IOp(96). Format of input: GHIJKLMM.
MM  Convergence in Davidson iterations. Convergence is set to 10^{MM}. The default value is MM=5. 
L  What is being diagonalized? This option is set explicitly in subroutines before calling the Davidson diagonalization code. Therefore, the value set in this IOp is ignored and serves only as a place holder. 
0  the Hessian 
1  the augmentedHessian/RFO matrix 
K  Check convergence on which roots? 
0  Default: For L=0, K=2; For L=1, K=1 
1  Check convergence on lowest root only. 
2  Check convergence on all roots. 
J  Appears to be unused. 
I  Number of Davidson iterations to store. 
0  Default: Keep all iterations. 
1  Keep only the last iteration. 
H  Number of new vectors to create in each Davidson iteration. 
0  Default: For L=0, H=1; For L=1, H=2. 
1  Iterate all roots/vectors. 
2  Iterate lowest root/vector only up to the maximum number of iterations that are default in the Davidson code (ignores J above) and keeping vectors from all iteratations (ignores I above). 
3  Iterate lowest root/vector only. Note that this option is essentially the same as H=2, though J and I option settings are honored. 
G  Initial approximation to use for Davidson diagonization. 
0  Default: G=2. 
1  Use a diagonal Hessian approximation together with the gradient vector. This is best used in RFO applications 
2  Use the inverted Hessian for the QM block 
3  Use a diagonal Hessian approximation. 
IOp(1/98)
Control of quadratic microiterations and coupled QM/MM quadratic macro step.
<0  Do not use dynamic convergence criteria for the microiterations. 
0  Default (11). 
1  Regular noncoupled macro step. 
2  Coupled macro step, full diagonalization. 
3  Coupled macro step, direct /w full Hessian in core. 
4  Coupled macro step, direct /w MM Hessian in core. 
5  Coupled macro step, fully direct. 
6  Go through the QuadMac machinery, but use the fully integrated ONIOM Hessian to calculate the Hessianvector products. Switch to regular microiterations at points without analytic second derivatives. 
7  Fully quadratic at 2nd derivative points (1st in CalcFC, all in CalcAll), QuadMac with integrated Hessian at non2nd derivative points. 
10  Regular microiterations. 
20  Quadratic microiterations, full diagonalization. 
30  Quadratic microiterations, direct with prepared Hessian in core. 
40  Quadratic microiterations, direct with raw MM Hessian in core. 
50  Quadratic microiterations, fully direct. 
60  No microiterations. 
IOp(1/99)
Accuracy used for the nonbonded interactions in the Hessianvector product calculations in the fully direct algorithms. Setting this IOp does not affect the MM energy and gradient calculations; only the direct evaluation in the QuadMac optimization step. When IOp(99) < 0, the full accuracy is used.
MM  Maximum multipole level (8) 
LLL00  Box size in FMM (12) 
KKK00000  Cutoff in van der Waals (30) 
IOp(1/101104)
Phase control in L115 and L118: N1, N2, N3, N4.
IOp(1/105)
Reaction direction.
DampedVelocity Verlet (DVV) options for Dynamic Reaction Path following.
00  Default (Same as 10). 
10  Forward direction. 
20  Reverse direction 
0  Default (Same as 2). 
1  Use DVV. 
2  Do not use DVV. 
00  Default (Same as 10). 
10  Follow the rxn path in the forward direction 
20  Follow the rxn path in the reverse direction. 
000  Default (Same as 200). 
100  Time step correction not used. 
200  Time step correction used but not to recalculate current DVV step. 
300  Time step correction used and current DVV step recalculated. 
0000  Default (Same as 1000). 
1000  Use DVV stopping criteria. 
2000  Do NOT use DVV stopping criteria 
IOp(1/106)
Damping constant for DVV Dynamic Reaction Path following (v0).
0  Default v0 = 0.04 (N=400). 
N  v0 is set to N*0.0001. 
IOp(1/107)
Error tolerance for DVV time step correction (Error).
0  Default Error = 0.003 (N=30). 
N  Error = N*0.0001. 
IOp(1/108)
Gradient magnitude for DVV stopping criteria (Crit1).
0  Default (N = 15). 
<0  Turn off this test. 
N  N*0.0001 
IOp(1/109)
Forcevelocity angle for DVV stopping criteria (Crit2).
0  Default (90 Degrees). 
<0  Turn off this test. 
N  Use N Degrees. 
IOp(1/110)
Scaling of rigid fragment steps during microiterations.
0  Do not scale. 
1  Scale with 1/NRA (NRA = number of atoms in fragment). 
2  Scale with 1/Sqrt(NRA). 
n  Scale with 1/n. 
IOp(1/111)
L103: Stepsize to use with steepest descent when L103 is having trouble.
N  Scale up to RMS step of N/1000 if DXRMS is less. 
1  Effectively disables the scaling. 
0  Default (50). 
N  Scale up or down to maximum change in a variable of N/1000. 
IOp(1/112)
L101: Temperature for thermochemistry.
0  Default (standard temperature, unless read in). 
N  N/1000 degrees. 
1  Default. 
N<1  N/1000000 degrees. 
N  N/1000000 degrees. 
IOp(1/113)
Pressure for thermochemistry.
0  Default (1 atmosphere, unless read in). 
N  N/1000 atmospheres. 
1  Default. 
N<1  N/1000000 atmospheres. 
IOp(1/114)
Scale factor for harmonic frequencies for use in thermochemistry and harmonic vibrationrotation analysis.
0  Default (1 unless specified by IOp in overlay 7 or read in). 
1  Force to 0 (default). 
N  N/1000000. 
IOp(1/115)
Compression for MOMM quadratic steps.
4  Second derivatives generated over active atoms, with real system terms done iteratively during microiterations. 
N¹4  Full second derivative matrices are used. 
IOp(1/116)
Options for ONIOM Conical intersections: which calculations have adiabatic couplings done.
0  Default (111 all component calculations). 
IOp(1/118)
Dump structures for each ONIOM system formatted as input.
0  Default (No). 
1  Yes. 
IOp(1/119)
Control Initial Lanczos Vector (ILzVec).
1  Read guess by card in input file. 
2  Use the largest elements of H as a guess. 
3  Use the five largest contributions of H as a guess. 
0x  For Opt, IRC, dynamics read guess from previous cycle. 
1x  For Opt, IRC, dynamics generate a fresh guess for each cycle 
IOp(1/120)
Flags for QM:QM embedding. NOTE: The standard embedding flags must also be set in the same way as necessary for QM:MM embedding calculations.
0  Default – Same as 1. 
1  Use Mulliken charges. 
1  Use the nuclear charge stored in array AtmChg. 
2  Set the charges to zero. 
00  Default (Same as 20). 
10  Just use the charges that already reside in AtChMM. 
20  Update AtChMM using current atomic charges on the RWF. 
This option is only employed immediately following lowlevel realsystem subcalculations.
IOp(1/121)
L106: Extra items to differentiate.
0  Default, none. 
1  Differentiate AO density and Fock matrices. 
2  NYI. 
IOp(1/122)
Read secondary structure information.
0  Default (4 or 3 if reading geometry from checkpoint or RWF file, otherwise 2). 
1  No. 
2  Yes, read from input stream if any residue information was provided on the atom definition lines. 
3  Yes, read from RWF. 
4  Yes, read from checkpoint. 
IOp(1/123)
Version of /Mol/ to save on disk.
0  Default (current, version 2). 
1  Version 1 (flag 12345). 
2  Version 2 (flag 12346). 
N<0  Flag value N. 
IOp(1/124)
Flavor of ONIOMPCM to use.
1  “A,” reaction field from the ONIOM integrated density; 
2  “B,” reaction field from the real system low level; 
3  “C,” reaction field in the real system low level only; 
4  “X,” reaction field in all subcalculations using the real system cavity. 
0x  Default (same as 1 unless a semiempirical method is involved); 
1x  Integrate the density for ONIOMPCMA; 
2x  Integrate the potential for ONIOMPCMA; 
1xx  Flag to indicate ONIOMPCMX as first iteration of ONIOMPCMA. 
IOp(1/125)
Solvent charge distribution to add to Hamiltonian.
0  None. 
1  Read charges and DBFs from input stream in input orientation. 
2  Read from RWF. 
3  Read from checkpoint file. 
4  Same as 1. 
5  Read charges and DBFs from input stream in standard orientation. 
10  Force units of Angstroms for coordinates. 
20  Force units of Bohr for coordinates. 
IOp(1/126)
Whether to read an input section with atom opt/freeze information.
0  Default (2). 
1  Yes. 
2  No. 
IOp(1/127)
Use of MM coordinates and forces in GEDIIS:
0  Default (3 for ME, 3 for EE). 
3  Don’t pass MM info. 
2  Zero MM forces and MM part of step, equivalent to not passing MM info. 
1  Zero MM forces but interpolate step in MM coords. 
N  Use MM forces scaled by 1/N. 
IOp(1/128)
Initial microiterations in L120 before first QM step, and microiterations in L120 during numerical differentiation in L103.
0  Default (No). 
1  Yes. 
2  No. 
IOp(1/129)
L123: Whether or not the final statistics printed in the job summary should include coordinate values at each IRC step.
0  Default (none). 
1  Read userdefined stats from the input file. 
0x  Default (do not report all Cartesian coordinates). 
1x  Report all Cartesian coordinates. 
0xx  Unused. 
0xxx  Default (do not report bond coordinates). 
1xxx  If redundant internals are on the RWF, then report values for bond coordinates along the IRC. 
0xxxx  Default (do not report angle coordinates). 
1xxxx  If redundant internals are on the RWF, then report values for angle coordinates along the IRC. 
0xxxxx  Default (do not report dihedral coordinates). 
1xxxxx  If redundant internals are on the RWF, then report values for dihedral coordinates along the IRC. 
IOp(1/130)
Eigenvector number to be followed during coordinate driving jobs [IOp(1/19 = 10)].
0  Default (1). 
N  Follow input structure Hessian eigenvector number N. 
IOp(1/131)
Options for corrector integration in predictorcorrector IRC calulcations (Link 123).
00  Should the corrector integration scheme be run in an (macrocycle) iterative fashion? Default = 2. 
01  After each corrector integration, evaluate the actual energy and derivatives, but do not actually use these. The final IRC will be the same as 1. 
01  Never check convergence of the corrector integration. Always do one corrector integration for each predictor integration. 
02  Always check for convergence of the corrector integration end point. Convergence is acheived when the change in corrector integration end point geometry is less than the convergence criteria indicated by IOp(7). 
03  Same as 2, but never accept convergence after the first corrector integration at a point. 
10  This flag iteratively refines the DWI fitted surface if multiple corrector integration macrocycles are taken by adding the largest standard deviation point from the previous BS cycle. 
11  This flag forces a PES evaluation step after each corrector integration. This is similar option 1, except that the actual energy and derivatives are used for the next predictor step rather than the values on the DWI fitted surface at the corrector endpoint. 
0xx  Should DWI surfaces employ numerical thridorder terms in Taylor series? Default = 1. 
1xx  Use DWI surface with Taylor series expansions truncated at secondorder. 
2xx  Use DWI surface with Taylor series expansions truncated at thirdorder using numerical thirdderivatives. 
Nxxx  What power should be used for DWI weights that include deltax raised to an even power. The value of this option setsthat power to 2*N. Default = N = 1. 
0xxxx  How should the first step off of the transition structure point be handled in corrector integration cycles? Note, in all cases, the transition vector is used to define the tangent at the transition structure. Default = 2. 
1xxxx  Run the requested number of Euler steps in the standard way. Only the first Euler step taken uses the transition vector. 
2xxxx  Take a large step off of the transition structure point along the transition vector. This step is taken to be half of the total requested step size given by TotStp. 
3xxxx  This the same as option 2 in concept. The only difference is that the first step off of the transition structure is taken as onethird the total requested step size given by TotStp. 
4xxxx  This the same as option 2 in concept. The only difference is that the first step off of the transition structure is taken as onefourth the total requested step size given by TotStp. 
0xxxxx  Should update vectors be used in DWI fits if possible? Default = 1. 
1xxxxx  Yes, when possible. 
2xxxxx  Never. 
IOp(1/132)
Whether to check for divalent link atoms in ONIOM input:
0  Default (yes). 
1  Yes. 
2  No. 
IOp(1/133)
Suppress integration/restore of quantities for Polar=Raman and Polar=ROA ONIOM jobs:
0  Default (1). 
1  Do not restore or integrate forces, force constants, static electric or magnetic field derivatives. 
2  Restore all files. 
IOp(1/135)
MM nonbonded switching function.
IOp(1/136)
MM van der Waals outer cutoff in Angstroms.
IOp(1/137)
MM Coulomb outer cutoff in Angstroms.
IOp(1/138)
MM soft cutoff range (applied to both vdW and Coulomb, in Angstroms.
IOp(1/139)
Number of MM microiterations allowed.
0  Default, based on NAtoms but at least 5000. 
N  N 
IOp(1/140)
Whether to restore the real system from chk file:
0  Default (yes if ONIOM). 
1  Yes. 
2  No. 
IOp(1/141)
Control of error choice during GEDIIS:
0  Default (1). 
1  Use RFO steps as error vectors, using the NR step at a point if the RFO fails or gives a Hessian eigenvector. 
2  If RFO fails for any point, use the gradient for all points. 
3  Always use the gradients as the errors. 
4  Use RFO steps as error vectors unless any RFO fails and unless Hessian is marked as untrustworthy; then use gradients instead. 
5  Drop back to RFO (no DIIS) if Hessian is untrustworthy. NR steps are used if RFO fails during DIIS. 
IOp(1/142)
Whether to copy MM charges to link atoms:
0  Default (3 if QEq is done; otherwise 1). 
1  Copy if link atom charge is zero. 
2  Do not copy. 
3  Always copy. 
IOp(1/143)
Hessian during IRC restarts.
0  No change in when Hessian is done. 
1  Do Hessian at first new point in each direction. 
IOp(1/144)
Whether to analyze residue geometry.
0  Default (Yes, if secondary structure present and N Atoms ≤ 10000). 
1  Yes. 
2  No. 
IOp(1/145)
Controls for the new internal coordinate data structure.
0  Should the new internal coordinate data structure be setup for use? (Default = 1). 
1  No. 
2  Yes. The coordinates are either generated by default (e.g., Option 9×2) or obtained from the userprovided input (e.g., Option 6×2). Alternatively, they can be taken from a chk file if they exist there and if geom=check without any GIC keyword (i.e., IOp(1/145=0) is used. 
3  Yes, but do two calls to CrdDef when building the coordinate definitions. The first call adds the coordinates in the same way that Option 2 does. The second call adds coordinates from a list of userprovided input coordinates in the same way as Option 6×2 does. In other words, there will be two sets of coordinates to be merged. The first set is generated by default (e.g., in the case of Option 9×3 without geom=check) or it comes from a chk file (Option 9×3 with geom=check). The second set is provided by the user as a separate input section, and it contains some additional coordinates or modifications to the first set. A blank line can be used instead of the second set’s input section. It is also possible to make the userprovided input’s reading to be the first step and the construction of complementary coordinates by CrdCon to be the second step. The latter is Option 6×3. 
0x  Which coordinate derivative terms should be included in the data structure? (Default = 3). 
1x  Include only values in the definitions list. 
6xx  Read the coordinate definitions from the input file using the symbolic form, (See Routine CrInp1), and then automatically differentiate the active coordinate derivatives as needed. 
7xx  Define (the full set of) distance matrix coordinates. This option can be combined with 2xxx to use inverse distance matrix coordinates. 
8xx  Read the coordinate definitions from the input file using the OLD symbolic form (See Routine CrDfSy), and then automatically differentiate the active coordinate derivatives as needed. 
9xx  Use the new coordinate generation code to create an intermediate RP structure based on the molecular connectivity (See Routine CrdCon), then have CrdDef build the data structure using the intermediate RP data from CrdCon and build derivatives wrt Cartesian coordinates. 
0xxx  When building the internal coordinate definitions, should any systematic modifications be done? (Default = 1). 
1xxx  No. Simply define the coordinates to be the same as they would have been using the old coordinate code. 2xxx … Invert all bond stretch coordinates. 
2xxx  Invert all bond stretch coordinates. 
0000000  Default for default is 1 (GIC). 
1000000  Default to 1 if not specified. 
2000000  Default to 2 if not specified. 
IOp(1/146)
Control of SCVS:
0  Default (03045011). 
1  Include forces in virial ratio. 
2  Do not include forces in virial ratio. 
1x  Use Murdoch’s extrapolation. 
2x  Do not use Murdoch’s extrapolation. 
Nxx  Apply SCVS when max force on nuclei is below 10^{N}. 
Mxxx  The convergence on the virial Eta. Default is 5. 
Lxxxx  The convergence threshhold on E * (2*Eta – 2), where E is the magnitude of the kinetic energy is 10^{M}. Default is 3. 
Kxxxxx  The initial order of extrapolation is 10^{L} 
JJJxxxxxx  Maximum of KKK SCVS iterations. 
IOp(1/148)
L106: Storage of derivatives.
0  Default (221). 
1  Normal storage of numerical first derivatives. 
2  Store numerical first and diagonal second derivatives. 
10  Print differences between numerical and analytic quantities. 
20  Do not print differences. 
N00  Threshold for printing differences is 6N. 
IOp(1/150)
L112: Initial scale factor.
0  Default (1.0) 
N  1 + N/100000000 
IOp(1/151)
How many vectors with negative curvature to use in downhill step during minimization:
0  Default (3) 
1  None, do regular RFO step. 
N  Up to N vectors. 
IOp(1/152)
L103: Control of MaxStp (allocated max number of steps in L103).
0  Default: compute based on number of variables, NStep, etc. 
N>0  Make MaxStp at least N. 
N<0  Make MaxStp at least N. 
IOp(1/153)
L103: Whether to fill OT file.
0  Default: do so for Cartesian and redundant internal coordinate optimizations. 
1  Yes 
2  No 
IOp(1/154)
Linear angle test during internal coordinate generation.
0  Default (15 degrees, applied to all 3 tests). 
N  Threshold N degrees, applied to the angle only. 
N  Threshold N degrees, applied to all 3 tests. 
IOp(1/155)
Number of steps to take in guessing a TS during QST2:
0  Default (10). 
N  Divide the overall step into N increments. 
IOp(1/156)
Automatic generation of internal coordinates.
0  Default (1). 
1  Generate bonds, angles, and dihedrals. 
2  Generate bonds and angles but no dihedrals. 
3  Generate bonds only. 
4  Generate no coordinates automatically. 
IOp(1/157)
Maximum step when going down an eigenvector:
0  Default (0.6) 
N  N/1000 
IOp(1/158)
L124: PCM defaults; L301: Details.
IOp(1/159)
Default maximum step in Cartesians during redundant coordinate optimizations.
1  Unlimited (10^{6} Bohr). 
0  Default (3 Bohr). 
N  N/10 Bohr. 
IOp(1/160)
Type of fragment calculation:
0  Default (11). 
1  Force use of new ghost atoms in counterpoise. 
2  Force use of old ghost atoms in counterpoise. 
10  Counterpoise. 
20  Fragment guess. 
80  Excitation energy transfer. 
0xxx  Default (same as 1). 
1xxx  Use fullsystem PCM cavity for fragments. 
2xxx  Use fragment PCM cavity for fragments. 
IOp(1/161)
L123: Hack to save results at each IRC point.
0  Default (No). 
1  Yes. 
IOp(1/162)
Frequency of analytic Hessians during IRC corrector cycles.
IOp(1/163)
Copy of external input section from chk file:
0  Default; copy if Geom=AllCheck. 
1  Copy regardless. 
2  Do not copy. 
IOp(1/164)
Read of atomic pair potential.
0  Default; copy from chk if Geom=AllCheck. 
1  Read from input. 
2  Do not read. 
3  Read from chk. 
Each pair is given on one line of the form: IA,IB,R0,A,B,ROn,ROff,C(n),C(n1),…,C(0).
The potential is of the form P(X)*Exp(A*X)*Exp(B*X*X)*Sw(Rab,ROn,ROff).
If A is zero, the exponential is omitted. If B is zero, the Gaussian is omitted, and if ROn is zero, the StratmannScuseria switching function Sw is omitted. P(X) is the polynomial of degree n given by the specified coefficients. n is determined from the # of C’s provided and can be 0 for a constant scale factor.
IOp(1/165)
Convergence of MM microiterations.
0  Default (10x tighter than macro, except 10x for FOSimult with MM included). 
N  Nx tighter. 
IOp(1/166)
L103: Maximum number of vectors in DIIS. Format of input: MMNN.
0  Default. 
NN  Save N vectors (default 10 for GEDIIS, 10 for SimOpt). 
MMNN  Mix up to MM < NN vectors in DIIS when mixing RFO steps. (Default NN). 
A negative value requests uses of the Hessian eigenvector basis for the step. This is the default and only choice for GEDIIS TS optimizations.
IOp(1/167)
Version of GEDIIS.
0  Default (2). 
1  Old. 
2  New. 
IOp(1/168)
GEDIIS switches.
NN  Switch from RFO to DIIS on RMS force (10^{NN}, default 1.d3). 
MM00  Switch toEnDIIS from RFODIIS on RMS step (10^{MM}, default 2.5d3). 
LLL0000  Maximum coefficient allowed in RFODIIS before space is reduced (LLL/10, default 10.0). 
KKK0000000  Maximum coefficient allowed in RFODIIS before coefficients are adjusted (KKK/10; default 3.0). The minimum value is KKK/10 + 1. 
IOp(1/170)
Control for selecting the initial conditions.
1  Initial Cartesian coordinates and velocities are readin from the input file. This data is read in a freeformat fashion. 
2  Initial Cartesian coordinates and massweighted velocities are readin from the input file. This data is read in a freeformat fashion. 
3  Initial coordinates are given by the userinput geometry; initial velocities are determined by selecting a random velocity which gives an kinetic energy equal to the value set in IOp(73). 
IOp(1/172)
Whether to print internal coordinates in L101:
0  Default (No). 
1  Print if present. 
IOp(1/173)
Extra files to differentiate in L106:
11  Differentiate the files related to groundstate postSCF 2nd derivatives (Lagrangian, P, and W). 
10  Differentiate the files related to CIS/TD second derivatives (Lagrangian, P, W, and T). 
2  Read two lists, each terminated by a blank line, the first the files to differentiate and the second the files where the numerical derivatives will be stored (0 for items which will not be saved). 
1  Read a list from an input section, terminated by a blank link. The numerical derivatives will be printed but not stored on disk. 
N>0  Do file number N, storing derivatives in file 795. 
IOp(1/174)
Reading parameters for classical electrostatics and dispersion.
0  Default, 3 if Geom=AllCheck else 2. 
1  Read selected parameters for elements from input. 
2  Use default parameters. 
3  Read userspecified parameters from chk file if present. 
4  Read both userspecified and default parameters from chk file if present. 
IOp(1/175)
Electronic embedding type:
0  Default (1) 
1  ONIOM style. 
2  QM/MM style. 
IOp(1/176)
Box size used for generating connectivity.
0  Default (8 Bohr). 
N  N Bohr. 
IOp(1/179)
Use of subprocesses for energy+derivatives, for debugging:
0  Default (1002). 
1  Yes. 
2  No. 
3  Yes, using loop over steps in L106. 
4  Yes, using loop over calls to utility in L106. 
5  Yes, using Linda. 
1x  Debug output and save of subprocess files. 
1xx  Full copy of rwf back and forth. 
1xxx  Iterate purturbations forward. 
2xxx  Iterate purturbations backward. 
IOp(1/180)
Stopping during subprocess execution, for debugging.
0  Don’t stop. 
N  Stop after running N subprocesses. 
IOp(1/181)
Maximum number of middle iterations during ONIOMEE:
0  Default (40). 
N  N 