All Classes Functions Variables Enumerations Enumerator Pages
User Manual

Table of Contents

Constants
FILE-IO
SYSTEM
DEBUG
GEOMETRY
LATTICE
PBC
BASIS
JOBS
Common
HF
DFT
MP
RPA
CI
PCM
PES-SCAN
NEB

    General:

      Several parts of this program are under construction, i. e. by now there is no warranty that what you want to compute is actually working.

      The input-file's content is case-insensitive.

      Basic keywords are of general nature, employed throughout the input file.

      keyword description
      # indicates a comment, if placed at the beginning of any line
      END end of any opened list input, e.g. a chunk section or any list
      Anything within an input file that is not recognised is cowardly ignored. This implies that you may freely put comments into the input file, also without any leading #. The # character may be employed to disable individual keywords.

    Constants:

      A list of constants employed by QuMuLuS++.

        Angstroem to Bohr conversion is a nasty topic...
        In fact, various program packages employ different numbers!

      variable value description
      angstroem_to_bohr__NIST 1 / 0.529177210818 NIST: http://www.u.arizona.edu/~stefanb/linkpages/conversions.html
      angstroem_to_bohr__psi4 1 / 0.52917720859 Psi4: http://sirius.chem.vt.edu/psi4manual/4.0b5/psithoninput.html
      angstroem_to_bohr__g09 1 / 0.5291772086 Gaussian: http://www.gaussian.com/g_tech/g_ur/k_constants.htm
      angstroem_to_bohr__tm 1 / 0.5291772083 Turbomole: http://www.turbo-forum.com/index.php?topic=316.0
      angstroem_to_bohr angstroem_to_bohr__NIST QuMuLuS++
      variable value description
      amu_to_kg 1.09776920066e-40 atomic mass unit in kg
      au_to_amu 5.485799097e-4 atomic unit to atomic mass unit
      u_to_kg 1.660538782e-27 unit mass in kg
      au_to_debye 2.54158025294e+0 atomic unit to Debye
      hartree_to_eV 2.721139613182e+1 hartree to eV
      hartree_to_kcalmol 627.5095 hartree to kcal/mol
      hartree_to_MHz 6.579684e+9 hartree to MHz
      hartree_to_wavenr 219474.6 hartree to Wave number
      hartree_to_J 4.3597482e-18 hartree to Joule
      hartree_to_K 3.157747e+5 hartree to Kelvin
      hartree_to_divcm 219474.63067 hartree to 1/cm
      hartree_to_Hx 6.5796839207e+15 hartree to Hertz
      molecule_to_mol 6.0221415e+23 Avogadro's number
      Planck_constant 6.62606896e-34 Planck constant in Js
      Boltzmann_constant 1.3806504e-23 Boltzmann constant in J/K
      Avagadro_number 6.02214179e+23 Avagadro's number
      debye_to_Cm 3.335640952e-30 Debye to Cm
      e0_in_F_per_m 8.854187817e-12 Vacuum permittivity in F/m
      me_in_kg 9.10938215e-31 Electron rest mass in Kg
      speedoflight_in_m_per_s 2.99792458e+8 speed of light in m/s
      twopic_in_m_per_s 1883651567.31 2.e+0 * M_PI * speedoflight_in_m_per_s
      twopic_in_cm_per_s 188365156731 2.e+2 * M_PI * speedoflight_in_m_per_s
      eigenvalue_to_one_per_s2 2.62550196655e+29 1.e+20 * hartree_to_J / u_to_kg
      udivcm2_to_millidyneperangstroem 1.49241795384e-07 1.e+3 * u_to_kg * speedoflight_in_m_per_s * speedoflight_in_m_per_s

    FILE-IO

    line-wise data

    END

    go to table of contents

      The FILE-IO chunk defines certain directories within the internal file-system.

      SCRATCH [path]

        This path defines where all scratch data is written to. Default is "./scratch_[puid]/" with [puid] as the unique process identifyer.

      OUTPUT [path]

        This path defines where the output is written to. Default is "./".

      BASIS [path]

        This path defines where external basis set data is read from. Default is "./".

      GEOMETRY [path]

        This path defines where external geometry data is read from. Default is "./".

    SYSTEM

    line-wise data

    END

    go to table of contents

      The SYSTEM chunk sets up very basic and overall properties to the system.

      MULTIPLICITY [integer]

        The system's multiplicity S = (2s + 1) with s being the sum of alpha / beta spins.

      CHARGE [integer]

        The system's overall charge, supposed to be an integer, as the sum of nuclear charges and eletrons must not be fractional.

    DEBUG

    line-wise data

    END

    go to table of contents

      The DEBUG chunk toggles various print-levels.

      Basically, individual sections of the overall workflow can be addressed and set to one of three levels of information depth.

      [section] [print-level]

        [print-level] description
        SILENT minimize the information to a still rational amount
        NORMAL default level of information
        VERBOSE maximize the information to a still rational amount
        [section] description
        GEOMETRY addresses the entire geometry setup
        BASIS addresses the entire basis set setup
        SHELLPAIRS addresses the entire shellpair assembly
        GUESS addresses the initial density guess phase
        DSCF addresses the direct SCF iteration in general
        POST-SCF addresses all post-SCF computations in general
        EINT-1 addresses the one-electron integral evaluation
        EINT-2 addresses the two-electron integral evaluation
        EGRAD-1 addresses the one-electron integral 1st derivative evaluation
        EGRAD-2 addresses the two-electron integral 1st derivative evaluation
        EHESS-1 addresses the one-electron integral 2nd derivative evaluation
        EHESS-2 addresses the two-electron integral 2nd derivative evaluation
        MP addresses all MP variants in general, excluding the integral evaluation
        DFT addresses all DFT variants in general, including the grid setup and pruning
        GEOM-OPT addresses any wrapping geometry optimization
        MATRICES addresses the output of matrices in general (SILENT: none, NORMAL: selected, VERBOSE: all of them)

    GEOMETRY

    line-wise data

    END

    go to table of contents

      The GEOMETRY chunk contains information on the employed geometry, respectively.

      Basically, atoms are provided via either some kind of symbol-x-y-z format or via internal coordinates (z-matrix).

      In any case, so-called ghost-atoms can be defined by adding a star (*) as a suffix or a prefix to the symbol (e.g. H*, *O, ...). Ghost-atoms carry all their assigned basis functions, but do not contribute to the computation otherwise. This is for example employed in BSSE computations. It might be quite nifty for other intents as well.

      Multiple input sources can be combined to sum up to a total geometry, i.e., the total geometry can be assembled from parts.

      FILE [format] [unit] [file]

        Geometry information is read from [file] expecting [format] in units of [unit].
        [format] description
        XYZ xyz file format
        PDB pdb file format
        CIF cif file format
        TURBOMOLE Turbomole coord file format
        ZMATRIX Z-matrix file format without variable definition
        [unit] description
        BOHR atomic unit, i.e., Bohr
        ANGSTROEM Angstroem, i.e., 100 pm

      Templates for [format]:

          XYZ (the default xyz-file is in Angstroem):

          [xyz-file content]
          PDB (the default pdb-file is in Bohr):

          [pdb-file content]
          CIF (the default cif-file is in Bohr):

          [cif-file content]
          TURBOMOLE (the default coord-file is in Bohr):

          [coord-file content]
          ZMATRIX (the default z-matrix-file is in Angstroem and degrees and CW (clock-wise) (instead of CCW: counter-clock-wise)):

          line entry
          1: [symbol]
          2: [symbol] [bond index] [bond value]
          3: [symbol] [bond index] [bond value] [angle index] [angle value]
          3 to n: [symbol] [bond index] [bond value] [angle index] [angle value] [dihedral index] [dihedral value] [0/1 for CW/CCW]

      LIST [format] [unit]

        Geometry information is read expecting [format] in units of [unit].
        Data is read, as from a respective external file, in between the start of the LIST and the next END keyword.

      POINTCHARGES FILE [unit] [file]

        Pointcharge information is read from [file] expecting [format] in units of [unit]. This toggles pointcharge embedding for all defined jobs.
        [format]:
            line entry
            1 to n: [x] [y] [z] [charge] ([sigma] [e-charge])

          The 5th and 6th columns may, optionally, contain defined sigma and e-charge parameters for so-called charge-blurring. In the attempt of charge-blurring, each pointcharge is thought of as an atom that carries its nuclear charge in the 4th column. It is then surrounded by an exponential function, respesenting an electron clound of negative e-charge in the 6th column. This could is damped in all its Coulomb interactions by the exponential Gaussian parameter sigma, i.e., a damping coefficient. The set of working equations contains exp(-sigma ...) factors, so usually sigma has a positive value. Furthermore, not all charges have to have their optional arguments, i.e., simple pointcharges coincide peacefully with blurred charges.

        NOT YET IMPLEMENTED:

          Charge blurring is not yet implemented. This feature is on the list of to-be-implemented algorithms.

      POINTCHARGES LIST [unit]

        Pointcharge information is read expecting [format] in units of [unit]. This toggles pointcharge embedding for all defined jobs.
        Data is read, as from a respective external file, in between the start of the LIST and the next END keyword.

      STDGEOM

        After the data has been read succefully, it is transformed in so-called standard geometry.
        That is, its moment if inertia is placed along the y-axis (per definition) and the orthogonal molecular plane is placed on the virtual xz-plane
        In case of pointcharge embedding, all pointhcharges are transformed alongside the geometry.
        In case of present LATTICE information, this feature is disabled, as such a transformation would totally screw up the geometry inside any crystal.

      POINTGROUP

        This keyword toggles four consecutive features:
        First of all, the brute-force-group algorithm tries to detect the geometry's pointgroup (BFG, developed in the cource of the author's PhD thesis).
        Second, an atomic permutation table is created on success.
        Third, shell and basis function permutation tables are created as a consequence.
        Fourth, the combined symmetry information is employed during integral evaluation.
        In case of pointcharge embedding, all pointhcharges are neglected during the detection step!
        In case of present LATTICE information, this algorithm is applied only to the unit-cell geometry.

      NOT YET IMPLEMENTED:

        Detection yes, but no symmetry adapted basis and no symetry in integral evaluation. This feature is on the list of to-be-implemented algorithms.

      LATTICE

      line-wise data

      END

      go to table of contents

        The LATTICE chunk is an extension to the geometry chunk to set up crystal structures.

        Its space is spanned via three lattice vectors a, b, and c with three compontens each (x,y,z) (i.e., a triclynic vector system).

        The LATTICE can be defined as follows, but if the geometry was read from a CIF file is created automatically and must not be specified any further.

        UNIT [unit]

          Sets the unit of the vectors. Default is BOHR.
          [unit] description
          BOHR atomic unit, i.e., Bohr
          ANGSTROEM Angstroem, i.e., 100 pm

        A/B/C ...

          All three vectors are mandatory input!
          A[a.x][a.y][a.z][number of replica in A >= 1][1/0 for periodic boundary conditions in A yes/no]
          B[b.x][b.y][b.z][number of replica in B >= 1][1/0 for periodic boundary conditions in B yes/no]
          C[c.x][c.y][c.z][number of replica in C >= 1][1/0 for periodic boundary conditions in C yes/no]
        The 5th column is used to create replica of the geometry along the vectors (a minimum of 1 per vector, respectively).

        All replicae are then fused to result in the so-called unit-cell of the crystal system.

        The 6th column specifies if periodic boundary condtions are to be employed in this axis.

        If requested, the actual type and parameters for PBC are to be set in the respective sub-chunk!

        As especially inorganic chemists rather prefer to have their crystal data given in vector lengths (a,b,c) and angles (alpha, beta, gamma), a second type of vector input is provided, more in concern with the common cif-file format. The vector lengths are supposed to be in the specified UNIT, whereas the angles are defined to be in degree:

          A[a][alpha in degree][number of replica in A >= 1][1/0 for periodic boundary conditions in A yes/no]
          B[b][beta in degree] [number of replica in B >= 1][1/0 for periodic boundary conditions in B yes/no]
          C[c][gamma in degree][number of replica in C >= 1][1/0 for periodic boundary conditions in C yes/no]
        NOT YET IMPLEMENTED, BUT UPCOMING:

        PBC

        line-wise data

        END

        go to table of contents

          The PBC chunk toggles and contains information on globally defined periodic boundary conditions.

    BASIS

    line-wise data

    END

    go to table of contents

        QuMuLuS++ sets up the basis set in three consecutive steps:
          1st, the basis set is read for elements and/or individual atoms from a specified source.

          2nd, this basis set information is combined with the actual geometry. If any problems occur the program will now terminate.

          3rd, the basis set is transformed into internal shellpair lists, as a first step to achieve linear wall time scaling behavior with respect to the system's size.

        The basis set data is read and applied via three basic sub-chunks: FILE, DATA and/or LIST:

        FILE

          Data is read line-wise, in between the start of the FILE and the next END keyword.
          Each per-line entry is [selection] [type] [name] [format] [filename].

          [selection]

          The complete selection is wrapped in curley brackets {}. Per-element selection is done via case-insensitive symbolic selection, according to the periodic table of elements. A special command in this selection context is ALL, to assign the specified basis set to all atoms, respectively. 1-based indices, and/or 1-based index-ranges can be used to select individual atomic indices. Range selections are sorted ascending, so e.g. 1-7 is the same as 7-1. For index and range selections, exclusions are marked by a prefix exclamation mark. Arbitrary combinations of element symbols, 1-based indices, and/or 1-based index-ranges are delimted by commata.

          Some examples: {all,!1-7}, {1-99,!15-8,!32-77}, {C,!6}

          NOT YET IMPLEMENTED:

            Index based selection is not yet implemented.

          [type]

          One of AO (default basis), AUX (auxiliary basis), AUX-J (auxiliary basis for J), AUX-K (auxiliary basis for K) or ECP (effective core potential).

          [name]

          The name of the basis set to be searched for (e.g. def2-SVP).

          [format]

          One of TURBOMOLE or GAUSSIAN.

          [filename]

          The either absolute or relative path to the basis set file to be read.
          Relative paths with respect to the program call directory, not the input file location.

        DATA

          Data is read line-wise, in between the start of the DATA and the next END keyword.
          In contrast to the FILE type selection, basis sets are read from the internal database, respectively.
          Each per-line entry is [selection] [type] [name] [format].

        LIST

          Data is read line-wise, in between the start of the LIST and the next END keyword.
          In contrast to the FILE type selection, basis sets are read from an arbitrary number of listed DATA chunks.
          Each per-line entry is either [selection] [type] [name] [format] or DATA {...} END.

          DATA

          basis set data, as it would be in an external file ("copy paste")

          END

        Besides the basis set assignment, other optional keywords can be used to modify the basis set creation:

        PURE ([list of moments])

          Toggles pure basis functions (i.e., 5d, 7f, ...). QuMuLuS++ employs cartesian basis functions by default.
          If no arguments are passed, all moments are set to be pure. Optionally, a space-delimited list may address individual moments.
          This enables the usage of Pople basis sets, that are defined to use combinations of cartesian and pure basis moments.

        PURE-AUX ([list of moments])

          Toggles pure auxiliary basis functions (i.e., 5d, 7f, ...).

        PURE-AUX-J ([list of moments])

          Toggles pure auxiliary basis functions (i.e., 5d, 7f, ...) for Coulomb fitting.

        PURE-AUX-K ([list of moments])

          Toggles pure auxiliary basis functions (i.e., 5d, 7f, ...) for exchange fitting.

        GTO-OVERLAP-THRESHOLD [threshold]

          This [threshold] influences the screening of significant Gaussian primitive pairs in the 3rd basis set setup step. Default is 1.e-14.

        SHELLPAIR-OVERLAP-THRESHOLD [threshold]

          This [threshold] influences the screening of significant shellpairs in the 3rd basis set setup step. Default is 1.e-7.

    JOBS

    line-wise data

    END

    go to table of contents

      Individual jobs can be defined as sub-chunks to the JOBS chunk.

      The job execution pipeline will be executed consecutively: later jobs may rely on previously converged.

      Common:

        Common keywords can be used inside any of the job specific chunk sections listed below.

        They control common algorithms, such as the SCF iteration, post-SCF analysis and derivative evaluation.

      go to table of contents

        metric assembly

          METRIC-TYPE [type]

            The metric X is used to orthogonalize the set of equations (i.e., the Fock matrix) during the SCF iteration.

            This keyword toggles the type of assembly of this matrix:

              [type] description
              GRAM-SCHMIDT S -> U*s*UT; X = U*s^(-1/2)*UT
              CHOLESKY S -> LT*L; X = L^(-1)
            Default is CHOLESKY, but if this fails, the program falls back to GRAM-SCHMIDT.

            If both algorithms fail, because of singularities in the basis set setup, the program terminates with a respective error message.

        go to common topics

        initial density guess

          GUESS-TYPE [type]

            Specifies the [type] for the initial density guess to the SCF iteration:

              [type] description
              CORE simple core-Hamiltonian (F = Hc) guess, i.e., the initial density is set to zero.; (default)
              EHT extended Hueckel theory guess (J. Comput. Chem., 39(1963)/40(1964), 1397/2474), quite good for organic species
              HARRIS Harris-like (J. Comput. Chem., 25(2004), 926) initial guess: F(r,s) = Hc(r,s) + 2.0*(rs|rs) - (rr|ss); quite good for metallic species
              SAD superposition of atomic densities (J. Comput. Chem., 27(2006), 637), computed on-the-fly with UHF
              READ read the density matrix from a previously converged run

          EHT-TYPE [type]

            If EHT is chosen as intitial guess type, the parameter set (J. Am. Chem. Soc., 100(1978), 3686) can be chosen via this keyword:

              [type] description
              WOLFSBERG-HELMHOLZ (original) parameterization set proposed by Wolfsberg and Helmholz
              BALLHAUSEN-GRAY parameterization set proposed by Ballhausen and Gray
              CUSACHS parameterization set proposed by Cusachs; (default)
              AMMETER parameterization set proposed by Ammeter et al.

          ORBITAL-MIX [type]

            In case of unrestricted computations, right after the guess step, HOMO and LUMO orbitals are mixed:

            HOMO(alpha) = (HOMO(alpha) + LUMO(alpha)) / sqrt(2)

            HOMO(beta) = (HOMO(beta) - LUMO(beta)) / sqrt(2)

            This is mandatory to converge single excitation states towards the 'real' solution. No mixing reproduces restricted calculations.

              [type] description
              NONE no orbital mixing; (default)
              A mixing of the alpha spin orbitals
              B mixing of the beta spin orbitals
              AB mixing of both spin orbitals
            Here is an examplary H-H PES-scan to depict the impact of orbital mixing:
            Please note that the 'real' HF solution is the blue curve, as the energy of two individual hydrogen atoms lost in space is about -1 hartree...

        go to common topics

        SCF general

          ITER-MIN [number]

            Sets the minimum [number] of SCF iteration steps. Default is 1.

          ITER-MAX [number]

            Sets the maximum [number] of SCF iteration steps. Default is 30.

        go to common topics

        SCF convergence type and threshold

          CONV-TYPE [type]

            Specifies the [type] of the SCF convergence criterion:

            In addition to this setting, the change in the one-electron energy is considered, as well as the change of the change of density matrix elements.

              [type] description
              DENSITY the highest absolute value in the difference of the density matrix and the previous; (default)
              ENERGY the highest absolute value in the difference of the MO energy vector and the previous
              GRADIENT the highest absolute value in the commutator norm [FPS - SPF]

          CONV-THRESHOLD [threshold]

            Specifies the [threshold] for the specified [type] of the SCF convergence criterion. Default thresholds are:

              [type] threshold
              DENSITY 1.e-08
              ENERGY 1.e-12
              GRADIENT 1.e-10

        go to common topics

        SCF convergence accelerations

          Several convergence aids and accelerations may be employed as follows:

          CG-DMS

          QuMuLuS++ performs a conjugate gradient density matrix search (CG-DMS) (J. Chem. Phys., 106(1997), 5569) in each SCF iteration cycle, in additon to one of the convergence aids defined as follows.

          CONV-AID [type]

              [type] description
              DIIS Direct Inversion of the Iterative Subspace (DIIS)
              ODA Optimal Damping Algorithm (ODA) (Int. J. Quant. Chem., 79(2000), 82)
              ODA-DIIS ODA and DIIS combined
              BROYDEN Broyden update scheme (J. Comput. Phys., 134(2011), 134109)
              DGTR Density-based globally convergent trust-region method (DGTR) (J. Math. Chem., 40(2006):4, 349)
          If either DIIS or ODA-DIIS was chosen as SCF convergence acceleration, the following set of keywords further specify the DIIS setup:

          DIIS-MAXSUBSPACE [dimension]

            Sets the maximum [dimension] of the iterative subspace. Default is 6.

          DIIS-TYPE [type]

            In addition to this (default is not to), this keyword can be used to add an additional SCF convergence acceleration:

              [type] description
              P-DIIS Pulay's DIIS formulation (J. Comput. Chem., 3(1982), 556)
              E-DIIS Int. J. Quant. Chem., 79(2000), 82 (RCA method)
              A-DIIS J. Chem. Phys., 132(2010), 054109 (RCA method)
              AP-DIIS A-DIIS + P-DIIS (in this sequence)
              EP-DIIS E-DIIS + P-DIIS (in this sequence)
              AE-DIIS A-DIIS + E-DIIS (in this sequence)
              AEP-DIIS A-DIIS + E-DIIS + P-DIIS (in this sequence)

          DIIS-MODE [mode]

            This is a special keyword for the P-DIIS formulation:

              [mode] description
              C1 Pulay's original formulation (J. Comput. Chem., 3(1982), 556); (default)
              C2 Seller's variation (Int. J. Quant. Chem., 45(1993), 31)

        go to common topics

        SCF Aufbau principle

          PFON

            Fractional occupation numbers (J. Chem. Phys., 110(1999), 695) for restricted methods. Turbomole's dscf calls it Fermi-smearing.

          DAMPING [factor]

            User defined density matrix damping: P'_k = (1 - [factor]) P_k + [factor] P_(k-1)

            Default is to use automated damping, based on energy minimization (line search).

          LEVEL-SHIFT [factor]

            User defined initial level shift factor.

          LEVEL-SHIFT-DEC [factor]

            User defined static cyclic decrement of the level shift factor.

          LEVEL-SHIFT-MIN [factor]

            User defined static minimum of the level shift factor.

        go to common topics

        SCF density purification

          DMP-TYPE

            This keyword defines the type of density purification applied after the Aufbau:
              [type] description
              McW McWeeney purification
              SP2 SP2-SCF
              SP4 SP4-SCF
              TRS4 TRS4-SCF; default
              PM1 PM1-SCF
              PM2 PM2-SCF

        go to common topics

        electron integrals and derivatives

          CHARGE-BATCHSIZE [batchsize]

            Sets a [batchsize] for the evaluation of one-electron Coulomb type integrals with respect to charges/cores.

            It clusters pointcharges to be transformed in one integral batch call. Default is to use up to 1000 charges/cores at once.

          SHELL-PAIR-BATCHSIZE [batchsize]

            Sets a [batchsize] for the evaluation of two-electron repulsion integrals with respect to shellpairs.

            It clusters shellpairs to be transformed in one integral batch call. Default is to use up to 1000 shellpairs at once.

          WOLF-DAMPING [factor]

            Toggles Wolf-type damping of external pointcharges (J. Chem. Phys., 110(1999), 8254) and sets the damping factor. Disabled by default.

          ERI-EVALUATION [type]

              [type] description
              STD-JK-AUTO analytic evaluation of both J and K type ERIs; QuMuLuS++ determines how...
              STD-JK-TOGETHER analytic evaluation of both J and K type ERIs; Combined evaluation.
              STD-JK-SPLIT analytic evaluation of both J and K type ERIs; Separate evaluation. default
              RI-J density fitting for J type ERIs, demands auxiliary basis for Coulomb fitting
              RI-K density fitting for K type ERIs, demands auxiliary basis for exchange fitting
              RI-JK density fitting for both J and K type ERIs
            Density fitting (DF) (formerly known as resolution of identity (RI)) has to be toggled explicitly, in addition to an AUX basis set, respectively.
            Please mind that in QuMuLuS++ Coulomb type RI/DF is explcitly multipole accelerated (MARI-J)!

          various integral thresholds: [keyword] [threshold]

              [keyword] description
              ERI-THRESHOLD-0 density adapted Schwarz screening threshold for J and K type ERIs. Default is 1.e-14.
              ERI-THRESHOLD-J density adapted Schwarz screening threshold for J type ERIs. Default is 1.e-14.
              ERI-THRESHOLD-K density adapted Schwarz screening threshold for K type ERIs. Default is 1.e-14.
              ERI-THRESHOLD-1 density adapted Schwarz screening threshold for two-electron 1st derivatives. Default is 1.e-10.
              ERI-THRESHOLD-2 density adapted Schwarz screening threshold for two-electron 2nd derivatives. Default is 1.e-10.

        go to common topics

        QQR screening

          QQR screening (J. Chem. Phys., 136(2012), 144107) of two-electron type integrals is an extension of the default Schwarz screening with respect to spatial aspects. Please mind though, that according to the author himself (Simon A. Maurer), this is "not a rigorous, but a highly approximative attempt". This is found to be a very powerful method to amplify linear scaling in local methods, but might easily screw up your computation as well... Sometimes you have to gamble a bit with the internal threshold, in order to make it work at its best...

          QQR

            Toggles the QQR screening. Default is not to use it.

          QQR-THRESHOLD [threshold]

            Sets the QQR internal threshold for screening. Default is 1.e-12.

        go to common topics

        post-SCF

          The post-SCF phase is basically divided into population analysis attempts, plotting and/or property computation.

          Only monopole analysis is default, all others have to be toggled explicily, by simply adding the [keyword]:

          simple analysis

              [keyword] description
              MOS plots the MO energies
              NPA plots the MO and AO occupation numbers

          population analysis

              [keyword] description
              MULLIKEN Mulliken population analysis
              MULLIKEN-E Mulliken electron population analysis
              LOEWDIN Loewdin population analysis
              LOEWDIN-E Loewdin electron population analysis
              BADER Bader population analysis (Comput. Mater. Sci., 36(2006), 254 and others)
              HIRSCHFELD Hirschfeld population analysis (J. Chem. Phys., 126(2007), 144111)
              STOCKHOLDER Stockholder population analysis (J. Chem. Phys., 131(2009), 144101)
              VORONOI Voronoi population analysis (J. Comp. Chem., 25(2004), 189)
              MAYER Mayer bond order analysis (Chem. Phys. Lett., 97(1983), 270)
              RESP Resp population analysis

          multipole analysis

              [keyword] description
              D-POLES dipole analysis
              Q-POLES quadrupole analysis
              O-POLES octupole analysis

          basis set analysis

              [keyword] description
              COMPLETENESS-PROFILE one-electron basis set completeness profiles weighted with MO coefficients

          spatial plots and spatial plot preparation

              [keyword] description
              MOLDEN prepares a MOLDEN input file from a converged SCF run
              EMD-CUBE plots the EMD profile of a converged SCF run into a cube file
              ESP-CUBE plots the electrostatic potential of a converged SCF run into a cube file
              MOS-CUBE [list] plots a [list] (space delimited 1-based indices) of MOs of a converged SCF run into cube files
              NOS-CUBE [list] plots natural orbitals (NOs) of a converged SCF run into cube files

          explicit derivatives

              [keyword] description
              GRADIENTS 1st cartesian derivatives with respect to atomic positions
              HESSIANS 2nd cartesian derivatives with respect to atomic positions
          NOT YET IMPLEMENTED:

          frequencies and normal modes

              [keyword] description
              FREQUENCIES computation of frequencies
              NORMAL-MODES computation of normal-modes (only in addition to frequencies)

        go to common topics

        geometry optimization

          OPTIMIZE

            Toggles the geometry optimization / minimization wrapper algorithm around the employed method.

          OPT-ITER-MIN [number]

            Sets the minimum [number] of geometry minimization steps. Default is 1.

          OPT-ITER-MAX [number]

            Sets the maximum [number] of geometry minimization steps. Default is 30.

          OPT-THRESHOLD-ENERGY [threshold]

            Sets the [threshold] for the change in two consecutive converged SCF energies. Default is 1.e-8.

          OPT-THRESHOLD-RMSD [threshold]

            Sets the [threshold] for the consecutive root-mean-square-distance change of the geometry. Default is 1.e-6.

          MINIMIZER [method]

              [method] description
              GRADIENT-DESCENT simple gradient following
              FLETCHER-REEVES conjugate gradient with Fletcher-Reeves update
              POLAK-RIBIERE conjugate gradient with Polak-Ribiere update
              MURTAGH-SARGENT quasi-Newton method with Murthagh-Sargent update
              MURTAGH-SARGENT-POWELL quasi-Newton method with Murthagh-Sargent-Powell update
              DAVIDSON-FLETCHER-POWELL quasi-Newton method with Davidson-Fletcher-Powell update
              BROYDEN-FLETCHER-GOLDFARB-SHANNO quasi-Newton method according to BFGS, but with the more stable Bofill update; default

        go to common topics

        3rd party tools used by QuMuLuS++ during the computation

          G-CP [basis set name]

            Stefan Grimme's generalized counterpoise correction. See it's manual for valid values for [basis set name].

          DFT-D2 (['-func' argument])

            Stefan Grimme's D-2 corrections to HF and DFT. The function argument is optional. If not specified, QuMuLuS++ will make an educated guess.

          DFT-D3 (['-func' argument])

            Stefan Grimme's D-3 corrections to HF and DFT. The function argument is optional. If not specified, QuMuLuS++ will make an educated guess.

          HF-3C (['-func' argument])

            Stefan Grimme's HF/DFT-3c [J. Comput. Chem. 34(2013) p.1672]. The '3c' denotes three corrections, i.e., a D3 correction, a gCP correction and an additional SRB correction, with the 'MINIX' basis set. QuMuLuS++ computes the first two contributions with the respective 3rd-party tools. The third term is computed internally, based on DFT-D3 parameters. Mind that HF-3c actually is not a method extension, but a method on its own, based on the 'MINIX' basis set! If you want to use this method with QuMuLuS++, therefore use "HF/DFT method + MINIX basis + HF-3c keyword". As mentioned, all corrections made are with the 'MINIX' basis set.
            Composition of the MINIX basis set.

              [element] basis
              H-He, B-Ne MINIS
              Li-Be MINIS+1(p)
              Na-Mg MINIS+1(p)
              Al-Ar MINIS+1(d)
              K-Zn SV
              Ga-Kr SVP
              Rb-Xe def2-SV(P) with ECP

          These tools are employed dynamically by QuMuLuS++ after the energy and derivative steps, respectively.

          The computed correction terms are printed out explicitly, and the original values (energy, derivatives) are updated accordingly.

        go to common topics

        gimmicks and development (DO NOT USE - or do not complain!)

          ERILISTER

            This keyword toggles the listing of electron repulsion integrals. This is of no need and, therefore, disabled by default.

        go to common topics


      HF [HF-TYPE]

      line-wise data

      END

      go to table of contents

        The HF chunk toggles and contains information on a Hartree-Fock job.

        [HF-TYPE]description
        RHF restricted Hartree-Fock, i.e., Roothan-Hall
        ROHF restricted open shell Hartree-Fock
        UHF unrestricted Hartree-Fock, i.e., Pople-Nesbet
        none QuMuLuS++ will decide, with its preference to RHF
        All common keywords can be used.

      DFT [HF-TYPE] [FUNCTIONAL]

      line-wise data

      END

      go to table of contents

        The DFT chunk toggles and contains information on a density functional job, employing the LibXC library (www.tddft.org).

        [HF-TYPE]description
        RHF restricted Hartree-Fock, i.e., Roothan-Hall
        ROHF restricted open shell Hartree-Fock
        UHF unrestricted Hartree-Fock, i.e., Pople-Nesbet
        [FUNCTIONAL]description
        [functional]setup via a predefined set of LibXC functionals, selected via a 'common' name (XALPHA, B3LYP, ...)
        XCK explicit assembly via exchange (X), correlation (C), and optional kinetic (K) LibXC functionals (xc_funcs.h, LIBXC_FUNCS.h)
          [functional] X C XC
          BLYP XC_HYB_GGA_XC_B1LYP XC_HYB_GGA_XC_B1LYP yes
          B3LYP XC_HYB_GGA_XC_B3LYP XC_HYB_GGA_XC_B3LYP yes
          B3LYP-CAM XC_HYB_GGA_XC_CAM_B3LYP XC_HYB_GGA_XC_CAM_B3LYP yes
          BP86 XC_HYB_GGA_XC_B3P86 XC_HYB_GGA_XC_B3P86 yes
          GOMBAS XC_LDA_X XC_LDA_C_GOMBAS no
          PBE XC_GGA_X_PBE XC_GGA_C_PBE no
          PBE0 XC_HYB_GGA_XC_PBEH XC_HYB_GGA_XC_PBEH yes
          PBE_SOL XC_GGA_X_PBE_SOL XC_GGA_C_PBE_SOL no
          TETER93 XC_LDA_XC_TETER93 XC_LDA_XC_TETER93 yes
          TPSS XC_MGGA_X_TPSS XC_MGGA_C_TPSS no
          TPSSH XC_HYB_MGGA_XC_TPSSH XC_HYB_MGGA_XC_TPSSH yes
          WIGNER XC_LDA_X XC_LDA_C_XALPHA no
          XALPHA XC_LDA_X XC_LDA_C_WIGNER no
          XLYP XC_GGA_XC_XLYP XC_GGA_XC_XLYP yes
        All common keywords can be used. Yet, some of them are not supported, such as derivatives and all related.

        Method specific keywords in this context are as follows:

        X-FUNC [name]

          If [FUNCTIONAL] was set to XCK, explicitly define an exchange type or combined echange-correlation type LibXC functional (xc_funcs.h, LIBXC_FUNCS.h).

        C-FUNC [name]

          If [FUNCTIONAL] was set to XCK, explicitly define a correlation type or combined echange-correlation type LibXC functional (xc_funcs.h, LIBXC_FUNCS.h).

        K-FUNC [name]

          If [FUNCTIONAL] was set to XCK, explicitly define a kinetic type LibXC functional (xc_funcs.h, LIBXC_FUNCS.h).

        GRID-TYPE-ANGULAR [type]

            [type] description / reference
            LOBATTO O. Treutler and R. Ahlrichs, J. Chem. Phys., 102(1994), 346
            LEBEDEV V.I. Lebedev and D.N. Laikov, Doklady Mathematics, 59(1999), 477; default

        GRID-TYPE-RADIAL [type]

            [type] reference
            EM ([m : 2,3,4]) C. W. Murray and N. C. Handy and G. J. Laming, Mol. Phys., 78(1993), 997; Euler-McLaurin with optional m (default m=2)
            M1 O. Treutler and R. Ahlrichs, J. Chem. Phys., 102(1994), 346; M1 grid
            M2 O. Treutler and R. Ahlrichs, J. Chem. Phys., 102(1994), 346; M2 grid
            M3 O. Treutler and R. Ahlrichs, J. Chem. Phys., 102(1994), 346; M3 grid
            M4 O. Treutler and R. Ahlrichs, J. Chem. Phys., 102(1994), 346; M4 grid
            LOG2 M. E. Mura and P. J. Knowles, J. Chem. Phys., 104(1996), 9848; LOG2 grid
            LOG3 M. E. Mura and P. J. Knowles, J. Chem. Phys., 104(1996), 9848; LOG3 grid; default
            LOG4 M. E. Mura and P. J. Knowles, J. Chem. Phys., 104(1996), 9848; LOG4 grid

        GRID-NPOINTS-ANGULAR [value]

          Demand a minimum of angular gridpoints, independent of the grid type. Default is: 80.

        GRID-NPOINTS-RADIAL [value]

          Demand a minimum of radial gridpoints, independent of the grid type. Default is: 300.

        GRID-WEIGHTS [type]

            [type] description / reference
            BECKE grid parameters according to Becke
            STRATMANN grid parameters according to Stratmann et al.; default

        GRID-PRUNE-TOLERANCE [threshold]

          Pruning threshold for the grid, after Becke weights have been applied. Default is quite tight: 1.e-14.
        NOT YET IMPLEMENTED:
        analytical 1st and 2nd derivatives
        no mGGA functionals: meta-GGA is not yet officially supportet by LibXC
        no kinetic functionals: not yet officially supportet by LibXC, except for LDA
        RPA functionals
        Coulomb damped functionals
        omega functionals
        explicitly correlated functionals
        double hybrid functionals

      MP [HF-TYPE] [MP-TYPE] ([DFT-TYPE])

      line-wise data

      END

      go to table of contents

        The MP chunk toggles and contains information on a Moeller-Plesset job.

        [HF-TYPE]description
        RHF restricted Hartree-Fock, i.e., Roothan-Hall
        ROHF restricted open shell Hartree-Fock
        UHF unrestricted Hartree-Fock, i.e., Pople-Nesbet
        [MP-TYPE]description
        MP2 Moeller-Plesset pertubation theory of order 2
        [DFT-TYPE]this argument is optional, and the very same as the [FUNCTIONAL] keyword in a DFT setup.
        All common keywords can be used. Yet, some of them are not supported, such as derivatives and all related.

        Method specific keywords in this context are as follows:

        AUTO-FREEZE

          Disabled by default.

          Occupied orbitals below a threshold energy of AUTO-FREEZE-THRESHOLD-OCC hartree are frozen.

          Virtual orbitals above a threshold energy of AUTO-FREEZE-THRESHOLD-VIRT hartree are frozen.

        AUTO-FREEZE-THRESHOLD-OCC [threshold]

          Occupied orbitals with energies lower than this threshold energy are frozen, if AUTO-FREEZE is toggled. Default is -20 hartree.

        AUTO-FREEZE-THRESHOLD-VIRT [threshold]

          Virtual orbitals with energies higher than this threshold energy are frozen, if AUTO-FREEZE is toggled. Default is +40 hartree.

        FREEZE [index numbers]

          Explicitly freezes the listed orbitals (1-based indices), delimited by spaces (e.g. "1 2 3 4"), independent of their energies.

        AO-CDD-MP

          Perform an AO-CDD-MPN computation instead of the default MO-MPN.

        RI-MP

          Employ resolution-of-identity / density fitting instead of the analytical integral evaluation. Demands a general purpose auxiliary basis (i. e., not Coulomb, or exchange basis).

        LAPLACE-POINTS [number]

          The number of Laplace points for AO-MP type computations. Default is 7.

      RPA [HF-TYPE] ([DFT-TYPE])

      line-wise data

      END

      go to table of contents

        The RPA chunk toggles and contains information on a Random-Phase-Approximation job.

        [HF-TYPE]description
        RHF restricted Hartree-Fock, i.e., Roothan-Hall
        ROHF restricted open shell Hartree-Fock
        UHF unrestricted Hartree-Fock, i.e., Pople-Nesbet
        [DFT-TYPE]this argument is optional, and the very same as the [FUNCTIONAL] keyword in a DFT setup.
        All common keywords can be used. Yet, some of them are not supported, such as derivatives and all related.

        Method specific keywords in this context are as follows:

        AUTO-FREEZE

          Disabled by default.

          Occupied orbitals below a threshold energy of AUTO-FREEZE-THRESHOLD-OCC hartree are frozen.

          Virtual orbitals above a threshold energy of AUTO-FREEZE-THRESHOLD-VIRT hartree are frozen.

        AUTO-FREEZE-THRESHOLD-OCC [threshold]

          Occupied orbitals with energies lower than this threshold energy are frozen, if AUTO-FREEZE is toggled. Default is -20 hartree.

        AUTO-FREEZE-THRESHOLD-VIRT [threshold]

          Virtual orbitals with energies higher than this threshold energy are frozen, if AUTO-FREEZE is toggled. Default is +40 hartree.

        FREEZE [index numbers]

          Explicitly freezes the listed orbitals (1-based indices), delimited by spaces (e.g. "1 2 3 4"), independent of their energies.

        AO-CDD-RPA

          Perform an AO-CDD-RPA computation instead of the default MO-RPA.

        RI-RPA

          Employ resolution-of-identity / density fitting instead of the analytical integral evaluation. Demands a general purpose auxiliary basis (i. e., not Coulomb, or exchange basis).

        LAPLACE-POINTS [number]

          The number of Laplace points for AO-RPA type computations. Default is 7.

      CI [CI-TYPE] ([DFT-TYPE])

      line-wise data

      END

      go to table of contents

        The CI chunk toggles and contains information on a Configuration Interaction job.

        [CI-TYPE]description
        CI-S single occupations
        CI-SD single, and double occupations
        [DFT-TYPE]this argument is optional, and the very same as the [FUNCTIONAL] keyword in a DFT setup.

        NATURAL-ORBITALS

          Compute and use natural orbitals with CI. This is disabled by default, as it is an approximation, however, this usually speeds up the computation to quite an amount.

        All common keywords can be used.

    PCM

    line-wise data

    END

    go to table of contents

      The PCM chunk toggles and contains information on a globally defined polarizable continuum model.

      Its information is considerd to be as global as the geometry and basis set information, and equally controls all jobs.

      CAVITY [type]

        Toggles the [type] of tabulated atomic VdW radii used for the cavity construction:
          [type] reference
          UFF Uniform Force Field paramaters, according to J. Am. Soc., 114(1992), 10024
          BONDI VdW radii proposed by Bondi in J. Phys. Chem., 68(1964), 441
          LEBEDEV Don't use any VdW radii, but employ Lebedev spheres. Q-Chem does that and it's fine; (default)

      RESOLUTION [value]

        The resolution parameter scales the number of tesserae, and is supposed to be in mathematical percent [0,1]. Default is 0.2. Not for LEBEDEV.

      HYDROGENS [1/0 for include/exclude]

        This keyword toggles the exclusion of Hydrogens from the cavity construction. Default is 1, i.e., Hydrogens are included.

      MODEL [type]

        Toggles the actual PCM method
          [type] description
          D-PCM the original formulation of the dielectric-PCM
          C-PCM the conductor-like PCM (i.e., the COSMO model); default
          IEF-PCM the (isotropic) integral equation formalism PCM
          SS(V)PE the surface and volume polarization for electrostatic model

      SOLVENT [solvent name]

        Specifies the PCM solvent, i.e., the epsilon value via an internal database. Default is 'water'.
          [solvent name] epsilon
          1,1,1-trichloroethane 7.0826
          1,1,2-trichloroethane 7.1937
          1,2,4-trimethylbenzene 2.3653
          1,2-dibromoethane 4.9313
          1,2-dichloroethane 10.125
          1,2-ethanediol 40.245
          1,4-dioxane 2.2099
          1-bromo-2-methylpropane 7.7792
          1-bromooctane 5.0244
          1-bromopentane 6.269
          1-bromopropane 8.0496
          1-butanol 17.332
          1-chlorohexane 5.9491
          1-chloropentane 6.5022
          1-chloropropane 8.3548
          1-decanol 7.5305
          1-fluorooctane 3.89
          1-heptanol 11.321
          1-hexanol 12.51
          1-hexene 2.0717
          1-hexyne 2.615
          1-iodobutane 6.173
          1-iodohexadecane 3.5338
          1-iodopentane 5.6973
          1-iodopropane 6.9626
          1-nitropropane 23.73
          1-nonanol 8.5991
          1-octanol 9.8629
          1-pentanol 15.13
          1-pentene 1.9905
          1-propanol 20.524
          2,2,2-trifluoroethanol 26.726
          2,2,4-trimethylpentan 1.9358
          2,4-dimethylpentane 1.8939
          2,4-dimethylpyridine 9.4176
          2,6-dimethylpyridine 7.1735
          2-bromopropane 9.3610
          2-butanol 15.944
          2-chlorobutane 8.3930
          2-heptanone 11.658
          2-hexanone 14.136
          2-methoxyethanol 17.2
          2-methyl-1-propanol 16.777
          2-methyl-2-propanol 12.47
          2-methylpentane 1.89
          2-methylpyridine 9.9533
          2-nitropropane 25.654
          2-octanone 9.4678
          2-pentanone 15.200
          2-propanol 19.264
          2-propen-1-ol 19.011
          e-2-pentene 2.051
          3-methylpyridine 11.645
          3-pentanone 16.78
          4-heptanone 12.257
          4-methyl-2-pentanone 12.887
          4-methylpyridine 11.957
          5-nonanone 10.6
          acetic_acid 6.2528
          acetone 20.493
          acetonitrile 35.688
          acetophenone 17.44
          aniline 6.8882
          anisole 4.2247
          benzaldehyde 18.220
          benzene 2.2706
          benzonitrile 25.592
          benzyl_alcohol 12.457
          bromobenzene 5.3954
          bromoethane 9.01
          bromoform 4.2488
          butanal 13.45
          butanoic_acid 2.9931
          butanone 18.246
          butanonitrile 24.291
          butyl_ethanoate 4.9941
          butylamine 4.6178
          n-butylbenzene 2.36
          sec-butylbenzene 2.3446
          tert-butylbenzene 2.3447
          carbon_disulfide 2.6105
          carbon_tetrachloride 2.2280
          chlorobenzene 5.6968
          chloroform 4.7113
          alpha-chlorotoluene 6.7175
          o-chlorotoluene 4.6331
          m-cresol 12.44
          o-cresol 6.76
          cyclohexane 2.0165
          cyclohexanone 15.619
          cyclopentane 1.9608
          cyclopentanol 16.989
          cyclopentanone 13.58
          decalin_(cis/trans_mixture) 2.196
          cis-decalin 2.2139
          n-decane 1.9846
          dibromomethane 7.2273
          dibutylether 3.0473
          o-dichlorobenzene 9.9949
          e-1,2-dichloroethene 2.14
          z-1,2-dichloroethene 9.2
          dichloromethane 8.93
          diethyl_ether 4.2400
          diethyl_sulfide 5.723
          diethylamine 3.5766
          diiodomethane 5.32
          diisopropyl_ether 3.38
          cis-1,2-dimethylcyclohexane 2.06
          dimethyl_disulfide 9.6
          n,n-dimethylacetamide 37.781
          n,n-dimethylformamide 37.219
          dimethylsulfoxide 46.826
          diphenylether 3.73
          dipropylamine 2.9112
          n-dodecane 2.0060
          ethanethiol 6.667
          ethanol 24.852
          ethyl_ethanoate 5.9867
          ethyl_methanoate 8.3310
          ethyl_phenyl_ether 4.1797
          ethylbenzene 2.4339
          fluorobenzene 5.42
          formamide 108.94
          formic_acid 51.1
          n-heptane 1.9113
          n-hexadecane 2.0402
          n-hexane 1.8819
          hexanoic_acid 2.6
          iodobenzene 4.5470
          iodoethane 7.6177
          iodomethane 6.8650
          isopropylbenzene 2.3712
          p-isopropyltoluene 2.2322
          mesitylene 2.2650
          methanol 32.613
          methyl_benzoate 6.7367
          methyl_butanoate 5.5607
          methyl_ethanoate 6.8615
          methyl_methanoate 8.8377
          methyl_propanoate 6.0777
          n-methylaniline 5.9600
          methylcyclohexane 2.024
          n-methylformamide_(e/z mixture) 181.56
          nitrobenzene 34.809
          nitroethane 28.29
          nitromethane 36.562
          o-nitrotoluene 25.669
          n-nonane 1.9605
          n-octane 1.9406
          n-pentadecane 2.0333
          pentanal 10.00
          n-pentane 1.8371
          pentanoic_acid 2.6924
          pentyl_ethanoate 4.7297
          pentylamine 4.2010
          perfluorobenzene 2.029
          propanal 18.5
          propanoic_acid 3.44
          propanonitrile 29.324
          propyl_ethanoate 5.5205
          propylamine 4.9912
          pyridine 12.978
          tetrachloroethene 2.268
          tetrahydrofuran 7.4257
          tetrahydrothiophene-s,s-dioxide 43.962
          tetralin 2.771
          thiophene 2.7270
          thiophenol 4.2728
          toluene 2.3741
          trans-decalin 2.1781
          tributylphosphate 8.1781
          trichloroethene 3.422
          triethylamine 2.3832
          n-undecane 1.9910
          water 78.355
          xylene_(mixture) 2.3879
          m-xylene 2.3478
          o-xylene 2.5454
          p-xylene 2.2705

      EPSILON [value]

        In case a user wants to employ a solvent that is not part in the internal database listed above, or simply wants to give the epsilon [value] explicitly, this can be done via this keyword. Default is 78.355 (for water). The solvent's name will be adapted, if epsilon is known, or otherwise set to 'unknown'.

      EPSILON_K [value]

        The PCM models, besides D-PCM, employ a quite empirical scaling function, determinated by comparing COSMO (unscaled) and correct electrostatic solute-solvenyt energies: f(epsilon) = (epsilon - 1) / (epsilon + epsilon_k). In the original COSMO literature, this factor was empirically set to the default value of 0.5. Later on some reports came up that for totally unpolarized systems, the factor should be reduced to zero. It's the user's choice to do so: default is 0.5.

      NOT YET IMPLEMENTED:
        Thermodynamic properties obtained from PCM contributions. This is on the list of to-be-implemented algorithms.
      IMPLEMENTED, BUT NOT YET VALIDATED:
        PCM is implemented on the d0, and d1 levels. d2 level is on the radar. All PCM usage has yet to be validated! The general disclaimer 'no warranty' holds even more in here!

    PES-SCAN

    line-wise data

    END

    go to table of contents

      The PES-SCAN chunk toggles a potential energy surface scan.

      This chunk is purposely set above all JOB chunks, so that any setting can be used to explore the PES!

      E. g. a geometry optimization can be performed at each step!

      The setup is linewise, each line containing infoamration on a bond, angle, or dihedral to be scanned:

      [1st index] [2nd index] ([3rd index]) ([4th index]) [start] [step] [number of steps]

        The indices (1-based indices!) refer to atoms in the defined geometry.
        Bonds are defined with two indices, angles with three and dihedrals with four indices, respectively.
        The [start] and [step] values describe the start point of the variable to be screened, and the incremental step, supposed to be in the defined units.

      UNIT [radial unit] [angular unit]

        Default is BOHR and DEGREE.
        [radial unit] description
        BOHR atomic unit, i.e., Bohr
        ANGSTROEM Angstroem, i.e., 100 pm
        [angular unit] description
        RAD rad
        DEGREE degree

    NEB

    line-wise data

    END

    go to table of contents

      The NEB chunk toggles a VERY SIMPLE nudged elastic band optimization.

      Use with great caution, and expect a lot of corner cutting...

      If this computation is performed, all other gradient computations are skipped!

      Any in-method optimization, gradient computation and PES-scan functionality is neglected!

      NUM-GEOMETRIES [number of geometries]

        The number of to be defined geometries in the NEB optimization.

      GEOMETRY-[1-based index] [source]

      GEOMETRY-[1-based index]-[1-based index] [source]

        Defines one geometry in the NEB setup.
        The first index (1) is the start-point, the last index (NUM-GEOMETRIES) is the end-point
        Two indices can be used to define a range, e.g. "GEOMETRY-2-5" defines the 2nd to 5th geometry.
        [source] description
        GEOMETRY take the geometry defined via the GEOMETRY chunk
        INTERPOLATED interpolate this geometry in between the endpoints
        NEW make this a new geometry; this expects a GEOMETRY child-chunk to come

      INTERPOL-TYPE [type]

        Defines the type of interpolation, if INTERPOLATED wa schosen as [source]. Default is LINEAR.
        [type] description
        LINEAR a linear, equidistant interpolation in between the start- and end-point

      OPTIMIZE-TYPE [type]

        Defines the type of NEB optimization for this nudge in the band. Default is TWO-SIDED..
        [type] description
        NONE do not optimize, i.e. keep this nudge as it is
        STATIC optimize the structure, but do not consider NEB gradients
        LEFT-SIDED optimize the structure, and take into account the "left-hand sided" NEB gradients
        RIGHT-SIDED optimize the structure, and take into account the "right-hand sided" NEB gradients
        TWO-SIDED optimize the structure, and take into account all NEB gradients

      NEB-MIN-ITER [value]

        Sets the minimum [number] of NEB iteration steps. Default is 1.

      NEB-MAX-ITER [value]

        Sets the maximum [number] of NEB iteration steps. Default is 30.

      NEB-THRESHOLD-ENERGY [threshold]

        Sets the [threshold] for the change in two consecutive converged SCF energies. Default is 1.e-8.

      NEB-THRESHOLD-RMSD [threshold]

        Sets the [threshold] for the consecutive root-mean-square-distance change of the geometry. Default is 1.e-6.