You are on page 1of 22

PHYSICAL REVIEW B

VOLUME 58, NUMBER 7

15 AUGUST 1998-I

Relativistic separable dual-space Gaussian pseudopotentials from H to Rn
C. Hartwigsen, S. Goedecker, and J. Hutter
Max-Planck-Institut fu¨r Festko¨rperforschung, D-70569 Stuttgart, Germany
~Received 20 March 1998!
We generalize the concept of separable dual-space Gaussian pseudopotentials to the relativistic case. This
allows us to construct this type of pseudopotential for the whole Periodic Table, and we present a complete
table of pseudopotential parameters for all the elements from H to Rn. The relativistic version of this pseudopotential retains all the advantages of its nonrelativistic version. It is separable by construction, it is optimal for
integration on a real-space grid, it is highly accurate, and, due to its analytic form, it can be specified by a very
small number of parameters. The accuracy of the pseudopotential is illustrated by an extensive series of
molecular calculations. @S0163-1829~98!05628-8#

I. INTRODUCTION

Pseudopotentials are a well-established tool in ab initio
structure calculations of molecules and solids. First, by replacing the atom by a pseudoatom, the number of orbitals
which have to be calculated is reduced, and, second, the size
of the basis set can substantially be reduced because the
pseudo-wave-functions are smoother than their all-electron
counterparts. In addition, relativistic effects which are relevant for heavier elements can be included in the pseudopotential construction, so that a nonrelativistic calculation can
reproduce these.
In 1982, Bachelet, Hamann, and Schlu¨ter1 published a list
of pseudopotentials for all elements up to Pu, that has found
widespread application. There have been many attempts
since to improve the pseudopotential transferability and their
numerical efficiency. One major advance was the introduction of a separable form by Kleinmann and Bylander,2 that
significantly reduces the computational effort for the calculation of the nonlocal part, especially when using a planewave basis set. Gonze, Stumpf, and Scheffler3 investigated
the Kleinmann-Bylander form carefully and computed a list4
of pseudopotentials for many elements up to Xe. Goedecker,
Teter, and Hutter proposed a dual-space Gaussian-type
pseudopotential which is separable and satisfies an optimality criterion for the real-space integration of the nonlocal
part. For large systems there is only a quadratic scaling with
respect to the system size if the integration of the nonlocal
part is performed on a real-space grid, compared to a cubic
scaling if a Fourier space integration is used.6 In contrast to
most other pseudopotential construction methods, the authors
of Ref. 5 also included unoccupied orbitals in their method,
thereby generating highly transferable pseudopotentials.
They gave the nonrelativistic pseudopotential parameters for
the first two rows of the periodic system, and showed that
their pseudopotentials give highly accurate results in molecular calculations. They obtained results which are much closer
to the quasi-exact all-electron LDA ~Refs. 7 and 8! ~localdensity approximation! value than what is obtained in allelectron calculations with a standard Gaussian 6-31G * basis
sets. In other words, the errors due to the pseudopotential
approximation were much smaller than the errors in an allelectron calculation introduced by incomplete basis sets.
0163-1829/98/58~7!/3641~22!/$15.00

PRB 58

In this paper we give the parameters of dual-space Gaussian pseudopotentials for all elements from H to Rn. In contrast to Ref. 5, all pseudopotentials are now generated on the
basis of a fully relativistic all-electron calculation, i.e., by
solving the two-component Dirac equation. The generalization of the norm-conservation property to the relativistic case
proposed by Bachelet and Schlu¨ter9 is used for the construction. We also introduced some slight modifications of the
analytic form of the pseudopotential. The parameters are
given in the context of the local-density approximation. Even
though the parameters change only slightly if the pseudopotential is constructed within the framework of a generalized
gradient approximation10,11 ~GGA! functional, we found that
molecular properties are less accurately described if LDA
pseudopotentials are inserted in a molecular calculation using GGA’s. Since it is not possible to construct pseudopotential tables for all current GGA schemes, a program that
can construct pseudopotentials for the most common GGA’s
can be obtained from the authors.
II. FORM OF THE PSEUDOPOTENTIAL

The local part of the pseudopotential is given by
V loc~ r ! 5

2Z ion
erf
r

F

S D F S DG

3 C 1 1C 2

r

A2r loc

1exp 2

1 r
2 r loc

2

S D S D S DG
2

r

r loc

1C 3

r

r loc

4

1C 4

r

r loc

6

,

~1!

where erf denotes the error function. Z ion is the ionic charge
of the atomic core, i.e., the total charge minus the charge of
the valence electrons. The nonlocal contribution V l (r,r8 ) to
the pseudopotential is a sum of separable terms
3

V l ~ r,r8 ! 5

3

1l

(( (

i51 j51 m52l

* ~ rˆ8 ! ,
Y l,m ~ rˆ! p li ~ r ! h li, j p lj ~ r 8 ! Y l,m
~2!

where Y l,m are the spherical harmonics, and l the angular
momentum quantum number. The projectors p li (r) are Gaussians of the form
3641

© 1998 The American Physical Society

3642

C. HARTWIGSEN, S. GOEDECKER, AND J. HUTTER

A2r l12 ~ i21 ! exp

AS

p li ~ r ! 5
rl

S

2

r2
2r l 2

D

4i21
G l1
2

l1 ~ 4i21 ! /2

32

D

~3!

,

p l51
3 ~ g !5

A

PRB 58

r 1 5 5/4
p g ~ 35214g 2 r 1 2 1g 4 r 1 4 !
1155
3 AV exp@ 21 ~ gr 1 ! 2 #

where G denotes the gamma function. The projectors satisfy
the normalization condition

E

8
p l52
1 ~ g !5

`

0

~4!

p li ~ r ! p li ~ r ! r 2 dr51.

16
It is a special property of our pseudopotential that it also
has an analytical form if expressed in reciprocal space. The
Fourier transform of the pseudopotential is given by
V loc~ g ! 52

4 p Z ion
Vg

2

e 2 ~ gr loc! /21 A8 p 3
2

1C 4 @ 1052105~ gr loc! 121~ gr loc! 2 ~ gr loc! # %
2

4

6

~5!
for the local part, and
3

V l ~ g,g8 ! 5 ~ 21 !

1l

3

(( (

i51 j51 m52l

Y l,m ~ gˆ! p li ~ g ! h li, j p lj ~ g 8 !

* ~ gˆ8 !
3Y l,m

~6!

for the nonlocal part. The Fourier transform of the projectors
p li (r) can be calculated analytically, and for the relevant
cases one obtains
p l50
1 ~ g !5

p l50
2 ~ g !5

16
p l50
3 ~ g !5

A

4 A2r 0 3 p 5/4

AV

A

AV

~7!

,

2r 0 3 5/4
8
p ~ 32g 2 r 0 2 !
15
exp@ 21 ~ gr 0 ! 2 #

~8!

,

2r 0 3 5/4
p ~ 15210g 2 r 0 2 1g 4 r 0 4 !
105
3 AV exp@ 21 ~ gr 0 ! 2 #

8
p l51
1 ~ g !5

16
p l51
2 ~ g !5

exp@ 21 ~ gr 0 ! 2 #

AV

A

A

r 1 5 5/4
p g
3

exp@ 21 ~ gr 1 ! 2 #

AV

exp@ ~ gr 1 ! #
1
2

2

~9!

~10!

,

r 1 5 5/4
p g ~ 52g 2 r 1 2 !
105

,

~11!

A

2r 2 7 5/4 2
p g
15

exp@ 21 ~ gr 2 ! 2 #

2r 2 7 5/4 2
p g ~ 72g 2 r 2 2 !
105

AV

A

r 3 9 5/4 3
p g
105

exp@ 21 ~ gr 3 ! 2 #

~12!

~13!

,

3 AV exp@ 21 ~ gr 2 ! 2 #

p l53
1 ~ g !5

,

.

~14!

~15!

In both real and Fourier space, the projectors have the
form of a Gaussian multiplied by a polynomial. Due to this
property the dual-space Gaussian pseudopotential is the optimal compromise between good convergence properties in
real and Fourier space. The multiplication of the wave function with the nonlocal pseudopotential arising from an atom
can be limited to a small region around the atom as the radial
projectors p li (r) asymptotically tend to zero outside the covalent radius of the atom. In addition, a very dense integration grid is not required, as the nonlocal pseudopotential is
reasonably smooth because of its good decay properties in
Fourier space.
The use of this form for the pseudopotential is also very
advantageous if atom-centered basis functions are used instead of plane waves. Because of the separability all threecenter integrals are products of two-center integrals, and so
only these two-center integrals have to be calculated. If
atom-centered Gaussian-type orbitals are used, these twocenter integrals can easily be evaluated analytically.
In the relativistic case the spin orbit coupling splits up all
orbitals with l.0 into spin-up and spin-down orbitals with
an overall angular momentum j5l61/2. So for each
angular-momentum l.0, one spin-up orbital and one spindown orbital with different wave functions and pseudopotentials exist. Following Bachelet and Schlu¨ter,9 we give a
weighted average and difference potential of these potentials.
The average pseudopotential is conveniently defined as
V l ~ r,r8 ! 5

1
@ lV l21/2~ r,r8 ! 1 ~ l11 ! V l11/2~ r,r8 !#
2l11
~16!

weighted by the different j degeneracies of the l6 21 orbitals.
The difference potential describes the spin-orbit coupling,
and is defined as
DV SO
l ~ r,r8 ! 5

,

AV

16

r loc3 2 ~ gr ! 2 /2
loc
e
3$C1
V

1C 2 ~ 32g 2 r loc2 ! 1C 3 @ 15210~ gr loc! 2 1 ~ gr loc! 4 #

l

p l52
2 ~ g !5

A

,

2
@V
~ r,r8 ! 2V l21/2~ r,r8 !# .
2l11 l11/2

The total pseudopotential is then given by

~17!

1 During the generation of our pseudopotentials we found that there is in general no single minimal parameter set that gives the best overall pseudopotential for one atom. III. For consistency we always performed a fully relativistic calculation for the all-electron atom. i. Analogously.14–16 such as extended norm conservation and hardness. The construction of our pseudopotential differs somehow from the usual method because we fit the pseudopotential parameters that give the best overall representation for the 3643 eigenvalues and charges of several orbitals. Identical parameter sets were used for comparable elements. Another. ~18! where V loc(r) and V l (r. IV. ~18! gives an average potential that contains all scalar parts of the relativistic pseudopotential. in our calculations we never exceeded l max53. For example. the local part of our pseudopotential does not correspond to a certain wave function.m ~19! DV SO l (r. It was already discussed in Ref. contrary to most other pseudopotential construction methods. so that there is a significant overlap with the 3(s.r8 ! 5 3 1l (( ( i51 j51 m52l * ~ rˆ8 ! .e.. rather than producing pseudo-wave-functions that are identical to their allelectron counterparts beyond some cutoff radius. IV! are an exception. This finding is different to the former study of Ref. The atom was put in an external parabolic confining potential to have well-defined unoccupied orbitals.. . the frozencore approximation underlying the construction of all pseudopotentials is not well satisfied. and Schlu¨ter. DETERMINATION OF THE PSEUDOPOTENTIAL PARAMETERS The parameters of the pseudopotentials were found by minimizing the differences between the eigenvalues and the charges within an atomic sphere of the all-electron atom and the pseudoatom. and for unoccupied orbitals to within 10 23 a. The pseudopotential parameters given in Table I typically reproduce the eigenvalues of the occupied orbitals.p) valence wave functions. The same is true for the 4d and 5d elements. We used a slow Simplex-Downhill algorithm17 for the optimization. The pseudopotential parameter r loc was set by hand. For many elements we generated and tested pseudopotentials with different values of r loc . Y l.r8 ! 1DV SO l ~ r. and meets several additional conditions. whereas the total potential contains relativistic effects up to order a 2 . Therefore. To express DV SO l (r. although the latter are much lower in energy than the 3d and 4(s. which themselves are constructed from their all-electron counterparts. Therefore. the same parameter set was used for all 3d or 4d elements.r8 ). the r loc values for the elements in between were interpolated so that no discontinuities occur.PRB 58 RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN . We always tried to use a minimum parameter set. After selection of the optimum pseudopotentials. but is much easier to compute. To ensure transferability of the pseudopotential. our pseudopotentials require significantly fewer parameters than those tabulated by Bachelet. The exchange and correlation energy was calculated with the functional given in Ref. In Table II we give the transferability errors for several excited and ionized states for some elements. more straightfor- . the 4 f wave functions of the 4 f elements are so localized that they overlap with the 5s and 5 p wave functions.m ~ rˆ! p li ~ r ! k li. as the errors for the semicore orbitals are usually larger than for the valence orbitals.r8 ! 5V loc~ r ! d ~ r2r8 ! 1 (l V l~ r. with an error of less than 10 25 a. 5 This functional reproduces very well the Perdew-Wang12 functional. the wave functions of the pseudoatom and the all-electron atom approach each other only exponentially.p) wave functions. ~2! just replacing the h li. as can been seen from Fig. that proved to be much more robust than more sophisticated methods.. where only the first two rows of the Periodic Table were considered. In all these cases. respectively. The 3d wave functions of the 3d elements are strongly localized. j . even when relativistic effects are negligible. because this parameter is not easily accessible by our fitting procedure. where a non-negligible overlap between valence and core wave functions exists. It is a special feature of our method that we fit our parameters directly against the all-electron eigenvalues and charges. V ~ r. as many local minima exist. j by different parameters k li. and the lowest orbital of the next two unoccupied angular momenta. 3 DV SO l ~ r. it is well known the n-1 p levels of the heavy alkali atoms are relatively shallow in energy and extended in space. One way to overcome this problem is the inclusion of a nonlinear core correction18 that considers the contribution of the core charge to the exchange-correlation potential. Nevertheless the difference is very small beyond the core region.e. . Pseudopotentials containing semicore electrons ~Sec. However.r8 ! L–S. ~1! and ~2!# as the nonrelativistic case. 1. In many cases we found it unnecessary to include all unoccupied orbitals in our fitting procedure. j p lj ~ r 8 ! Y l. except for the first row.r8 ) are now scalar relativistic quantities but with the same form @Eqs.r8 ) Neglecting the contributions from in Eq. thereby leading to pseudopotentials of a very high quality. so that sometimes up to some 105 pseudopotential evaluations are necessary until one finds good parameter values.u. For most cases the inclusion of only the first unoccupied orbital for an angular momentum results in comparably good results for the following higher unoccupied orbitals. In most cases the radius of the atomic sphere was taken to be the covalent radius of the atom. i. The fitting of the pseudopotential parameters is numerically demanding. Hamann. we also use Eq. 5 that our fitting procedure yields pseudopotentials that obey the normconservation13 condition.u. A second consequence of our fitting procedure is that. 5. Nevertheless we always checked all-electron eigenvalues. SEMICORE ELECTRONS For many atoms there is no unambiguous separation of the electronic system into a well-isolated core and valence shell. and pseudoeigenvalues and charges of the unoccupied orbitals to verify this. which is sufficient to reproducing the desired accuracy of the fitted eigenvalues and charges. we also considered the next two or three higher unoccupied orbitals for each angular momentum. rather than fitting analytical or numerical potentials that reproduce pseudo-wave-functions.

038386 22.006154 0.005906 21.329941 2.289179 0.105469 Al 3 0.787553 0.238867 1.556987 210.766488 0.247621 0.221786 0.256829 F 7 0.552243 0.506098 20.004104 212.246318 0.179488 0.228432 1.528797 0.325000 0.553476 0.141547 0.C.580318 18. S.556478 0.885509 0.666375 1.864297 2.200000 0.125997 1.180237 29.210950 0.256605 0.000019 214.005152 219.000000 0.088340 2.165419 1.677569 Mgsc 10 0.000000 0.513771 9.002623 27.000878 28.096277 28.000090 0.395701 3.858811 20.491351 5.076245 0.214913 Na 1 0.266917 0.857119 Nasc 9 0.725075 1.034868 22.892612 1.804251 23.000000 0.692852 28.679700 0.004476 221.354839 21.270134 O 6 0.522842 0.141251 0.419008 40. HARTWIGSEN.400000 0.307361 23.545593 36.010336 21.316626 210. LDA pseudopotential parameters in atomic units.084370 17.072869 4.326390 0. H He Li 1 2 1 0.460104 0.651812 0.079306 Lisc Be 3 2 0.005895 0.348830 0.584942 0.000000 0.204014 0.218525 0.049881 0.234820 13.392083 0.200000 0.139668 Mg 2 0.304553 0.739009 0. The meaning of the entries is given in the text.015041 25. GOEDECKER.190000 0.000000 0.015106 227.000000 0.360393 C 4 0.471133 0.871331 2.061666 0.536744 24.970957 1.766407 2.233928 0.592951 3.193438 0.450000 0. HUTTER 3644 PRB 58 TABLE I. AND J.003131 216.003947 .698368 0.433930 0.578642 6.286060 9.232677 N 5 0.661104 0.847271 0.373843 0.658153 Besc B 4 3 0.112023 21.195567 0.000129 224.092462 0.174268 Ne 8 0.891113 0.582004 1.

000000 0.065682 0.022062 5.425036 6.471698 0.249487 4.238705 5.597079 0.000000 0.400000 0.029171 0.938947 0.528411 0.PRB 58 RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .405285 0.950000 0.010261 20.336103 5.000373 26.856693 0.523491 0.005372 26. ~Continued!.012779 0.946474 0.000354 24.317381 0.906928 2.020913 1.016395 0.657455 0.955364 1.281909 0.001430 7.020892 0.418483 20.361757 0.032321 0.489852 22.914612 0.015795 1.258196 0.068194 0.658172 .866579 0.928146 12.654220 6.004812 1.978801 0.310345 0.351619 0. 3645 TABLE I.338208 0.000000 0.720606 Ksc 9 0.014437 3.800000 0.252945 Ti 4 0.784127 20.859241 0.376137 0.025784 5.440233 0.750000 0.529514 0.440000 0.756048 7. .002544 26.563453 26.720000 0.016806 0.062240 4.835768 0.866613 0.389803 0.904330 Sc 3 0.067779 0.923591 0.465876 0.727013 0.246733 0.390000 0.093017 1.410000 0.000371 1.287551 0.905303 3.989348 11.232854 7.048923 1.083053 28.847994 0.385000 0.126329 0.440796 0.000000 0.008463 23.322359 Ca 2 0.722423 0.282606 0.256702 0.016415 0.294826 0.000000 0.430000 0.004343 21. Si P S Cl Ar K 4 5 6 7 8 1 0.645014 0.864754 9.359707 0.376618 0.585479 20.791146 27.243234 0.669737 0. .422738 0.021701 0.056661 1.602516 0.842136 3.352340 5.086411 0.454653 Scsc 11 0.315462 20.420000 0.400000 0.100000 10.967916 3.526550 Casc 10 0.119585 6.260687 3.734309 0.295996 21.300224 20.484278 0.010020 27.017895 4.000814 24.003362 23.554492 7.

221422 Fe 8 0.588377 23.201450 212.229057 20.743504 0.113362 8.137695 25.334235 0.122395 29.360000 0. ~Continued!.828518 0.834982 25.010322 0.003514 5.035625 2. GOEDECKER.032906 0.941291 7. S.374890 Vsc 13 0.241090 0.748683 9.610000 0.286649 0.030816 3.257635 .316763 0.583038 0.254536 0.615878 0.371176 0.014909 29.892229 25.219577 Mn 7 0.499642 20.145354 0.326651 0.005822 2.486101 0.096443 23.009590 3.669304 0.072337 0.234954 20.009333 24.454482 0.3646 PRB 58 C.157868 0.208670 1.005722 5.169781 211.010973 25. AND J.193723 0.640000 0.365000 0.617835 3.392507 10.026941 0.000411 2.156408 28.013176 26.145613 0.925740 5.223021 20.638903 0.079086 0.125896 0.941212 0.002010 7.306011 0.242947 V 5 0.576569 25.327763 Mnsc 15 0.400756 1.280753 0.057447 1.354341 Crsc 14 0.115751 20.284281 0.002717 4.115385 0. HARTWIGSEN.006548 2.565630 20.256230 0.659390 4.370000 0.548789 6.075829 3.375000 0.995418 0.025291 20.242416 0.009007 2.719768 0.161588 0.013685 27.030066 26.278236 0.368776 20.379532 0.690000 0.498578 0.308732 Fesc 16 0.326874 0.896763 0.799031 1.247686 0.575280 24.164188 212.841633 0.826456 0. HUTTER TABLE I.049793 20.076377 20.408712 Tisc 12 0.004536 6.240792 Cr 6 0.246407 0.008030 2.142005 26.481246 0.145557 20.039396 2.514704 0.269268 0.042938 2. 0.016640 1.646819 24.660000 0.952179 20.380000 0.

249015 20.551796 4.075354 0.940527 4.252151 Ga 3 0.006734 20.358800 0.833238 3647 .610311 12.384713 0.888050 1.163546 0.214950 Cu 1 0.975787 0.610048 0.078929 26.338368 0.088557 0.065292 20.001486 20.560000 0.510000 0.010486 0.290091 2.091028 0.572177 0. .584081 0.967605 1.133237 3.245105 0.000873 20.586130 3.513132 20.751272 20.000225 20.RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .356083 0.299738 2.004876 20.386341 0.278710 2.000730 20.742220 20.627342 0.431742 0.449638 29.395062 0.551464 0.560000 0.425399 0.207915 212.608428 0.195226 20.012139 0.831779 1.007137 3.374943 0.610791 0.259140 0.076481 2.539618 0.227086 20. PRB 58 TABLE I.982580 Gasc 13 0.251425 0.266143 Zn 2 0.020384 211.334978 1.218270 20.024580 0. .058598 0.029607 0.058766 3.580000 0.108249 3.530000 0.440457 0.013538 2.400866 0.746305 0.330352 Znsc 12 0.580000 0.355000 0.090915 2.234741 0.704596 0.010792 20.676957 0.278113 Nisc 18 0.634005 20.161131 20.350000 0.873150 0.088965 0.490000 0.025759 214.843283 1.619651 1.249001 20.012767 3.143442 0.269572 213.291602 Cusc 11 0. ~Continued!.849567 2.089543 1.011475 3.845825 24.238168 20.075437 0.029071 20.418391 11.639237 20.482078 27.369325 0.024067 212.023884 20.570000 0.941317 4.068770 3.008708 3. Co 9 0.423734 0.291661 Cosc 17 0.221665 Ni 10 0.820624 0.017521 210.640712 0.822070 20.010489 3.276584 0.

007144 1.903088 0.145131 2.362938 0.005784 2.020562 21.362518 0.439983 0.004273 1.502045 0.664323 Rbsc 9 0. HARTWIGSEN.191205 0.536329 2.096207 0.820292 20.524357 0.282301 0.200395 0.156681 0.486634 5.926675 0.826891 1.026949 215.275441 0.052009 20.015503 0.052466 0.867379 0.3648 PRB 58 C.209592 20.369203 5.490000 0.955699 1.576265 0.632202 0.008991 1.472473 0.743175 Srsc 10 0.741018 9.057876 20.887460 0.432531 0.043443 21.430256 0. S.010000 0.043981 0.557847 Kr 8 0.008685 0.500000 0.002918 1.685283 Se 6 0.079963 9.504151 9. ~Continued!.510000 0.312373 0.087011 0.858806 0.691198 0.748120 0.480000 0.449376 0.911194 3.002846 5.990062 20.613420 Br 7 0.520000 0.006750 20.009267 0. AND J.837564 1.684749 1.646727 0.847333 0.571455 9.100231 20.336679 4.590671 0.004022 21.003708 4.323974 5.064907 21.428207 0.795675 0.336113 0.517120 Rb 1 1.174178 0.995135 3.456400 0.062125 0.315858 .386990 0. HUTTER TABLE I.108823 0. GOEDECKER.074007 0.601064 0.022207 21.627370 0.629228 0.475249 1.240803 Ge 4 0.009802 21.692389 20.434829 0.344218 4.493743 0.461734 20.169126 20.062196 0.031400 3.086008 21.018267 0.068787 21.514895 Sr 2 1.540000 0.812247 0.455323 0.410759 0.049231 0.43728 0.550562 0.398837 3.788369 As 5 0.967372 20.560761 1.301886 0.237532 20.302243 0. 0.373804 5.500000 0.555903 0.458069 20.

Y 3 0.782342 2.627616 20.315381 0.513644 Mo 6 0.368739 0.140668 0.721822 20.301412 20.658363 0.820946 20.363769 23.009198 13.794123 0.348460 0.379571 0.699000 0. PRB 58 3649 TABLE I. .571003 0.075368 20.653851 Ysc 11 0.484368 0.475000 0.630668 Zrsc 12 0.714509 23.548402 0.028673 20.080760 25.251526 21.014064 7.826127 20.153906 26.433648 24.782611 1.007507 20.074388 20.012450 20.104495 0.009380 15.125243 21.043336 21.649998 0.505394 3.696830 0.098654 0.900000 0.525828 Tc 7 0.362426 20.403626 0.189013 2.782457 0.388208 0.528223 0.077005 0.057486 21.750000 0.984224 0.086659 0.428264 0.949864 0.822037 0.018294 20.343891 1.516072 Nbsc 13 0.823817 20.388558 0.080816 22.025653 13.496536 25.678126 0.522621 20.014769 0.247366 20.453384 Mosc 14 0.021058 1.694967 20.393708 0.543211 0.995868 1.724000 0.569552 0.173911 0.677612 0.519890 20.237452 3.376255 0.532651 0.289276 24.470000 0.489848 0.172114 20.741615 0.596788 0.205349 1.011657 1.571767 20. .784275 0.414360 0.124561 1.739877 1.RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .998113 0.819218 0.406442 0.430000 0.041776 .800771 0.513304 Zr 4 0.020044 20.353178 24.068972 16.222025 20.809708 0.846672 0.067713 20.256930 0.075591 1.361734 0.599342 22.289607 0.012055 0.699708 0.396540 0.378681 21.752434 24. ~Continued!.011973 0.013187 4.520496 Nb 5 0.104124 22.031541 0.520655 0.301247 21.246821 21.460000 0.079989 20.010136 1.609675 20.780950 20.014460 13.212179 20.217914 2.874408 0.673000 0.026025 1.

127750 0.029787 3.277495 20.420000 0.377131 0.177686 20.387660 Cd 2 0.650376 0.130081 0.033797 15.016485 1.017773 5.364084 0.410000 0.006821 0.746425 0.095454 24.164257 23.026732 20.304450 0.268679 23.114357 21.103467 1.720259 5.397962 2.357772 0.132307 20.209665 2.369721 0.688787 0.193923 210.242111 1.155278 20.165495 20.373040 20.685594 0.883365 0.110764 1.407986 PRB 58 .258728 20.017053 2.448111 20.387707 20.446609 20.140765 25.046381 25.376061 0.019725 2.201869 23.061808 0.912395 1.065948 1.469208 1.372561 0.598079 0.415911 25.442835 Pdsc 18 0.496950 Pd 10 0.227253 20.018875 5.972873 14. S.570000 0.748323 20.012705 1.3650 C.424753 20.608273 0.715292 20.318920 0.268399 23.813968 20.270000 0.647214 0.737771 20.990284 1.852819 23.319708 21.494916 Ag 1 0.059714 20. GOEDECKER.650000 1. HARTWIGSEN.510487 Ru 8 0.015790 8.046652 13.343111 0.430000 0.485292 0.480632 20. AND J.923960 20.063746 20.630009 0.020374 22.406984 21.596227 25.586709 0.220805 0.364053 0.045193 21.053058 0.337322 21.136909 24.430000 0.369207 Rhsc 17 0.038842 0. ~Continued!.411076 1.135036 15.621429 0.637866 0.495850 Rh 9 0.582571 4.625000 0.019692 20.151597 0.205847 20.016159 Agsc 11 0.828465 0.413271 20.350052 0.709586 0.687723 1.420076 0.389629 21.596000 0.625656 0.704846 0.235842 1.047798 0.440358 Rusc 16 0.009455 1. Tcsc 15 0.122856 1.340791 21.805525 23.342151 0.019949 20.022837 2.498900 0.796838 1.329335 0. HUTTER TABLE I.597870 0.320372 0.639012 20.504694 0.758021 0.910011 3.897931 0.051565 2.350253 0.582204 0.225012 4.917408 20.

865777 1.088691 Insc 13 0.049329 2.610000 0.769355 0.037714 0.577787 0.225115 0.065412 0.805101 I 7 0.130043 0.107522 0.648791 0.306488 24.239830 0.300103 0.611527 0.610912 1.132507 0.745865 0.103931 0.542163 0.213967 0.794325 Xe 8 0.200000 1.530000 0. ~Continued!. .944459 Sb 5 0.773785 20.240949 Cdsc 12 0.033478 0.584234 20.003109 2.481172 0.382713 3.172997 0.223664 20.023513 20.050641 20.560000 0.029333 20.515892 0.376830 20.729821 Cs 1 1.010180 12.507371 0.005563 20.003259 20.970313 0.108473 21.241409 24.004448 2.010809 14.236451 1.046890 1.387085 2.380789 20.256194 0.140963 1.224345 0.590000 0.881119 20.280131 0.560000 0. 1.093313 20.856557 Te 6 0.278329 4.605000 0.597684 0.107627 20.338054 0.395404 3.129208 0.770156 20.674496 0.073414 1.541024 0.066367 0.661825 1.556456 0.360488 Sn 4 0.734280 2.491505 0.672122 0.005057 20.207932 1.002009 4.940150 20.467438 20.227279 0.951477 0.615262 0.550000 0.224737 1.754135 2.013362 0.559819 0.244893 20.035278 0.680228 1.479435 20.770602 0.403551 20.397255 20.576546 0.752764 0.294024 .466731 0.663544 0.552830 0.037537 20.377874 In 3 0.565040 20.445070 0. .PRB 58 RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .834851 20.280478 1.474081 0.138132 20.236603 0.009432 9.058765 4.858132 1. 3651 TABLE I.598565 0.554411 2.317411 0.494459 0.575000 0.141974 20.562251 0.139222 0.566414 0.058835 1.786631 20.190072 0.007066 6.

165254 20.474830 21.458897 0.000000 1.016187 1. GOEDECKER.153732 21.489879 0.514776 0.665403 0. AND J.000000 0.204825 20.010905 217.599763 0.543240 0.300773 Ndsc 14 0.540000 0.470324 0.017660 218.747610 0.000000 20.160512 0.183754 0.100446 0.520228 0.717982 21.529110 0.308978 0.304920 Lasc 11 0.142077 20.318070 22.290940 1.293272 1.333465 22.937158 Basc 10 0.030901 0.832670 20.700444 21.478653 1.151081 0.269439 0.521790 0.214790 0.046729 20.500850 20.000000 20.507751 21.594798 22.362467 0.765636 21.611413 218.121819 20.657669 21.529167 0.680331 0. Cssc 9 0.350982 0.503000 0.202439 20.249880 0.647499 1.492107 22.074241 0.535000 0.780956 22.972641 0.172527 0.540000 0.703593 0. HARTWIGSEN.463710 0.299310 Cesc 12 0.696733 22.761462 0.565305 0.467013 0.012566 20.476308 0.306717 Prsc 13 0.451337 0.472260 35.007193 18.948259 0.000000 0.718934 0.017571 0.282378 22.325290 0.214396 20.314280 0. ~Continued!.897119 0. S.234438 20.447168 20.010841 217.117060 0.088800 0.251723 1.090257 0.526850 0.333507 Ba 2 1.321616 0.748528 0.415083 0.004760 23.828810 20.029857 0.005321 24.763495 0.927065 21.020900 218.006151 19.008547 17.009598 18.795208 0.200000 1.007568 18.375190 0.041017 20.000000 .535000 0.C.847470 1.532083 0.526250 0.424739 1.328377 0.909308 1.815069 0.524623 0.012621 1.378419 0. HUTTER 3652 PRB 58 TABLE I.456821 0.626672 0.294743 Pmsc 15 0.013265 217.852109 0.977217 0.551775 0.721553 20.922593 0.815030 1.

145883 20.467229 0.135186 0.490598 0.291527 Smsc 16 0.054041 20.000000 20.252258 20.512556 1.508750 0.RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .828080 23. PRB 58 3653 TABLE I.144503 0.434642 0.456953 0.268260 Dysc 20 0.432212 0.124371 20.046238 1.473709 0.562601 0.016197 16.697711 0.019830 17.648468 22.052797 0.435239 22.054301 20.111009 1.058286 223.144007 0.424220 0.860425 20.603616 1.206792 1.255135 1. ~Continued!.535540 20.940320 0.470840 0.014315 20.173514 23.518046 0.469043 0.517500 0. 0.063352 219.462014 0.448694 0. .719534 23.891960 0.532803 22.781570 2.482368 0.749717 .014666 17.994331 1.430953 23.284040 Eusc 17 0.763638 0.482809 0.625802 0.911697 0.490038 0.447131 0.562400 20.200867 0.922358 0.261670 Hosc 21 0.668924 0.327404 0.916932 22.092142 21.440590 0.051754 222.426120 0.955298 23.082403 20.781827 1.305057 0.254992 Ersc 22 0.984292 0.373516 1.000000 0.053128 221.520417 0.674534 22.253295 0.008725 20.054087 1.000000 20.420138 0.480586 20.584324 0.355289 20.017938 16.410630 0.523333 0.010619 17.000000 20.000000 21.514583 0.197211 0.946528 0.479677 0.064044 219.069483 225.013267 17.895947 20.278401 Gdsc 18 0.294477 20.275149 0.240300 20.505833 0.273390 Tbsc 19 0.359017 21.429952 0.559798 22.414455 20.718481 0.058347 0.205729 21.419948 0.511667 0.923490 0.051028 220.659367 21.723635 20.984203 22.000000 21.339955 20.011924 17.742863 0.000000 0.551856 20.445907 0.068703 21.022024 0. .105293 2.

517063 0.485635 22.243917 Ybsc 24 0.409923 0.751739 20.923212 20.719819 0.094635 0.529191 26.719000 0.218316 0.449555 0.773916 24.000000 20.025275 0.502917 0.414358 0.947663 0.789137 20.413373 0.901544 26. S.027036 1.741500 1.152450 21.247293 1.121283 3.472681 0.872548 20.202200 0.418385 0.028006 5.534370 Tasc 13 0.121556 22.999006 0.385339 25.902736 21.121021 0.184678 20.399602 20.349470 20.518567 20.534959 Wsc 14 0.173245 0.451680 20.070076 229.947196 20.627697 24.119299 228.692204 20.560000 0.134801 2.086716 4.232629 Hfsc 12 0.238298 Lusc 25 0.C.623116 2.238744 0.607504 1.421853 0.779422 25.392870 0.037053 2.215211 20.074295 0.790959 22.000000 20.077093 .553621 20.802990 21.386947 20.744000 0.224087 22.695148 21.348495 0.102353 2.075772 4.161166 0.770646 0.072130 231. HARTWIGSEN.353308 0.661610 25.177436 0.500000 0.546236 2.422810 0.249126 Tmsc 23 0.852262 0.702084 0.357144 2.013732 21.086193 0.299943 0.459363 0.222119 3.444025 0.000000 21.410994 W 6 0.000000 21.402309 0. 0.331637 25.045370 21.185378 0.678540 2. AND J.715480 1.582463 0. ~Continued!.666675 20.101322 1.094493 2.120771 20.025718 17.426388 Ta 5 0.000000 1.023833 17.109760 0.205344 4.581801 0.932854 0.100046 1.708136 20.800251 2.393896 0.347237 0.104159 0. GOEDECKER.021783 17.742307 0. HUTTER 3654 PRB 58 TABLE I.889967 0.395416 0.550000 0.058450 2.724853 0.461345 0.723799 1.391206 0.814297 21.022637 22.418570 0.600973 20.282764 20.649992 1.696809 0.091862 226.540000 0.205860 21.000000 20.509800 22.005338 0.564442 21.336371 25.497000 0.436518 0.432346 22.

590006 0.538614 0.000000 20.684971 0.440706 8.720016 2.370589 23.667000 0.140351 0.379865 25.000000 20.701113 2.616000 0.471229 0.PRB 58 RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .925829 0.528529 0.219517 11.410578 0.209737 9. TABLE I.348213 25.448772 23.635428 20.641000 0.440951 0.216440 10.403252 0.511211 20.760720 20.658976 0.788715 0.142035 0.530000 0.102862 3.018323 20.717553 0.672543 20.504403 20.116941 5.719070 3.111557 5.792039 20.994366 20.422395 0.613073 2.169306 1.919308 1.180816 2.046706 1.153427 3655 .745839 0.870817 0.103221 2.811037 1. Re 7 0.017012 2.754315 0.440459 2.496693 20.247034 0.000000 20.963712 1.894601 20.262716 4.510307 0.447430 0.471745 Irsc 17 0.065883 22.146912 5.102396 20.693000 0.510000 0.376428 Pt 10 0.520000 0.509960 0.017060 0.331919 20.499523 20.156382 27.112145 2.398652 0.304726 0.755431 20.000000 20.131595 0.226472 21.930121 0.394853 20.139506 2.692130 23.630613 0.632067 0.039349 11.151950 1. .402367 0.590000 0.640360 1. .243278 20.960055 24.552295 0.500000 0.785758 20.900546 0.053284 0.404469 0.243095 0.647716 1.698123 20.489984 25.412886 25.057831 1.943957 26.132970 2.875580 20.411426 0.689630 0.313786 27.451243 Ptsc 18 0.875251 0.868732 0.409942 0.390395 Os 8 0.114346 21.956978 0.732892 20.666100 20.880133 0.763296 24.604428 2.139761 0.209758 20.461792 20.521180 0.468779 20. ~Continued!.509816 0.921955 26.565347 24.650000 0.296571 21.859620 21.408453 20.592660 2.776974 20.545484 0.027417 2.380574 0.500954 Resc 15 0.367964 Au Ausc 1 11 0.445832 2.650429 0.616765 21.269379 0.539599 0.445999 0.315509 23.225181 1.701752 21.487586 Ossc 16 0.520132 0.380252 Ir 9 0.904509 3.430746 24.127199 4.497538 0.396415 22.

303680 20.480774 0.527078 0.092330 0.798673 0.903742 1.474056 0.305314 0.301886 3.008296 20.880468 At 7 0.471028 20.979927 0.433232 0.570000 0.374967 0.945557 0.992206 0.549896 0.655578 0.621648 0.122282 20. ~Continued!.630000 0.864420 0.520411 0.038963 20.502423 0.012948 20.120456 PRB 58 .353595 20.170651 20.072771 22.679437 1.550000 0. HARTWIGSEN.396354 0.247614 0.037544 14.647950 0.981832 0. Hg 2 0.029256 6. GOEDECKER.753143 1.033886 13.875766 0.377634 0.344122 20.669886 0.971939 Bi 5 0.799941 0.612279 0.617500 0.468948 0.781337 0. S.045488 20.293920 2.031462 20.402932 20.806060 20.748947 0.235846 1.186849 0.318821 20.053714 1.934683 Po 6 0.001118 4.164960 20. AND J.031664 23.594562 0.629185 0.705259 0.000000 0.155759 21.148484 23.846641 0.586721 0.019164 20.3656 C.765041 0.272807 0.605000 0.012233 23.034954 20.689988 0.473265 20.572016 0.341390 2.531816 20.521802 0.411731 1.168641 0.144203 0.063512 Tlsc 13 0.466127 20.023760 20.120638 0.615182 0.788337 0.754005 0.592500 0.557746 0.640000 0.000000 1.676697 0.378476 0.004459 4.513697 20.812108 1.169493 20.580000 0.029217 10.709823 0.393185 Pb 4 0.339386 20.134572 3.653348 20.759668 0.627827 0.401894 Tl 3 0.735851 20.570000 0.190429 21.965903 20.100960 0.678858 0.022941 7.100000 Hgsc 12 0.084989 21.207711 0.296329 1.838365 Rn 8 0.540969 0.661001 1.023134 20.200652 0.020931 2.326560 1. HUTTER TABLE 1.

35310 23 20.2462 0.2510 0.3612 0.1198 0.28310 23 0.46310 23 0.7216310 21 0. but still yield accurate results for the valence wave functions.31310 22 20.43310 24 20. The choice of which electrons are treated as semicore electrons also depends on the required accuracy. The difference between them is shown by the dotted line on a logarithmic scale. but only 0. Therefore the inclusion of semicore electrons does not inevitably require the use of extremely large basis sets.4294310 21 0.2681 0.17310 23 20.2970 20.38310 24 20. Pseudopotentials with semicore wave functions always require higher computational effort. . For most applications the need to use semicore pseudopotentials depends on the required .1 hartree.28 310 22 hartree for the 12e semicore pseudopotential. In many cases it is not quite clear if semicore electrons play an important role or not.RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .15310 23 ward.1879 0.3073 0.2605 20. which corresponds to a Ti41 ion. It is unnecessary that the eigenvalues and charge distribution of the semicore wave functions have the same accuracy as the valence wave functions of the pseudoatom.2529 20.58310 24 20. For the 4e pseudopotential the error is 0.u.1166310 1 0.1084 0.41310 25 20. In this work we decided on the second method. we tried to include not too many semicore electrons into our pseudopotentials. the explicit inclusion of the semicore electrons ensures that our pseudopotentials still work well for systems where nonlinear core corrections fail. 1.3181 20.28310 22 0.2505 0. In many applications like molecular structure calculations semicore pseudopotentials yield converged results with comparably small basis sets even if the calculated total energy is still far from its converged value. including the 3s and 3p semicore electrons. in our molecular calculations the highest plane-wave energy cutoff was needed for calculations with the fluorine pseudopotential which has no semicore electrons at all.34310 24 20. Therefore the error for the eigenvalues of semicore wave functions for our pseudopotentials is within 10 23 –10 22 a.3882 0.3258310 1 0. In Table III the transferability error of two Ti pseudopotentials is listed for several states.2673 0.5293310 22 0.40310 23 0. This is most significant for the 4s 0 4 p 0 3d 0 state. which is about three orders of magnitude worse than the typical error for the valence wave functions. PRB 58 3657 TABLE II. As we were interested in generating pseudopotentials that can be used together with plane-wave basis sets within a reasonable computational effort.3258310 1 0..23310 23 0.33310 24 20.3593310 21 20. In fact.11310 23 20. the transition metals of groups IIIb–VIIIb. In addition. For most states the calculated excitation energies are much closer to the all-electron values for the Ti semicore pseudopotential.40310 23 0. all FIG.35310 23 0.3593310 21 0.1682310 21 0.52310 24 20. This ensures that our semicore pseudopotentials still can be used with programs where nonlinear corrections are not considered. and the lanthanides treat the (n21! s and the (n21!p electrons as semicore electrons.2482310 21 0.46310 24 20. and IIIa ~except B and Al!. They contain more electrons. and larger basis sets are necessary for a sufficient description of the localized semicore wave functions. Our semicore pseudopotentials for the group Ia and IIa elements.10 0. Element C Si Ge Sn Pb Ti Tisc state DE ~Ha! Error ~Ha! 2s 1 2p 3 3d 0 2s 2 2 p 3 3d 0 2s 2 2 p 1 3d 0 2s 2 2 p 2 3d 1 3s 1 3 p 3 3d 0 3s 2 3 p 3 3d 0 3s 2 3 p 1 3d 0 3s 2 3 p 2 3d 1 4s 1 4 p 3 4d 0 4s 2 4 p 3 4d 0 4s 2 4 p 1 4d 0 4s 2 4 p 2 4d 1 5s 1 5 p 3 5d 0 5s 2 5 p 3 5d 0 5s 2 5 p 1 5d 0 5s 2 5 p 2 5d 1 6s 1 6 p 3 6d 0 6s 2 6 p 3 6d 0 6s 2 6 p 1 6d 0 6s 2 6 p 2 6d 1 4s 1 4 p 1 3d 2 4s 2 4 p 0 3d 0 4s 0 4 p 0 3d 0 4s 2 4 p 1 3d 2 4s 1 4 p 1 3d 2 4s 2 4 p 0 3d 0 4s 0 4 p 0 3d 0 4s 2 4 p 1 3d 2 0.95310 24 20. IIb.48310 23 0.37310 24 20. solution is the explicit inclusion of the semicore electrons into the pseudopotential. electrons of the completely filled nd shell are treated as semicore electrons.44310 23 0. For all elements mentioned above semicore wave functions improve the description of highly positive charged ions. .1166310 1 0.65310 24 20.20310 23 0. For the elements of groups Ib. Transferability errors.1198 0. We always tried to generate semicore pseudopotentials with semicore pseudo-wave-functions that are as smooth as possible.3587310 21 0. Relativistic all-electron ~solid line! wave function and pseudo-wave-function ~dashed line! of the valence electrons of gold.47310 21 20.

997b 3.861 2.991 2.072 2.994b 3.373 2.706 sc .553 2.082 5.174 2.116 sc 3.409 3.529 6.996 sc .070 1.381 6.028 2. HUTTER 3658 PRB 58 TABLE III.029 4.566 7.303a 4.192 2. Comparison of bond lengths of molecules calculated with our pseudopotentials ~PSP! and obtained with GAUSSIAN 94 or NUMOL ~Ref.477 2.918 3.901 2.668 1.182 3.052 2.015 4.307 3.969 3.140 2.200 sc .605 2.76 3.696 2.64 3.756 2.936 2.613 1. 2.31 3.132 2.704 2.038 1.808 2.271 3.052 3.542 2.820 2.069 5.853 2.41 3.152 3.C.827 2.63 2.612 2.018 sc .082 5.497 2. 3.169 2.733 2.313 3.067 1.929 4.725 2.374 2.943 sc .120 3.934 3.016 3.912 2.074 1.072 2.516 sc .100 sc .371 2.003 3.413 2.524 2.922 3. 1.351 3.836 2.024 21 21 21 21 28 21 29 2.809 2.818 3.609a 2.469 2.897 2.952 3.279 3.027 2. 2.871 21 21 21 22 23 21 24 23 23 21 23 25 25 23 21 23 21 21 23 23 21 21 21 21 21 26 21 23 21 27 21 2. 4.959b 2.595 2. 3.304 sc .114 2.436 3.030 2.447 5.401 5.127 2.726 1.845 2.331 2.126a 3.072 6.783 23 21 23 21 21 23 23 21 21 21 3.318 sc .882 2. S.363 2.789 3.330 2.547 2.932 3.051 3. Molecule H2 Li2 LiH Be2 BF3 BH CH4 C2 H2 N2 NH3 HCN H2 O CO CO2 F2 HF CH3 F Na2 NaH Mg2 MgH AlH SiH4 SiO SiF4 P2 PH3 PN S3 H2 S CS CS2 Cl2 HCl CH3 Cl K2 KH CaH ScH ScO TiO VO CrO MnO FeO CoO NiO NiH r~CC! r~HC! r~CN! r~HC! r~CF! r~HC! r~CCl! r~CH! PSP AE Expt.013 1. 5.037b 3.062 3.187 3.174 2.032b 2.840 2.072 7.592 2.060 7.169 2.787 21 .3pd) basis set if not otherwise mentioned.263 2. 2.955 sc .795 2.97 3.029 sc .179 2.447 5.671 2.760 2.932 2.035b 1.109b 3.808 3. All GAUSSIAN 94 calculations were done with a 6 – 31111G(3d f .834 2.470 2.300b 3. GOEDECKER.020b 2.578 2.903 2. AND J.764 2.270 2.932 2.440 3.558 2. Bond lengths are given in bohr.930 3.534 sc . 21! ~AE! and the experimental data. 2. Ref.005 2.039 1.510 2.130 2.936 3.358a 2. HARTWIGSEN.274 2. 3.958 2. 3. 2.742 2. 2.622 1.195 2.196 2.

143 21 21 3.705 2.783 4.284 3.951 4. 3.161b 3.861 3.701 2.879 3.073c 3.146 3.167 sc .133 3.755b 4.394 3.635 3. 3. 3.038 4.437 3. 3.600 sc .69 3.368 sc . 7.670 2.278 sc .173 3. .146 3.858 3.074 3.887 2.898 sc .314 2.322 3.027 2.749 2. 2.049 2. 2.176b 2.976 4.373 3.688 4.475 30 21 21 23 23 21 23 31 21 21 23 21 21 21 21 21 21 21 21 21 21 32 33 21 23 21 21 34 21 21 21 21 21 21 21 21 21 21 .010 sc .434 3.173 3.411 4.882 3.072 4.731 sc .463 4. 3.052 3.217 3.473 4.371 2.073 5.252 7.522 5.108 3.063 3.420 3. 2.472 3.504 sc .824 3.896 3.591 sc 3.602 2.473 4. 3.174 4.326 3.329 3.234 3.056 4.956 4.653 3. 3.551 2.573 2.258 2.440 4.025b 3.783 4.872 3.097 4. 2.176 sc .031 5.479 sc .285 3.512c 3.176 sc . 2.323 sc .327 3.157 sc . CuO CuH Cu2 ZnO ZnH GaH GaO GaF GeH4 GeO As2 As4 AsH3 H2 Se Se2 Se3 (C2 v ) SeO2 Br2 HBr CH3 Br r~CBr! BrF BrF3 Rb2 RbH SrH SrI2 SrO ZrO RhC PdH AgH Ag2 CdH InH SnH SnO SnS Sb2 SbCl3 Te2 H2 Te HI I2 Cs2 CsH BaH IrC PtH PtC AuH HgH TlH TlCl PbH r~CH! r1 ~BrF! r2 ~BrF! 3.022 3.848 3. .RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN .560 sc .414 4.110b 3.171 4.888 3.636 3.041 4.534 4. 3.630 3.760 3. 2.788 4.70 4.81 3.013 3.331 3.855 2.376 sc .624 4.832 3.324 2.403 3. 3.424 3.765 4.296 2. 3.875 2.067 3.67 3.076 30 23 21 21 23 23 23 21 3.142 3.069 3.717 sc . 3.086 sc .401 4.433 3.039 8.010 3.195 21 21 21 3.089 4.299 2.865 3.713 4.180 2.023 sc .053 5.289 7.612 2.974 4.912 4.029 8. 5. 7.183 2.318 3. PRB 58 3659 TABLEL III.472 3. 2.156c 3.671 sc .367 2.153 sc . 3.889 3.123 4. 3.118 3.045 4.009 sc .886 3.333 4.759 4.473 3.095 3. ~Continued!.353 2.235 sc .630 sc .

7 2.411 22 34 22 a with 3 – 21G basis set. GOEDECKER.1 h 11. HUTTER TABLE III ~Continued!.3 k 13. together with their experimental values. the error in the calculated bond lengths is quite large. HARTWIGSEN.3 h 23.237 4. 21 3. To estimate the error arising from the pseudopotential approximation. PbS PbCl2 BiH 4. or if the computational resources are limited.352 A 1 2 h 11. With a few exceptions the values calculated with GAUSSIAN 94 agree within a few thousandths of a bohr. Parameters for elements marked with sc correspond to semicore pseudopotentials.2 h 03. This clearly demonstrates that relativistic effects do not influence the bond lengths for these molecules on a relevant scale.3 A 5 1 h .3660 PRB 58 C.1 r0 r1 r2 C2 C3 h 02. Only the nonzero parameters are shown in Table I.2 h 22. our calculated values should be compared with these LDA results rather than with the experimental bond lengths. Unfortunately exact LDA values for molecules containing heavier elements often are not available because of the lack of a sufficiently accurate basis set.1 k 11. 5 2. extremely large boxes and high plane-wave energy cutoffs were needed.2 k 12. we took the all-electron results from Dickson and Becke. These pseudopotentials require only small basis sets which is sometimes a necessity for the study of large systems. In all calculations we used our scalar relativistic pseudopotentials. Our calculated bond lengths containing only first or second row atoms also agree to within one or two thousandths of a bohr to those obtained with nonrelativistic versions of these pseudopotentials that have already been published5 ~differences in the case of HCN are due to the choice of a different simulation box!. the coefficients h li. neglecting the terms for spin-orbit interaction. The error of the pseudopotential approximation for first row atoms is nearly ten times smaller than the LDA error.3 A C4 .i and k li.20 and therefore should be close to the LDA limit.2 k 22.2 ~20! 5 0 h .1 k 21. For some non-semicore pseudopotentials. with the LDA results published by Dickson and Becke. VI. For some molecules where no high precision basis sets are available.252 h 01. are listed in Table III. the missing h li. agree very well with the all-electron values obtained with GAUSSIAN 94. In all cases except for the non-semicore pseudopotentials. Whenever possible we tried to determine values for the bond lengths.609 3. AND J. Therefore we constructed both types of pseudopotentials for most elements where semicore electrons can play a significant role in electronic structure calculations. MOLECULES We tested our pseudopotentials by calculating the bond lengths of a large number of molecules. however. PARAMETERS In the following we list the parameters for all elements up to Rn. and it is comparable to or better than the results obtained with other all-electron methods. To obtain this high level of accuracy. The entries in Table I have the following meaning: Element Z ion r loc C 1 h 01. The bond lengths calculated with our pseudopotentials.2 h 12. as our pseudopotentials also include relativistic effects. As a reference for the quasiexact LDA value. j of the nonlocal projectors for i5 ” j are not listed.3 ~21! A 1 2 100 0 h . including semicore electrons where necessary. In order to keep the table as small as possible.35 1 2 h 02. and should be tested carefully. GAUSSIAN 94 b GAUSSIAN 94 accuracy and necessary computational effort.552 3. j and k li. j are 1 2 h 01. j have to be calculated from h li.252 A 3 0 h .526 4. c Reference 20. better than the results obtained with standard Gaussian 6 – 31G * basis sets.i . with 6 – 31111G(3d.3pd) basis set. To obtain the full parameter set.3 k 23. we also list the bond lengths calculated with GAUSSIAN 94 ~Ref.3 h 13. It must be mentioned that our results especially for molecules with heavier elements are not exactly comparable to the values obtained with GAUSSIAN 94 or the values of Dickson and Becke. V. no f polarization functions have been used!. 19! using a 6 – 311G 11(3d f . so that for some molecules the accuracy of the calculations was limited by our computational resources.2 ~22! ~23! . 63 3.374 4. S. Nevertheless these pseudopotentials may still be of interest for electronic structure calculations if no high precision is required. the accuracy relative to the exact LDA value is. The relevant equations for the h li. that are converged to '1023 bohr. The calculated bond lengths. j and k li.1 h 21.3pd) basis set ~for the 3d elements. and for the heavier elements at least comparable to the LDA error.

8503 ~1991!. Maschke. and M. Phys. 200 ~1988!. T. I. Vol. 136. Y. Ernzerhof. The pseudopotentials are highly accurate and transferable. Scheffler. J. 15 R. and H. Teter. Carter. Rev. Phys. GAUSSIAN 94. C. 1976!. G. Becke. 7406~E! ~1986!. Hohenberg and W. Burke. edited by K. 80. 28 R. X. 97. j by k li. P. . Miehlich. B. Phys. Phys. Rev. Peng. M. 12 J. J. Defrees. J. ibid. H. P. Gomperts. D. and M. 5031 ~1993!.352 A 35 1 h . R. Rev. Ayala. Phys. G. 13 065 ~1992!. B864 ~1964!. Rev. ~29! ACKNOWLEDGMENTS 7 2 h . Phys. Dickson and A. 19 M. 20 R. J. Chem. B 45. 3 X. Pittsburgh. Johnson. C. Gordon and A. Phys. Phys. Bachelet. S. M. 116. 34. Stumpf. Andres. D. 88 ~1992!. and no tabulated functions are needed. C. 4199 ~1982!. Phys. and J. Rev. 1986!. S. W. Louie. 8822 ~1986!. P. Y. 8800 ~1986!. Allan.3 1 2 A A ~24! gives the equations for k li. K. and J. Trucks. Phys. Gerry. Can. 12 264 ~1990!.. Yang. Savin. B 25. Schlu¨ter. J.35 1 2 h 22.2 ~26! 63 2 h . P. J.252 h 21. J. M. Allan and P. 1703 ~1996!. S. 1. Rev. Wong. Perdew. 11 3. ~20!–~29! 1 3661 G. 3898 ~1993!. Rev. Lett. Baker. and have been tested in extensive atomic and molecular calculations. J. 27 G. Phys. Sham. S. Mills. Phys. and A. Scheffler. Phys. in Atomic and Molecular Physics. G. j for iÞ j as independent pseudopotential parameters does not improve the results. M. M. New Series. Replogle. ~25! VII. PRB 58 h 11. Gill. M. Stoll. 11 A. The necessary programs are available from the authors upon request. Spectrosc. Hamann. M. Singh. Financial support by Hoechst AG is gratefully acknowledged.-H. Krakauer. Y. 3098 ~1988!. Robiette. B. Mol. D. PA. 9 G. H. April 1990 ~unpublished!. 33. 3865 ~1996!. J. A. Chem. Goedecker. Gaussian. Herzberg. j in Eqs. 18 S. T. 14 E. Vetterling. Rev. 6 X. Kleinmann and D. A1133 ~1965!. Cohen. Phys. A. Data 8. P. J. 1738 ~1982!. J. Bachelet and M. Fox. Handy. D. 13 D. G. Head-Gordon. D. We thank Sverre Froyen and Mike Teter for making available their relativistic atomic codes. D. Goedecker and K. J. R. 143 3. H. Ortiz. 10 J. All terms for both Fourier and real space are given analytically. Phys. Chem. McRae. ibid. J. J. English. 3652 ~1989!. A. Schlu¨ter. P. 1425 ~1982!. P. M. J. Gonze. 22 V. Ref.3 ~27! 1 18 2 h . j . Perdew and Y. Hamann. Montgomery. Fritz-Haber-Institute Research Report No. L. L. Shirley. 59. B 54. 1606 ~1992!. Merer. 1136 ~1987!. 48. E. K. R. Inc. Frisch. B. Chem.. G. Lett. P. J. Keith. P. Schlu¨ter. Lett. Stumpf. 9 2. B. 4 R. Gonzalez. Bondybey and J. L. Structure and Data of Free Polyatomic Molecules. W. C. and N. Replacing h 1i. and M. H. 21 K. New York. A. 619 ~1979!. Haas. 2 A143 3. Rev. J. Chem. Liu. Kohn. D. Martin. C. Al-Laham. 17 W. and B. V. 568 ~1984!. M. Becke. A. 99. B 41. Cohen. 785 ~1988!. New York. Mol.i . Berlin. 2 L. 23 J. 1494 ~1980!. Wang. Kaekell. Wang. H. Zakrzewski. H. Robb. V. Hellwege. .3 h li. Lee. Teter. Gaussian-type pseudopotentials for other exchange correlation functionals or gradient-corrected functionals can easily be constructed using our LDA parameter sets as an initial guess. Rev. C. M. A. Press. J. B 48. L. Teter. Petersson. 140. C. Rev. R. Phys. Phys. P. 58 ~1986!. 7 ~Springer. and J. 6 A11 3. 43. J. J. A. 25 S. G. Schlegel. Rev. j 5h lj. W. Phys. Gonze. Hellwege and A. Phys. R. Cheeseman. 59. 5 S. R. 13 244 ~1992!. Phys. 1901 ~1979!. Group II. M. J. J. and E. 2103 ~1982!. B 44. A. Parr and W.RELATIVISTIC SEPARABLE DUAL-SPACE GAUSSIAN . 1995. Gonze. Gray and A. Callomon et al.35 1 6 h 12. B 26. Rev. Stewart. Parr. 1979!. By this procedure we minimize linear dependencies of the h li. Rev. G. M. ibid. 1989!. SUMMARY ~28! We developed a complete set of relativistic LDA pseudopotentials for the whole periodic system up to Rn. Preuss. 26 M. and M. B. Phys. G. and M. M. S.3 1 14 1 h . 642 ~1980!. We found that the treating all h li. C. Density-Functional Theory of Atoms and Molecules ~Oxford University Press. Wang. Chiang. M. The pseudopotentials are easy to use as only a few parameters are necessary. and R. W. Bylander. Foresman. 157. Rev.352 h 21. Lett. B 26. Numerical Recepies ~Cambridge University Press. Teukolsky. Perdew and Y. Harmony et al. New York. E. W. 7 P. L. W. M. R. and C. Pople. Kohn and L. 24 D. Rev. B 37. Martin. Phys. Huber and G. and M. A. D. j during the fit. Froyen. 37. Rev. B. A 38. Rev. 77. Binkley. Joannopoulos. Scheffler. B 40. A 45. Raghavachari. D. Phys. Perdew. 16 . L.. Flannery. 46. D. Landolt-Bo¨rnstein. Phys. Hutter. Constants of Diatomic Molecules ~Van Nostrand Reinhold. Phys. B 33. Chen. 8 W.

and J. Phys. Phys. LandoltBo¨rnstein. Berlin. Hellwege. HARTWIGSEN. J. GOEDECKER. 474 ~1979!. S. AND J. Reference Data on Atoms. 1985!. in Atomic and Molecular Physics. in Atomic and Molecular Physics. New Series. C. Smirnov. Kuchitsu. Demaison et al. Berlin. edited by K. Vol.3662 C. 33 E. 71. M. Hellwege and A. Berlin. 1974!.-H.. J. Group II. M. Molecules and Ions ~Springer. Molecular Constants. N. P.. HUTTER T. Landolt-Bo¨rnstein. . 21 ~Springer. 29 30 32 PRB 58 A.. Gallaher. 34 W. Verges. Vol. A. 1992!. Hirota et al. 121. Lett. J. Chem. Crozet. Amiot. Chem. Group II. 5354 ~1985!. Weickenmeier et al. New Series. Phys. 82. DeVore and T. edited by K. 6 ~Springer. 31 C. Structure Data of Free Polyatomic Molecules. 390 ~1985!. Radzig and B. Chem.