
==== Front
J Mol Model
J Mol Model
Journal of Molecular Modeling
1610-2940
0948-5023
Springer Berlin Heidelberg Berlin/Heidelberg

39305376
6136
10.1007/s00894-024-06136-3
Original Paper
Hyperhardness and hypersoftness of atoms and their ions
Zaklika Jarosław 1
Ordon Piotr piotr.ordon@upwr.edu.pl

2
Komorowski Ludwik 1
1 https://ror.org/008fyn775 grid.7005.2 0000 0000 9805 3178 Department of Physical and Quantum Chemistry, Wrocław University of Science and Technology, Wyb. Wyspiańskiego 27, 50-370 Wrocław, Poland
2 https://ror.org/05cs8k179 grid.411200.6 0000 0001 0694 6014 Department of Physics and Biophysics, Wrocław University of Environmental and Life Sciences, Ul. Norwida 25, 50-373 Wrocław, Poland
21 9 2024
21 9 2024
2024
30 10 34423 3 2024
3 9 2024
© The Author(s) 2024
2024
https://creativecommons.org/licenses/by/4.0/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Context

The theory of reactivity based on cDFT has been supplemented with the new method of calculating the atomic and local indices. With the use of previously derived relationship of the electron density gradient to the softness kernel and to the linear response function, we deliver theoretical analysis to obtain significant reactivity indices—the electron density derivatives: local softness and local hypersoftness together with the global hyperhardness index and the derivative of the global softness with respect to the number of electrons. The local derivatives have been applied in the calculation of responses of atoms to perturbation by an external potential by the alchemical approach. The vital role of the local softness has been confirmed; the potential role of the hypersoftness has been indicated.

Method

Our original theoretical scheme has been numerically illustrated with the results obtained with electron density calculations with B3LYP method implemented in Gaussian 16 package. The aug-cc-pvqz basis set has been routinely applied, except for the Ca atom (cc-pvqz). Using the pVTZ basis set recommended by Sadlej was necessary for the potassium atom.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s00894-024-06136-3.

Keywords

Atoms
Hyperhardness
Hypersoftness
issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2024
==== Body
pmcIntroduction

The conceptual density functional theory (cDFT) has been widely recognized as the potential apparatus for theoretical exploration aiming at a priori prediction of chemical reactivity of atoms and molecules. Its birth was marked by the landmark paper by Parr et al. [1] where chemical potential derived from the Hohenberg and Kohn theorem [2] has been identified as electronegativity. Further development was clearly divided between the theory and application. In the theoretical advances, the significant step has been done with the paper by Perdew et al. [3] where Hohenberg–Kohn theory was formulated for fractional number of electrons, for systems described by a quantum mixed state. The practical applications of these theoretical achievements were remarkably initiated by the paper by Parr and Pearson [4]. While innumerable successes have been reported in both directions, coupling the two fields together into the coherent system of “molecular thermodynamics” is still in progress [5, 6].

The proposal of axiomatic approach to the chemical concepts by Ayers et al. [7] exposed once again the ever-growing disparity between the language of chemistry with its traditional concepts (atoms, bonds) and the wealth of quantum chemistry tools. The short paper with the exhausting bibliography has confirmed the longstanding interest of the group in bridging the fields of practical chemistry and quantum theory [8, 9]. Interesting practical guidelines to possible merging these two separately developed and powerful methodologies have been presented: the role of observables, the free choice of basic axioms, and the role of their mathematical formulation for the concept based upon them. Interestingly, the authors concentrate on the conceptual integrity of the possible results of that approach, rather than on the computational accuracy on the output from a theoretical effort.

An example of the intuitive, independently developed application of the approach along those lines has been presented by Ordon and Komorowski on the ground of conceptual density functional theory (cDFT) [10] The Hohenberg and Kohn theorem [2] on the electron density and the existing mathematical apparatus of the DFT have been applied to construction of the density functional (DF) connectivity matrix for any molecular system in its stationary state [11, 12]. The matrix elements are given as divergences of the atomic Hellmann–Feynman forces [13] and are numerically obtained from the energy Hessian of the system. They have strict physical interpretation as the cumulative (or effective [14–16]) force constants between any pair of localized nuclei, providing the clear and unique physical measure for the bonding status between atoms situated in a molecular array. Individual bonded atoms have been characterized accordingly: the diagonal terms of the DF connectivity matrix reflect the valence of the atom in a molecule and the variation of the elements of the DF connectivity matrix with the reaction progress has been demonstrated to provide the much appealing picture of the modification of the bonding status of atoms along IRC (reaction fragility spectra) [17–20]. The vibrational atomic modes emerging from this analysis [10] are in close relation to the calculated adiabatic internal vibrational modes AIMO by Kraka et al. [21, 22]. Review of many other attempts and approaches to the reaction force constant may be found in the recent papers [23, 24].

The results of the reaction fragility spectra have prompted us to developing the theoretical concept even further. We have derived the following gradient theorem from the basic properties of the connectivity matrix: [10, 25].1 ∇ρr=∫ωr,r′εr′dr′=-∫sr,r′εr′dr′

whereωr,r′=δρ(r)δvr′N  is the linear response function,

εr=∑AεAr  is the total external electric field, and

sr,r′=δρ(r)δvr′μ  is the softness kernel.

This exact formula represents the much-needed relation between the electron density and the electric field from the nuclei in every point of a system εr, very much in the spirit of the analysis by Bader [26]. The clear mathematical form of this gradient theorem and the unveiled equivalence between the use of the linear response function ωr,r′ and the softness kernel sr,r′; the nonlocal electronic response functions [27–30]  Eq. 1 has opened a way to exploration of its further consequences. We combined it with another cDFT axiom—the nearsightedness of matter, that can be applied to sr,r′ but not to ωr,r′ [31–33]. Vela and Gázquez [34] proposed the approximate expression for the softness kernel:2a sr,r′≈δr-r′sr′

Substituting to Eq. 1 and the integration of the delta function with the local softness and the electric field gives obvious result:2b ∫δr-r′sr′εr′dr′=srεr

Finally, this resulted in the recently published Ansatz of great potential utility (albeit no longer exact) allowing for calculation of the local softness: [10, 25]. 2c ∇ρr=-srεr

The route to calculation of the local softness sr=∂ρr∂μv=-δNδvrμ describing the response of the electron density to the electric field has been opened. The necessary test of this approach was achieved by calculation of the global softness (S) of atoms and ions directly by integration of the local softness s(r) resulting from Eq. 2 and calculated from the computable electron density alone. Consequently, another important local response function has also become available directly from the density function: the Fukui function f(r) = s(r)/S [25].

The step towards the search for higher response functions has been made by the analytical calculation of the derivatives [∂fr/∂N]vand[∂sr/∂N]v for the hydrogenic orbitals [35]. This present work delivers the theoretical and computational scheme to obtain the local hypersoftness of atoms and their ions. This is possibly non-negligible reactivity index that contributes to the global changes of the state functions (ΔN, ΔE, Δμ), formally replacing the nonlocal response functions, as demonstrated in our previous paper [35]. Numerical tests of this concept present a collection of results for atoms and ions, and they are aimed at estimation of the role of first and second local derivatives of their density over N for the physically observable effects of change ΔN, ΔE, and Δμ induced in chemical reactions.

Hyperhardness and hypersoftness

Nalewajski has developed and presented a complete theory of chemical reactivity derived from DFT and from the information theory [5, 36]. However, hyperhardness, the global energy derivative of the third order over the number of electrons γ=∂η/∂Nv and other third energy derivatives, has rarely attracted attention in cDFT [37–39]. The first attempt for calculation γ has been provided by Fuentealba and Parr [40]. Ordon and Tachibana have published original approach to obtaining γ with the use of the nuclear reactivity indices and the Maximum Hardness Principle [41]. Later, another in-depth analysis of the local energy derivatives of higher orders has been presented in the work by Cardenas et al. [42] and the systematic and explicit approach to the whole body of cDFT energy derivatives was presented by Heidar-Zadeh et al. [43] However, no attempt to tracing a general way to their numerical values has been indicated. The widely recognized derivative proposed by Morell f2r≡∂fr∂Nv=∂2ρr∂N2v (the dual descriptor) plays a significant role within the theory of nuclear reactivity indices [44–46]. It also represents the third energy derivative and has been calculated by the finite difference method applied to the lower order derivatives [47–49]. f2r has been tested in many chemical applications ever since [50–52]. Another work by Cardenas et al. provided formal arguments for application of the dual descriptor as the useful tool for predicting the regioselectivity and nucleophilic attack simultaneously [53]. However, better predictive power of the softness/hypersoftness over FF/dual index has recently been advocated by Martinez-Araya [54]: “With the use of s(r) and s(2)(r) we have sufficient certainty that we are carrying out a more appropriate analysis since these two local reactivity descriptors are not affected by differences in size of the systems (…)”.

The local hypersoftness s2r≡∂sr∂μv=∂2ρr∂μ2v defined in the grand canonical ensemble has not yet been calculated for atoms, either free or bonded; analytical approach to this index for the hydrogenic orbitals has been presented in our previous paper [35].

Theoretical approach

Equation 2c allows for an alternative approach to the third energy derivatives. The basic equation for direct calculation of the local softness reads (Eq. 3): [25, 35]3 sr=-∇ρr·εrεr2

Hence:4 ds(r)dNv=-∇fr·εrεr2

This is transformed into working equation: [35]5 ds(r)dNv=-1S∇2ρrεr2

This result is computable once the global softness S has been determined at the preceding step by integration of the local softness (Eq. 3); the global derivative ∂S/∂Nv is obtained from Eq. 5 accordingly. Calculation of this global derivative is sufficient for finding the global hyperhardness [41]:6a γ=∂η∂Nv=-1S2∂S∂Nv

Hyperhardness is defined within the canonical ensemble (closed-system representation [42, 47]) with the energy function as canonical potential, dependent on the external potential and the number of electrons E[vr,N]. On the other hand, the definition of the hypersoftness reads:6b S(2)≡∂S∂μv=-∂3Ω∂μ3v

As the global softness itself, S(2) is defined in the grand canonical ensemble (open-system representation [42, 47]), where the chemical potential μ is an independent variable for the grand potential: Ω[v(r),μ]=E-μN [42, 47]. Since the variables N and μ are related, the hypersoftness S(2) may be expressed in both ensembles: [35]7 S(2)=∂S∂μv=∂S∂NvS

The relationship of hypersoftness with the hyperhardness follows by combining Eq. 7 with Eq. 6a.8 S(2)=-γS3

The local hypersoftness is defined and expressed accordingly:9 s(2)r≡∂sr∂μv=∂2ρr∂μ2v=∂sr∂NvS

It has been demonstrated in the preceding work that the dual descriptor can be expressed in a similar form: [35]10 f(2)r=1S∂sr∂Nv-frS∂S∂Nv

As evidenced by Eqs. 7–9, the computational access to the single derivative ∂sr∂Nv is sufficient for reproducing both the global hyperhardness γ and hypersoftness S(2); hence for both ensembles, we obtain the operating formulas for the whole collection of reactivity indices being the third-order energy derivatives [35].

cDFT analysis of the state functions

In the previous papers [25, 35], we have obtained analytical results for global parameters, vital for the description of the chemical reaction: ΔN, ΔE, and Δμ. They were calculated within the alchemical approach [55–57], up to the second order for both open and closed systems, using the operating approximations to the local softness (Eq. 3) and its derivative over N (Eq. 5) [35]. This has motivated the computational results of this present paper. With the use of sr and [∂sr/∂N]v, we have derived new formulas for hyperhardness and hypersoftness that come from Eq. 6a and Eq. 6b and we have tested them for atoms and ions.

The local approximation [25] and the Berkowitz-Parr formula [27] allowed for replacing the kernels by the local derivatives in the Taylor expansion into the derivatives over vr. Canonical ensemble expansion has been presented for energy ΔEN and chemical potential ΔμN at constant N (closed system), while for ΔNμ, the grand canonical ensemble/open system with the constraint of constant μ was appropriate. The results are: [35]11 ΔEN=∫ρ(r)Δv(r)dr-12∫s(r)Δv(r)2dr-1S∫s(r)Δv(r)dr2

12 ΔμN=1S∫s(r)Δv(r)dr++12-∫∂s(r)∂NμΔv(r)2dr++γ∫s(r)Δv(r)dr2++2S∫∂s(r)∂NμΔv(r)dr∫s(r′)Δv(r′)dr′

13 ΔNμ=-∫s(r)Δv(r)dr-S2∫∂sr∂NvΔv(r)2dr

The energy formula (Eq. 11) does not involve the third-order density derivatives in the expansion to the second order, since its first derivative over v(r) is the density itself.

Within the alchemical approach to atoms and atomic ions, the variation of the potential Δvr comes uniquely from the variation of the atomic number ΔZ [35]. The integrals necessary in Eqs. 11–13 are transformed and the brief notation for the five new necessary integrals has been introduced (Eqs. 14–19):14 ∫ρrΔvrdr=-ΔZ∫ρrrdr=-ΔZIρ

15 ∫s(r)Δv(r)dr=-ΔZ∫srrdr=-ΔZIs

16 ∫s(r)Δv(r)2dr=-ΔZ2∫srr2dr=-ΔZ2Is2

17 ∫∂sr∂NμΔvrdr=-ΔZ∫∂sr∂Nμ1rdr=-ΔZIN

18 ∫∂s(r)∂NμΔv(r)2dr=-ΔZ2∫∂s(r)∂Nμ1r2dr=-ΔZ2IN2

Three basic equations (Eqs. 11–13) may now be transformed to the explicit functions of the atomic number ΔZ.19 ΔEN=-ΔZIρ+12Is2+1SIs2ΔZ2=AEΔZ+BEΔZ2

20 ΔμN=-ΔZIsS+12IN2-γIs2+2SINIsΔZ2=AμΔZ+BμΔZ2

21 ΔNμ=ΔZIs+S2IN2ΔZ2=ANΔZ+BNΔZ2

The significance of sr and [∂sr/∂N]v is clearly exposed in Eqs. 11–13: they both are sufficient to obtain all necessary integrals (Eqs. 14–18). The abbreviated symbols for various integrals within the alchemical approach have been introduced for the sake of clarity (Eqs. 14–18). The results for ΔEN, ΔμN, and ΔNμ are all parabolic functions of ΔZ with the local hypersoftness function contributing to the variation of each state function in the second order only. The significance of this term for the quantitative results of ΔEN, ΔμN, and ΔNμ has been analyzed in detail for the group of atoms representing the highest values of global softness S and global hypersoftness and [∂S/∂N]v.

Numerical results

Computational results of this present work have been obtained for the 36 atoms (1–4 row of the periodic table) and corresponding cations and anions [25].

Computational methods

Numerical analysis was executed with the Gaussian 16 code [58]. The B3LYP method has been chosen, following its former successful tests in the electron density calculations in atoms [59–61]. The aug-cc-pvqz basis set has been routinely applied, except for the Ca atom (cc-pvqz). Using the pVTZ basis set recommended by Sadlej was necessary for the potassium atom [62]. The ground states of atoms have been identified by the specification of atomic electronic terms [25, 63].

The numerical DFT method has been used to calculate the electron density and electron density gradient to overview the local softness of atoms and ions. The effect of degeneracy of frontier orbitals has been avoided by using the integral electron density for every atom and ion. As it has been proved by Kohn, the ground state density is unique even for systems, where frontier orbitals are degenerate [64, 65]. The simple practical method has been applied to circumvent another well-known difficulty in reproducing spherical symmetry of the electron density in atoms. Two steps have been involved in this procedure. Integration of the electron density has been routinely made by the spherical algorithm for a density variable in one direction only. The result of integration to proper number of electrons served as evidence of a spherical symmetry of the density. For atoms whose basic electronic terms were other that S type, the numerical averaging procedure over principal coordinate axis and all diagonal directions has been applied to the raw computational results. The subsequent integration provided proof for the sufficient quality (symmetry) of the averaged density for the purpose of this study. The density gradient has only been calculated in one direction from the spatially averaged numerical density data. Equation 3 and Eq. 5 serve as basic computational algorithms.

The integration of the electron density and its derivatives has been obtained with the grid of the density function. The accuracy of the grid was controlled by the integration of the resulting radial distribution of the density to the proper number of electrons. The grid for atoms/ions in the 1st and 2nd periods was 0.05ao with the integration radius 15 ao. For the 3rd and 4th periods, the grid was 0.02 ao within the radius 20 ao.

In the traditional cDFT formalism, the global derivatives of energy over N have been expressed in units of energy, typically in eV. This comes from the basic formulation of the electron density, hence also the number of electrons as dimensionless and this method makes the formalism clear. However, it has been noted by March that it leads to non-physical sign of the energy ( +) while the binding energy of any system of electrons and nuclei is negative. The correction, by introducing the charge factor with the electron density: -eρr, had been proposed by March [66, 67], following the earlier analysis by Feynman [13]. The units of electric field are clearly V/m and the local approximation leading to working equation (Eq. 2) does not affect the physical meaning of the local softness s(r). By integrating s(r) to global softness S, the physical unit for global softness is disclosed: S [e/V], equivalent to the units of capacity [C/V] = [F] = [m]. The unit of the same origin had once been proposed within the chemical approximation to the basic cDFT concepts [68–72]. Neglecting the corrections indicated by Feynman and March would lead to global softness in [V−1], hardly physical, in terms of the meaning of this quantity [73].

The above conclusion calls for reconciliation between the physical rigor and the widely accepted practice in cDFT. It does not seem reasonable to abandon the custom of expressing the chemical potential, global hardness (and further global derivatives in the canonical ensemble) in the specific units of energy (eV); their conversion to a.u. is evident. Therefore, since the units for softness have not yet been consolidated, it seems practical to propose keeping S and its derivatives in atomic units; the inconsistence between units of hardness and softness becomes immaterial in a.u.: Sau=ηeV27.211[eV/au]-1. This method has been applied in the presentation of the results of this work.

Local hypersoftness calculated for atoms and ions

The basic derivatives are presented in Figs. 1 and 2: ∂sr/∂Nv=s2r/S and the dual descriptor f(2)r=[df(r)/dN]v for atoms in rows 2 and 3 of the periodic table, respectively. The derivatives are interconnected by Eq. 10.Fig. 1 A Radial distribution of the derivative dsrdNvfor atomsin a.u. (Eq. 5). B Radial distribution of the dual descriptor f(2)r=[dfr/dN]v for atomsin a.u. (Eq. 10). Note the expansion of the ordinate axis by 1:4 ratio

Fig. 2 A Radial distribution of the derivatives [∂sr/∂N]v for atoms in a.u. (Eq. 5). B Radial distribution of the dual descriptor f(2)r=[∂fr/∂N]vfor atoms in a.u ( Eq. 10)

The diagrams of the [∂sr/∂N]v functions for 36 elements are presented in the Supporting Material, together with the diagrams of the dual descriptors of atoms; results for cations and anions have also been included. Except the rows 1 and 2, the derivatives [∂sr/∂N]v tend to be flat functions reaching to considerably greater distance than the local softness functions [25]. They all span positive values, except for narrow area near the nucleus, where shallow minima are typically observed (Fig. 1A and Fig. 2A). This represents the meaningful feature as compared to the dual descriptor, the analogous function that is integrated to zero, hence showing the minima and maxima of equal importance (Fig. 1B and Fig. 2B). The global values [∂S/∂N]v resulting from integration are all positive numbers, typically reaching no more than 30% of the global softness (also positive), except the very light elements (rows 1 and 2), cf. Figure 5A and Fig. 6A and 6.

Global hyperhardness γ and hardness η

Hyperhardness for atoms and ions is calculated by Eq. 6a. The comparison of hardness (positive values) and hyperhardness (negative values) has been presented as two functions of the atomic number in joint diagrams for 36 neutral atoms (Fig. 3) and for their cations (Fig. 4A) and anions (Fig. 4B). The range on the ordinate axis has been adjusted to expose the details of the diagrams. Several common di-cations and di-anions are also included in Fig. 4A and Fig. 4B, respectively. They provide an additional quick test for the rational character of the results: hardness increasing with further ionization for cations, decreasing for anions.Fig. 3 Hardness (η = 1/S) and hyperhardness (γ, Eq. 6) for atoms calculated from electron density in eV

Fig. 4 Hardness (η = 1/S) and hyperhardness (γ, Eq. 6) for cations (A) and for anions (B) in eV. The common di-cations and di-anions have also been marked by the green diamonds

Global hypersoftness

Calculated derivative [∂S/∂N]v has been confronted with the global hypersoftness sensu stricto S(2) (Eq. 8) in Fig. 5A and Fig. 5B, respectively. The rough linear correlation between both quantities has been observed for of atoms and cations; the effect results from the interrelation between these two derivatives (Eq. 7). The global softness parameter linking both derivatives spans only modest range of 3 ÷ 6 a.u. for atoms and 1–3 a.u. for cations and it is slowly lowering with increasing Z. (In the group of anions, with S parameter 5–20 a.u., the effect is considerably less pronounced, though still observed.) For practical reasons, only the derivative [∂S/∂N]v has been presented for cations (Fig. 6A) and for anions (Fig. 6B). Calculated [∂S/∂N]v data for atoms have all been included with the diagrams available in the Supporting Material.Fig. 5 A [∂S/∂N]v derivative for atoms in a.u. B Global hypersoftness for atoms S(2) (Eq. 6. in a.u.). Note the contraction of the ordinate axis by 5:1 ratio

Fig. 6 A [∂S/∂N]v derivative calculated in a.u. for cations. B [∂S/∂N]v derivative calculated in a.u. for anions. Note the contraction of ordinate axis by 5:1 ratio. Results for di-cations (in A) and di-anions (in B) have also been included (green diamonds) whenever substantially different from the univalent cation

Variations of the state functions ΔEN, ΔμN, ΔNμ

The local softness s(r), its derivative ∂sr/∂Nv, and the global softness S available for atoms allow for testing the possible sensitivity of atoms/ions in the model conditions available in the alchemical analysis. Calculations have been limited to elements in rows 1 and 2 of the periodic table, as the well-recognized variations in electronic properties of atoms in this group allow for critical evaluation of the results. The integrals in Eqs. 14–18 have been calculated first (Table 1). In Table 2, the calculated parameters of the parabolic functions of ΔZ (A and B in Eqs. 19–21) have been collected. Table 1 Calculated values of integrals introduced in Eqs. 14–18 (a.u.)

Atom	Z	Iρ	Is	Is2	IN	IN2	
H	1	1.00	2.86	1.90	0.0259	0.0000237	
He	2	3.36	2.76	3.08	0.0353	0.00196	
Li	3	5.67	2.74	3.34	0.0141	0.0023	
Be	4	8.32	2.72	3.54	0.0123	0.00418	
B	5	11.20	2.70	3.70	0.0116	0.0066	
C	6	14.37	2.68	3.85	0.0115	0.00964	
N	7	17.81	2.66	4.00	0.0118	0.0131	
O	8	21.47	2.64	4.14	0.012	0.0167	
F	9	25.39	2.62	4.28	0.0122	0.0206	
Ne	10	29.56	2.62	4.42	0.0127	0.0244	

Table 2 Calculated coefficients for the relations: ΔEN(ΔZ), ΔμNΔZ, ΔNμΔZ in Eqs. 19–21 (a.u.). The necessary values for integrals are listed in Table 1; global softness S and hyperhardness γ are from the table in the Appendix

Atom	ΔEN =AEΔZ+BEΔZ2 (Eq. 19)	ΔμN =AμΔZ+BμΔZ2 (Eq. 20)	ΔNμ =ANΔZ+BNΔZ2 (Eq. 21)	
AE	BE	Aμ	Bμ	AN	BN	
H	 − 1.00	1.26	 − 0.490	33.0	2.86	0.0000690	
He	 − 3.36	2.90	 − 0.793	27.0	2.76	0.00341	
Li	 − 5.67	2.56	 − 0.436	15.5	2.74	0.00723	
Be	 − 8.32	2.86	 − 0.471	10.4	2.72	0.0120	
B	 − 11.2	3.18	 − 0.524	8.61	2.70	0.0170	
C	 − 14.4	3.57	 − 0.596	7.73	2.68	0.0217	
N	 − 17.8	4.02	 − 0.673	7.20	2.66	0.0259	
O	 − 21.5	4.48	 − 0.742	6.85	2.64	0.0297	
F	 − 25.4	4.98	 − 0.814	6.58	2.62	0.0332	
Ne	 − 29.6	5.56	 − 0.897	6.44	2.62	0.0356	

Discussion

The results presented in this work are all founded on the extension of our gradient theorem (Eq. 1 and Eq. 2c). On the first step of this work, the softness of atoms and ions has been calculated from the electron density only. This is well in contrast to the standard method introduced by Parr and Pearson and founded on the energy parameters exclusively. The efforts to find another source of the atomic hardness parameters have long history [68–70]. The problem has been analyzed in depth by Ayers [74]; the old relations between the size and polarizability and softness/hardness have been reminded and many more arguments have been raised, leading to the conclusion that the original concept η ∝ I-A might be reconsidered. A solution has been proposed in the following papers by Cardenas et al. who introduced the benchmark values of the chemical hardness for atoms [75, 76]. These values are also based on the energy parameters exclusively; nevertheless, they represent most reasonable set of data to which our results might be compared. This has been demonstrated in Fig. 7A in the η(Z) relationship that shows the asymptotic behavior of the index resulting from this work. The change in hardness of atoms in rows 1 and 2 is parallel; in periods 3 and 4, this trend is less pronounced, but still visible. The turning points at completion of the electronic shells are reproduced in both approaches. In period 4 (Z > 18), the hardness parameters from the density (this method) are gradually increasing, while the benchmark data undergo stepwise variations with no actual explanation.Fig. 7 A Hardness of atoms η = S−1 (eV) calculated by this method (ref. 25) as compared to the benchmark values (ref. 76). B Hyperhardness of atoms (eV): γ—this work; 6γF by Fuentealba and Parr (ref. 40)

The review of global hardness data (η) for atomic cations and anions, as compared to those for atoms, provides the support for the coherence of the results (Fig. 3 and Fig. 4). The anions are all considerable softer that the neutral atoms (the di-anions are even softer still), while cations are all harder. With the increase of the atomic number, the difference between hardness of ions vs atoms is lowered.

The pioneering results for hyperhardness by Fuentealba and Parr γF=1/6∂3E/∂N3v have been obtained by the sophisticated interpolation based on three experimental parameters: electron affinity A1 and ionization energies I1 and I2 [40]. The hyperhardness γF was found to be negative, except for the atoms in their S electronic state (e.g., noble gases), where exceedingly small, positive γF>0 was found. While no correlation between the present results (γ) and 6 γF was observed (Fig. 7B), both values span over similar numerical data with neatly equivalent average values: < 6γF >  =  − 1.86 eV, < γ >  =  − 1.77 eV. This observation is meaningful, given the much different origin of the two sets of data. The notable advantage of the method applied hereby is rather regular change in γ(Z) function equally for neutral atoms (Fig. 3) and their ions (Fig. 4) whereas the dispersed 6γF(Z) data could not be fitted with any feasible rational interpolation (Fig. 7B).

The applied method of calculation of the global energy derivatives (η, γ) from the properties of the stationary electron density alone represents an explicit example that these indices are density functionals and that electron density is the basic variable in cDFT. The well-known problems with the degeneracy of the electronic states that invalidates the usual reactivity indicators associated with the cDFT have been circumvented [77, 78]. Although this approach leads merely to a lower bound to the response parameters, the simple availability of the profound DFT descriptor may be valuable for the chemical practice, once the phenomenological method of their use is available in calculation of the variations of the state functions ΔEN, ΔμN, ΔNμ. Recent efforts by Miranda-Quintana et al. may bring a fresh interest in this matter [79, 80], and the problem of indexing the chemical properties by the density-based derivatives has also been focused upon in other recent papers [81, 82]. The interplay between the electric field and the electron density function has been analyzed in detail in the classical work by Liu and Hunt, where the set of new connections between molecular softness and the electromagnetic properties of molecules has been analyzed in detail [83]. Specifically, the role of the local softness and hypersoftness to long-range intermolecular forces has been exposed.

The hyperhardness calculated in the framework of the local approximation for atoms increases along the rows of the periodic table, but it tends to be rather smooth flat function of the atomic numbers for Z > 10 within their periods. The hardness data from the same method show distinct discontinuities at critical points for the electron number, corresponding to complete filling consecutive electronic shells (Fig. 3: Z = 2, 10, 18, 25). The hyperhardness for cations (Fig. 4A) is smaller as compared to the hardness of atoms except the lightest atoms. The contribution from hyperhardness may have considerable effect for anions (Fig. 4B), with absolute value of hyperhardness γ reaching ca. 10% of hardness itself.

The numerical values of the local function [∂sr/∂N]v for light atoms (Z = 1 ÷ 10) are considerably smaller than the values of the local softness sr and their orbital structure is less exposed (cf. Supporting Material). For heavier atoms (Z > 10), the derivative [∂sr/∂N]v becomes negligibly small as compared to the local softness. Notably, they span much broader region around atoms than the local softness functions and their role for many anions may not be neglected; complete integration of these functions could not be achieved in the standard grid size [0 ÷ 20 a. u.].

The alchemical approach methodology [84–86] allows for assessing the role of softness and hypersoftness for the energy E and chemical potential μ for a closed system (polarization effect) and for the possible flow of charge ΔNμ to/from an atom and a reservoir at constant μ (an open system), the handy model for an electron exchange in a chemical reaction. The role of derivatives representing the susceptibilities of atoms to a disturbance of the external potential has been determined, by calculating the integrals involved in the expansion (Eqs. 14–18 and Table 1). Integrals IN and IN2 involving the local function [∂sr/∂N]v (Eq. 17 and Eq. 18) are small as compared with other entries in this table. They appear to have little effect for the state functions, as their role for the coefficient Bμ and BN is marginal. However, the disturbance in the second order cannot be entirely neglected, except for ΔNμ(ΔZ) (Eq. 21). Calculated parameters for the ΔENΔZandΔμNΔZ relationships presented in Table 2 prove the role of the second-order derivatives, since the local softness s(r) plays quite significant role here. The second-order effect provides large term of the energy change and has been proved to dominate the effect Δμ in the chemical potential (at constant N). Only for the charge flow (at constant μ) it is limited to the first-order effect.

The local approximation systematically explored in this work allows also to estimate the role of the third-order effects on the energy ΔEN, associated with the disturbance of the electric field on the electron density of an atom. The third derivative of the energy E[v(r), N] over v is transformed to the density derivative, as demonstrated in the work by Heidar-Zadeh et al. [43] E′′′=δω(r1′r2)δv(r3)N. By exploring the Berkowitz and Parr relation followed by the local approximation, the final result for Δ(3)EN effect involves the same set of integrals as listed in Eqs. 15–18, with a dominant role of the IN and IN2 that have been proved to play only the minor role (Table 1). The complete result is presented in the Appendix. At the level of alchemical approach adopted in this work, including the third-order correction to the energy expansion leads to the supplemented ΔEN(ΔZ) function of the second order.

The characteristic feature of the ΔN(ΔZ) relationship represented by its slope for the limit of ΔZ→0 was found to be unique for every orbital ΔN/ΔZ→3a.u. [35]. The results for actual atoms were found reasonably close to this prediction ΔN/ΔZ=>AN (Table 2), with 2.62 < AN < 2.86 (a.u.). The difference could be attributed to the effect of electron–electron interaction, as the prediction for orbitals was for single occupied orbitals only.

These results help to understand the effects of the actual reactions in some indirect way. In the alchemical method, the disturbing potential leading to changes in a system originates from a hypothetical change ΔZ that occurs at the center of symmetry of the electron density. The response to this disturbance involves preferentially the inner part of the electron density [55]. This is clearly observed by the integrals Iρ in Table 1: 99% of their values come from the electron density within a sphere of radius not exceeding 5 a.u. (Li). In a real situation, the disturbing potential acts from a distance to an atom or molecule and the outer sphere of the electron density is affected. In this region, the contribution from the local softness s(r) will also be dominating the consequences of such disturbances. The role of hypersoftness [∂sr/∂N]v that contributes to the integrals IN and IN2 (Eq. 17 and Eq. 18) will be exposed by the strengths of the external field dominating in the outer sphere of an atom and by the large volume of integration. The effects of confinement on a reaction are evident field for testing the role of this parameter [87]. This may be especially important for calculation of the chemical effect given by the charge flow (at constant μ) ΔNμ, hence reaching further than the predictions exploring the dual descriptor, defined for the closed system (canonical ensemble).

Calculated derivatives of the local softness [∂s(r)/∂N]v allow also for calculation of the new version of the double descriptor f(2)(r) via Eq. 10. Results for f(2)(r) in atoms in the 2nd and 3rd periods are demonstrated in Figs. 1B and 2B. The results are systematically negative at short distances from a nucleus and positive at the outer region around a nucleus; they fall to zero at infinity and they do integrate to zero, as required. However, unlike the original double finite-difference quantities Δf(r) [48], the local derivatives f(2)(r) themselves are not expected to reflect the susceptibility of an atom to a nucleophilic or an electrophilic attack, neither by their sign, nor by their value, as it has been claimed for Δf(r) descriptors in the original work of the authors [49]. The potential prognostic value of the present method is limited to the observables calculated for a given perturbance of the electric field in the space of an atom in electronic ground state: ΔEN—interaction energy at constant N; ΔμN—the change in the chemical potential at constant N; ΔNμ—the modification of the population at constant chemical potential.

The fundamental requirement introduced by Eq. 2c is that the density gradient vector and electric field vector are parallel in every point in space. This condition is obviously fulfilled for free atoms and their ions. According to our preliminary computations, this condition is also met for the valence region of a molecule. On the other hand, the gradient theorem (Eq. 1) holds only for the density resulting from computational methods that comply to both Hellmann–Feynman theorem and Feynman electrostatic expression for atomic force [13]. The vital role of this condition for the accuracy of the quantum chemical computations has been discovered and discussed in other works from this laboratory [88, 89].

Conclusion

The method described in this paper opens a new way to exploration of the vital cDFT derivatives: the linear response function ωr,r′ and the softness kernel sr,r′. Applications ωr,r′ in chemistry have been subject of many efforts [90, 91], hampered by the nonlocal character of this derivative. The Fukui function f(r) and the dual descriptor f(2)(r), another energy derivative of the third order, have been found more promising for chemical use, due to their local character [51, 53]. This present work has demonstrated that the non-local response functions may be replaced (to some extent) by exploring the local derivatives of higher order. This resulted not only in calculations of the local descriptors, but also the method for their phenomenological use has been proposed: calculation of the disturbances of the state functions ΔEN, ΔμN, ΔNμ. While the first two describe merely the polarization of a system due to the external potential, ΔNμ describes the truly chemical effect: the charge flow. The nonlocal kernels ωr,r′ and sr,r′ have not been neglected but replaced by the higher order local derivative: [∂s(r)/∂N]v. Calculation thereof was then necessary for the completeness of the method.

The novel local approach for calculation of the basic derivative of the local softness [∂s(r)/∂N]v enables to derive analytical formulas for cDFT reactivity indices of higher order. The integral of [∂s(r)/∂N]v gives the meaningful positive value of [∂S/∂N]v that opens an access to the canonical global energy derivatives of the third order: the hyperhardness (Eq. 22a) and the hypersoftness (Eq. 22b):22a γ=∂3E/∂N3v=-S-2[∂S/∂N]v

22b S(2)=∂3Ω/∂μ3v=S[∂S/∂N]v

Calculated data for γ in atoms have generally reproduced the first suggested results obtained by the inconsistent extrapolation of the energy function E(N) for atoms; the difficulties and arbitrary assumptions of the former method have been avoided. The radial distribution of [∂s(r)/∂N]v derivative (integrated to positive global [∂S/∂N]v) is characteristically different from the analogous function that has also been calculated: the dual descriptor f(2) (integrated to zero). The detailed results are presented in the Supporting Material available with this paper.

Application of 1st and 2nd density derivatives s(r) and [∂s(r)/∂N]v confirmed the vital role of the local softness s(r) for a response of the energy (ΔEN) and the chemical potential (ΔμN) of an object (at constant N) to perturbation by an external potential Δv(r), and also for the electron flow between an atom and an external reservoir (at constant μ)—ΔNμ. The tiny contributions from hypersoftness in these studies are well understood at the level of the alchemical analysis; it may play a role in studies focused on the outer space of an atom. Calculated results for the state functions represent the promising application for the results of the axiomatic approach.

Supplementary Information

Below is the link to the electronic supplementary material.Supplementary file1 The visual collection containing the radial distribution diagrams is available in form of the pdf presentation. It contains visualization of the derivatives: s(r),dsr/dNv ,and   f(r), f2r for 36 atoms in rows 1÷4 of the periodic table; their cations and anions (selected di-cations and di-anions) have also been included. (PDF 4602 KB)

Appendix

The third-order component to the energy change Δ(3)EN under the rigor exposed in ref. 43 is:A.1 Δ(3)EN=∫∫∫δω(r1,r2)δv(r3)NΔv(r1)Δv(r2)Δv(r3)dr1dr2dr3

Exploring the Berkowitz and Parr relation and the local approximation allows for transformation of Eq. A.1 leads to:Δ(3)EN=-∫∫δs(r1)δv(r3)NΔv(r1)2Δv(r3)dr1dr3+

A.2 +∫∫∫δ{Sf(r1)f(r2)}δv(r3)NΔv(r1)Δv(r2)Δv(r3)dr1dr2dr3

The following expansions are available under the cDFT scrutiny for the derivatives to be integrated: [92]A.3 δs(r1)δv(r3)N=-s(r1)∂s(r3)∂Nv+f(r1)s(r3)∂S∂Nv+S∂ω(r1,r3)∂Nv

A.4 δ[Sf(r1)f(r2)]δv(r3)N=-f(r1)f(r2)S∂s(r3)∂Nv+s(r3)∂S∂Nv++Sf(r2)∂ω(r1,r3)∂Nv+Sf(r1)∂ω(r2,r3)∂Nv

The integration of the derivative of the linear response function over N has been presented in our earlier work [35].A.5 ∫∫∂ω(r,r′)∂NvΔv(r)Δv(r′)drdr′=-∫∂s(r)∂NvΔv(r)2dr++γ∫s(r)Δv(r)dr2++2η∫∂s(r)∂NvΔv(r)dr∫s(r′)Δv(r′)dr′

With the use of Eqs. A.3, A.4, and A.5, the integration in Eq. A.1 contains the single integrals exclusively, with the integrands provided by the softness (local and global) and their derivatives over N, now avaiable by the method presented in this paper.

Acknowledgements

JZ acknowledges his friendly cooperation with Mr Mateusz Jędrzejewski. The use of resources of Wrocław Center for Networking and Supercomputing in these projects is acknowledged (WCSS Grants No. 249 and GW 036). PO was supported by Wrocław University of Environmental and Life Science (Poland) under the project POMOST, grant N110/0004/21. This manuscript has been published as preprint: ChemRxiv. 10.26434/chemrxiv-2024-0vttl.

Author contribution

JZ— prepared numerical results, figures and supplementary material PO—concept, theoretical analysis, LK—concept, wrote the manuscript All authors reviewed the manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing Interests

The authors declare no competing interests.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References

1. Parr RG Donnelly RA Levy M Palke WE Electronegativity: the density functional viewpoint J Chem Phys 1978 68 3801 3807 10.1063/1.436185
Parr RG, Donnelly RA, Levy M, Palke WE (1978) Electronegativity: the density functional viewpoint. J Chem Phys 68:3801–3807
2. Hohenberg P Kohn W Inhomogeneous electron gas Phys Rev 1964 136 8864 8871 10.1103/PhysRev.136.B864
Hohenberg P, Kohn W (1964) Inhomogeneous electron gas. Phys Rev 136:8864–8871
3. Perdew JP Parr RG Levy Mel Balduz JL Jr Density-functional theory for fractional particle number: derivative discontinuities of the energy Phys Rev Lett 1982 49 1691 1694 10.1103/PhysRevLett.49.1691
Perdew JP, Parr RG, Levy Mel, Balduz JL Jr (1982) Density-functional theory for fractional particle number: derivative discontinuities of the energy. Phys Rev Lett 49:1691–1694
4. Parr RG Pearson RG Absolute hardness: companion parameter to absolute electronegativity J Am Chem Soc 1983 105 7512 7516 10.1021/ja00364a005
Parr RG, Pearson RG (1983) Absolute hardness: companion parameter to absolute electronegativity. J Am Chem Soc 105:7512–7516
5. Nalewajski RF A coupling between the equilibrium state variables and open molecular and reactive systems Phys Chem Chem Phys 1999 1 1037 1049 10.1039/a806523c
Nalewajski RF (1999) A coupling between the equilibrium state variables and open molecular and reactive systems. Phys Chem Chem Phys 1:1037–1049
6. Geerlings P Chamorro E Chattaraj PK De Proft F Gázques JL Liu S Morell C Toro-Labbé A Vela A Ayers PW Conceptual density functional theory: status, prospects, issues Theor Chem Acc 2020 139 36 10.1007/s00214-020-2546-7
Geerlings P, Chamorro E, Chattaraj PK, De Proft F, Gázques JL, Liu S, Morell C, Toro-Labbé A, Vela A, Ayers PW (2020) Conceptual density functional theory: status, prospects, issues. Theor Chem Acc 139:36
7. Ayers PW Fias S Heidar-Zadeh F The axiomatic approach to chemical concepts Comp Theor Chem 2018 1142 83 87 10.1016/j.comptc.2018.09.006
Ayers PW, Fias S, Heidar-Zadeh F (2018) The axiomatic approach to chemical concepts. Comp Theor Chem 1142:83–87
8. Ayers PW Atoms in molecules, an axiomatic approach. I. Maximum transferability J Chem Phys 2000 113 10886 10898 10.1063/1.1327268
Ayers PW (2000) Atoms in molecules, an axiomatic approach. I. Maximum transferability. J Chem Phys 113:10886–10898
9. Ayers PW Strategies for computing chemical reactivity indices Theor Chem Acc 2001 106 271 279 10.1007/PL00012385
Ayers PW (2001) Strategies for computing chemical reactivity indices. Theor Chem Acc 106:271–279
10. Ordon P, Komorowski L (2023) Reaction fragility method: monitoring evolution of atoms and bonds on a reaction path, [Chapter 9 in:] Chemical reactivity vol. 1: theories and principles; [Kaya S, Von Szentpaly L, Serdaroglu G, Guo L, Editors.]; Elsevier
11. Ordon P Komorowski L Jędrzejewski M Zaklika J The connectivity matrix: a toolbox for monitoring bonded atoms and bonds J Phys Chem A 2020 124 1076 1086 10.1021/acs.jpca.9b10145 31962040
Ordon P, Komorowski L, Jędrzejewski M, Zaklika J (2020) The connectivity matrix: a toolbox for monitoring bonded atoms and bonds. J Phys Chem A 124:1076–108631962040
12. Ordon P Komorowski L Jędrzejewski M Conceptual DFT analysis of the fragility spectra of atoms along the minimum energy reaction coordinate J Chem Phys 2017 147 134109 10.1063/1.4995028 28987090
Ordon P, Komorowski L, Jędrzejewski M (2017) Conceptual DFT analysis of the fragility spectra of atoms along the minimum energy reaction coordinate. J Chem Phys 147:13410928987090
13. Feynman RP Forces in molecules Phys Rev 1939 56 340 343 10.1103/PhysRev.56.340
Feynman RP (1939) Forces in molecules. Phys Rev 56:340–343
14. King WT Zelano AJ Sum rule for molecular frequencies J Chem Phys 1967 46 3197 3199 10.1063/1.1712374
King WT, Zelano AJ (1967) Sum rule for molecular frequencies. J Chem Phys 46:3197–3199
15. King WT Calculation of molecular force constants J Chem Phys 1968 49 2866 2867 10.1063/1.1670505
King WT (1968) Calculation of molecular force constants. J Chem Phys 49:2866–2867
16. King WT Calculation of atomic force constants from electron densities J Chem Phys 1972 57 4535 4539 10.1063/1.1678111
King WT (1972) Calculation of atomic force constants from electron densities. J Chem Phys 57:4535–4539
17. Komorowski L Ordon P Jędrzejewski M The reaction fragility spectrum Phys Chem Chem Phys 2016 18 32658 32663 10.1039/C6CP06519H 27892571
Komorowski L, Ordon P, Jędrzejewski M (2016) The reaction fragility spectrum. Phys Chem Chem Phys 18:32658–3266327892571
18. Zaklika J Komorowski L Ordon P The bond fragility spectra for the double proton transfer reaction, in the formic acid type dimers J Phys Chem A 2019 123 4274 4283 10.1021/acs.jpca.9b00595 31008601
Zaklika J, Komorowski L, Ordon P (2019) The bond fragility spectra for the double proton transfer reaction, in the formic acid type dimers. J Phys Chem A 123:4274–428331008601
19. Zaklika J Komorowski L Ordon P Evolution of the atomic valence observed by the reaction fragility spectra on the reaction path J Mol Model 2019 25 134 10.1007/s00894-019-4029-0 31028500
Zaklika J, Komorowski L, Ordon P (2019) Evolution of the atomic valence observed by the reaction fragility spectra on the reaction path. J Mol Model 25:13431028500
20. Ordon P Zaklika J Jȩdrzejewski M Komorowski L Bond softening indices studied by the fragility spectra for proton migration in formamide and related structures J Phys Chem A 2020 124 328 338 10.1021/acs.jpca.9b09426 31815477
Ordon P, Zaklika J, Jȩdrzejewski M, Komorowski L (2020) Bond softening indices studied by the fragility spectra for proton migration in formamide and related structures. J Phys Chem A 124:328–33831815477
21. Cremer D Wu A Kraka E The mechanism of the reaction FH+H2C=CH2→H3C–CFH2. Investigation of hidden intermediates with the unified reaction valley approach Phys Chem Chem Phys 2001 3 674 687 10.1039/b007733j
Cremer D, Wu A, Kraka E (2001) The mechanism of the reaction FH+H2C=CH2→H3C–CFH2. Investigation of hidden intermediates with the unified reaction valley approach. Phys Chem Chem Phys 3:674–687
22 Kraka E Zou W Tao Y Decoding chemical information from vibrational spectroscopy data: local vibrational mode theory WIREs Comput Mol Sci 2020 10 e1480 10.1002/wcms.1480
Kraka E, Zou W, Tao Y (2020) Decoding chemical information from vibrational spectroscopy data: local vibrational mode theory. WIREs Comput Mol Sci 10:e1480
23. Politzer P Muray J A look at bonds and bonding Struct Chem 2019 30 1153 1157 10.1007/s11224-019-01364-3
Politzer P, Muray J (2019) A look at bonds and bonding. Struct Chem 30:1153–1157
24. Zhao L Zhi M Frenking G The strength of a chemical bond Int J Quant Chem 2022 122 8 e26773 10.1002/qua.26773
Zhao L, Zhi M, Frenking G (2022) The strength of a chemical bond. Int J Quant Chem 122(8):e26773
25. Zaklika J Hładyszowski J Ordon P Komorowski L From the electron density gradient to the quantitative reactivity indicators: local softness and the Fukui function ACS Omega 2022 7 7745 7758 10.1021/acsomega.1c06540 35284764
Zaklika J, Hładyszowski J, Ordon P, Komorowski L (2022) From the electron density gradient to the quantitative reactivity indicators: local softness and the Fukui function. ACS Omega 7:7745–775835284764
26. Bader RFW (1994) Atoms in molecules. A Quantum theory. A Clarendon Press Publication: International Series of Monographs in Chemistry, vol. 22
27. Berkowitz M Parr RG Molecular hardness and softness, local hardness and softness, hardness and softness kernels, and relations among these quantities J Chem Phys 1988 88 2554 2557 10.1063/1.454034
Berkowitz M, Parr RG (1988) Molecular hardness and softness, local hardness and softness, hardness and softness kernels, and relations among these quantities. J Chem Phys 88:2554–2557
28. Fias S Ayers PW De Proft F Geerlings P Properties of the density functional response kernels and its implications on chemistry J Chem Phys 2022 157 11 114102 10.1063/5.0094653 36137804
Fias S, Ayers PW, De Proft F, Geerlings P (2022) Properties of the density functional response kernels and its implications on chemistry. J Chem Phys 157(11):11410236137804
29. Senet P Nonlinear electronic responses, Fukui functions and hardness’s as functionals of the ground-state electronic density J Chem Phys 1996 105 6471 6489 10.1063/1.472498
Senet P (1996) Nonlinear electronic responses, Fukui functions and hardness’s as functionals of the ground-state electronic density. J Chem Phys 105:6471–6489
30. Ayers PW Parr RG Beyond electronegativity and local hardness: higher-order equalization criteria for determination of a ground-state electron density J Chem Phys 2008 129 054111 10.1063/1.2957900 18698892
Ayers PW, Parr RG (2008) Beyond electronegativity and local hardness: higher-order equalization criteria for determination of a ground-state electron density. J Chem Phys 129:05411118698892
31. Prodan E Kohn W Nearsightedness of electronic matter Proc Natl Acad Sci USA 2005 102 11635 11638 10.1073/pnas.0505436102 16087868
Prodan E, Kohn W (2005) Nearsightedness of electronic matter. Proc Natl Acad Sci USA 102:11635–1163816087868
32. Fias S Heidar-Zadehb F Geerlings P Ayers PW Chemical transferability of functional groups follows from the nearsightedness of electronic matter Proc Natl Acad Sci USA 2017 114 11633 11638 10.1073/pnas.1615053114 29078266
Fias S, Heidar-Zadehb F, Geerlings P, Ayers PW (2017) Chemical transferability of functional groups follows from the nearsightedness of electronic matter. Proc Natl Acad Sci USA 114:11633–1163829078266
33. Bader RFW Nearsightedness of electronic matter as seen by a physicist and a chemist J Phys Chem A 2008 112 13717 13728 10.1021/jp806282j 19032142
Bader RFW (2008) Nearsightedness of electronic matter as seen by a physicist and a chemist. J Phys Chem A 112:13717–1372819032142
34. Vela A Gazquez JL A relationship between the static dipole polarizability, the global softness, and the Fukui function J Am Chem Soc 1990 112 1490 1492 10.1021/ja00160a029
Vela A, Gazquez JL (1990) A relationship between the static dipole polarizability, the global softness, and the Fukui function. J Am Chem Soc 112:1490–1492
35. Ordon P Zaklika J Hładyszowski J Komorowski L Analytical approximation to the local softness and hypersoftness and to their applications as reactivity indicators J Chem Phys 2023 158 17 174110 10.1063/5.0142656 37132529
Ordon P, Zaklika J, Hładyszowski J, Komorowski L (2023) Analytical approximation to the local softness and hypersoftness and to their applications as reactivity indicators. J Chem Phys 158(17):17411037132529
36. Nalewajski RF Perspectives in electronic structure theory 2012 Berlin Heidelberg Springer-Verlag
Nalewajski RF (2012) Perspectives in electronic structure theory. Springer-Verlag, Berlin Heidelberg
37. Parr RG Yang W Density functional theory of atoms and molecules 1989 Oxford, UK Oxford University Press
Parr RG, Yang W (1989) Density functional theory of atoms and molecules. Oxford University Press, Oxford, UK
38. Chermette H Chemical reactivity indexes in density functional theory J Comput Chem 1999 20 129 154 10.1002/(SICI)1096-987X(19990115)20:1<129::AID-JCC13>3.0.CO;2-A
Chermette H (1999) Chemical reactivity indexes in density functional theory. J Comput Chem 20:129–154
39. Geerlings P De Proft F Langenaeker W Conceptual density functional theory Chem Rev 2003 103 1793 1874 10.1021/cr990029p 12744694
Geerlings P, De Proft F, Langenaeker W (2003) Conceptual density functional theory. Chem Rev 103:1793–187412744694
40. Fuentealba P Parr RG Higher order derivatives in density-functional theory, especially the hardness derivative ∂η/∂N J Chem Phys 1991 94 5559 5564 10.1063/1.460491
Fuentealba P, Parr RG (1991) Higher order derivatives in density-functional theory, especially the hardness derivative ∂η/∂N. J Chem Phys 94:5559–5564
41. Ordon P Tachibana A Use of nuclear stiffness in search for a maximum hardness principle and for the softest states along the chemical reaction path: a new formula for the energy third derivative gamma J Chem Phys 2007 126 234115 10.1063/1.2741535 17600412
Ordon P, Tachibana A (2007) Use of nuclear stiffness in search for a maximum hardness principle and for the softest states along the chemical reaction path: a new formula for the energy third derivative gamma. J Chem Phys 126:23411517600412
42. Cardenas C Echegaray E Chakraborty D Anderson JSM Ayers PW Relationships between the third-order reactivity indicators in chemical density-functional theory J Chem Phys 2009 130 244105 10.1063/1.3151599 19566140
Cardenas C, Echegaray E, Chakraborty D, Anderson JSM, Ayers PW (2009) Relationships between the third-order reactivity indicators in chemical density-functional theory. J Chem Phys 130:24410519566140
43. Heidar-Zadeh F Richer M Fias S Miranda-Quintana RA Chan M Franco-Pérez M González-Espinoza CE Kim TD Lanssens C Patel AHG Yang XD Vöhringer-Martinez E Cárdenas C Verstraelen T Ayers PW An explicit approach to conceptual density functional theory descriptors of arbitrary order Chem Phys Lett 2016 600 307 312 10.1016/j.cplett.2016.07.039
Heidar-Zadeh F, Richer M, Fias S, Miranda-Quintana RA, Chan M, Franco-Pérez M, González-Espinoza CE, Kim TD, Lanssens C, Patel AHG, Yang XD, Vöhringer-Martinez E, Cárdenas C, Verstraelen T, Ayers PW (2016) An explicit approach to conceptual density functional theory descriptors of arbitrary order. Chem Phys Lett 600:307–312
44. Ordon P Komorowski L Nuclear reactivity and nuclear stiffness in density functional theory Chem Phys Lett 1998 292 1–2 22 27 10.1016/S0009-2614(98)00645-9
Ordon P, Komorowski L (1998) Nuclear reactivity and nuclear stiffness in density functional theory. Chem Phys Lett 292(1–2):22–27
45. Luty T Ordon P Eckhardt CJ A model for mechanochemical transformations: applications to molecular hardness, instabilities, and shock initiation of reaction J Chem Phys 2002 117 4 1775 1785 10.1063/1.1485968
Luty T, Ordon P, Eckhardt CJ (2002) A model for mechanochemical transformations: applications to molecular hardness, instabilities, and shock initiation of reaction. J Chem Phys 117(4):1775–1785
46. Ordon P Komorowski L Fluctuations in electronegativity and global hardness induced by molecular vibrations J Mol Structure (Theochem) 2003 630 25 32 10.1016/S0166-1280(03)00166-0
Ordon P, Komorowski L (2003) Fluctuations in electronegativity and global hardness induced by molecular vibrations. J Mol Structure (Theochem) 630:25–32
47. Ordon P Komorowski L DFT energy derivatives and their renormalization in molecular vibrations Int J Quantum Chem 2005 101 6 703 713 10.1002/qua.20327
Ordon P, Komorowski L (2005) DFT energy derivatives and their renormalization in molecular vibrations. Int J Quantum Chem 101(6):703–713
48. Morell C Grand A Toro-Labbé A New dual descriptor for chemical reactivity J Phys Chem A 2005 109 205 212 10.1021/jp046577a 16839107
Morell C, Grand A, Toro-Labbé A (2005) New dual descriptor for chemical reactivity. J Phys Chem A 109:205–21216839107
49. Morell C Grand A Toro-Labbé A Theoretical support for using the descriptor Chem Phys Lett 2006 425 342 346 10.1016/j.cplett.2006.05.003
Morell C, Grand A, Toro-Labbé A (2006) Theoretical support for using the descriptor. Chem Phys Lett 425:342–346
50. De Proft F Ayers PW Fias S Geerlings P Woodward-Hoffmann rules in density functional theory: Initial hardness response J Chem Phys 2006 125 214101 10.1063/1.2387953 17166009
De Proft F, Ayers PW, Fias S, Geerlings P (2006) Woodward-Hoffmann rules in density functional theory: Initial hardness response. J Chem Phys 125:21410117166009
51. Ayers PW Morell C De Proft F Geerlings P Understanding the Woodward-Hoffmann rules by using changes in electron density Chem Euro J 2007 13 29 10.1002/chem.200700365
Ayers PW, Morell C, De Proft F, Geerlings P (2007) Understanding the Woodward-Hoffmann rules by using changes in electron density. Chem Euro J 13:29
52. Geerlings P Ayers PW Toro-Labbe A Chattaraj PK De Proft F The Woodward-Hoffmann rules reinterpreted by conceptual density functional theory Acc Chem Res 2012 45 5 10.1021/ar200192t
Geerlings P, Ayers PW, Toro-Labbe A, Chattaraj PK, De Proft F (2012) The Woodward-Hoffmann rules reinterpreted by conceptual density functional theory. Acc Chem Res 45:5
53. Cardenas C Rabi N Ayers PW Morell C Jaramillo P Fuentealba P Chemical reactivity descriptors for ambiphilic reagents: dual descriptor, local hypersoftness, and electrostatic potential J Phys Chem A 2009 113 8660 8667 10.1021/jp902792n 19580251
Cardenas C, Rabi N, Ayers PW, Morell C, Jaramillo P, Fuentealba P (2009) Chemical reactivity descriptors for ambiphilic reagents: dual descriptor, local hypersoftness, and electrostatic potential. J Phys Chem A 113:8660–866719580251
54. Martínez-Araya JI Why are the local hyper-softness and the local softness more appropriate local reactivity descriptors than the dual descriptor and the Fukui function, respectively? J Math Chem 2023 62 2 1 15
Martínez-Araya JI (2023) Why are the local hyper-softness and the local softness more appropriate local reactivity descriptors than the dual descriptor and the Fukui function, respectively? J Math Chem 62(2):1–15
55. Lesiuk M Balawender R Zachara J Higher order alchemical derivatives from coupled perturbed self-consistent field theory J Chem Phys 2012 136 034104 10.1063/1.3674163 22280741
Lesiuk M, Balawender R, Zachara J (2012) Higher order alchemical derivatives from coupled perturbed self-consistent field theory. J Chem Phys 136:03410422280741
56. Balawender R Lesiuk M De Proft F van Alsenoy Ch Geerlings P Exploring chemical space with alchemical derivatives: alchemical transformations of H through Ar and their ions as a proof of concept Phys Chem Chem Phys 2019 21 23865 10.1039/C9CP03935J 31647077
Balawender R, Lesiuk M, De Proft F, van Alsenoy Ch, Geerlings P (2019) Exploring chemical space with alchemical derivatives: alchemical transformations of H through Ar and their ions as a proof of concept. Phys Chem Chem Phys 21:2386531647077
57. Balawender R, Geerlings P (2023) Exploring chemical space with alchemical derivatives, [Chapter 2 in:] Chemical reactivity vol. 2: approaches and applications; [Kaya S, Von Szentpaly L, Serdaroglu G, Guo L, Editors]; Elsevier
58. GAUSSIAN 16, Frisch M, Trucks G, Schlegel H, Scuseria G, Robb M, Cheeseman J, Scalmani G, Barone V, Mennucci B, Petersson G (2009) Inc., Wallingford, CT
59. Szarek P Komorowski L Lipiński J Fukui functions for atoms and ions: polarizability justified approach Int J Quant Chem 2010 110 2315 2319 10.1002/qua.22606
Szarek P, Komorowski L, Lipiński J (2010) Fukui functions for atoms and ions: polarizability justified approach. Int J Quant Chem 110:2315–2319
60. Komorowski L Lipiński J Szarek P Ordon P Polarization justified Fukui functions: the theory and applications for molecules J Chem Phys 2011 135 014109 10.1063/1.3603449 21744890
Komorowski L, Lipiński J, Szarek P, Ordon P (2011) Polarization justified Fukui functions: the theory and applications for molecules. J Chem Phys 135:01410921744890
61. Szarek P Komorowski L Modeling the electron density kernels J Comput Chem 2011 32 1721 1724 10.1002/jcc.21754 21328410
Szarek P, Komorowski L (2011) Modeling the electron density kernels. J Comput Chem 32:1721–172421328410
62. Sadlej AJ Basis Set: pTVZ Theor Chim Acta 1992 81 339 10.1007/BF01118573
Sadlej AJ (1992) Basis Set: pTVZ. Theor Chim Acta 81:339
63. https://physics.nist.gov/PhysRefData/ASD/levels_form.html
64. Kohn W (1985) Highlights of condensed matter theory, ed. Bassani F, Fumi F, Tosi MP. North Holland Physics Pub., Amsterdam
65. Dreizler RM Gross EKV Density functional theory 1990 Berlin-Heidelberg Springer Verlag
Dreizler RM, Gross EKV (1990) Density functional theory. Springer Verlag, Berlin-Heidelberg
66. March NH Relation between electrostatic potential of electron cloud at nucleus and chemical potential in atomic ions Phys Lett 1981 82A 73 74 10.1016/0375-9601(81)90941-5
March NH (1981) Relation between electrostatic potential of electron cloud at nucleus and chemical potential in atomic ions. Phys Lett 82A:73–74
67. March NH (1993) The ground state energy of atomic ions and its variation with the number of electrons [in:] Structure and Bonding, vol. 80, KD Sen ed., 71–86, Springer-Verlag
68. Komorowski L Empirical evaluation of chemical hardness Chem Phys Lett 1987 134 536 540 10.1016/0009-2614(87)87188-9
Komorowski L (1987) Empirical evaluation of chemical hardness. Chem Phys Lett 134:536–540
69. Komorowski L Electronegativity and hardness in the chemical approximation Chem Phys 1987 114 55 71 10.1016/0301-0104(87)80019-8
Komorowski L (1987) Electronegativity and hardness in the chemical approximation. Chem Phys 114:55–71
70. Komorowski L Chemical hardness and L. Pauling’s scale of electronegativity Z Naturforsch A 1987 42 767 773 10.1515/zna-1987-0718
Komorowski L (1987) Chemical hardness and L. Pauling’s scale of electronegativity. Z Naturforsch A 42:767–773
71. Komorowski L Electronegativity through the energy function Chem Phys Lett 1983 103 201 204 10.1016/0009-2614(83)80381-9
Komorowski L (1983) Electronegativity through the energy function. Chem Phys Lett 103:201–204
72. Komorowski L (1993) Hardness indices for free and bonded atoms, [in:] Structure and bonding, vol. 80, KD Sen ed., 46–69, Springer-Verlag
73. Yang W Parr RG Hardness, softness, and the Fukui function in the electronic theory of metals and catalysis Proc Nat Acad Sci USA 1985 82 6723 6726 10.1073/pnas.82.20.6723 3863123
Yang W, Parr RG (1985) Hardness, softness, and the Fukui function in the electronic theory of metals and catalysis. Proc Nat Acad Sci USA 82:6723–67263863123
74. Ayers PW The physical basis of the hard/soft acid/base principle Faraday Discuss 2007 135 161 190 10.1039/B606877D 17328428
Ayers PW (2007) The physical basis of the hard/soft acid/base principle. Faraday Discuss 135:161–19017328428
75. Cardenas C Ayers PW De Proft F Tozer DJ Geerlings P Should negative electron affinities be used for evaluating the chemical hardness Phys Chem Chem Phys 2011 13 2285 2293 10.1039/C0CP01785J 21113528
Cardenas C, Ayers PW, De Proft F, Tozer DJ, Geerlings P (2011) Should negative electron affinities be used for evaluating the chemical hardness. Phys Chem Chem Phys 13:2285–229321113528
76. Cardenas C Heidar-Zadeh F Ayers PW Benchmark values of chemical potential and chemical hardness for atoms and atomic ions (including unstable anions) from the energies of isoelectronic series Phys Chem Chem Phys 2016 18 25721 25734 10.1039/C6CP04533B 27711480
Cardenas C, Heidar-Zadeh F, Ayers PW (2016) Benchmark values of chemical potential and chemical hardness for atoms and atomic ions (including unstable anions) from the energies of isoelectronic series. Phys Chem Chem Phys 18:25721–2573427711480
77. Cardenas C Ayers PW Cedillo A Reactivity indicators for degenerate states in the density-functional theoretic chemical reactivity theory J Chem Phys 2011 134 174103 10.1063/1.3585610 21548669
Cardenas C, Ayers PW, Cedillo A (2011) Reactivity indicators for degenerate states in the density-functional theoretic chemical reactivity theory. J Chem Phys 134:17410321548669
78. Bultinck P Cardenas C Fuentealba P Johnson PA Ayers PW How to compute the Fukui matrix and function for systems with (quasi-)degenerate states J Chem Theor Comp 2014 10 202 210 10.1021/ct400874d
Bultinck P, Cardenas C, Fuentealba P, Johnson PA, Ayers PW (2014) How to compute the Fukui matrix and function for systems with (quasi-)degenerate states. J Chem Theor Comp 10:202–210
79. Miranda-Quintana RA Ayers PW Heidar-Zadeh F Reactivity and charge transfer beyond the parabolic model: the “|Δμ| big is good” principle J Phys Chem Lett 2018 9 4344 4348 10.1021/acs.jpclett.8b01312 29944379
Miranda-Quintana RA, Ayers PW, Heidar-Zadeh F (2018) Reactivity and charge transfer beyond the parabolic model: the “|Δμ| big is good” principle. J Phys Chem Lett 9:4344–434829944379
80. Heidar-Zadeh F Miranda-Quintana RA Verstraelen T Bultinck P Ayers PW When is the Fukui function not normalized? The danger of inconsistent energy interpolation models in density functional theory J Chem Theor Comp 2016 12 5777 5787 10.1021/acs.jctc.6b00494
Heidar-Zadeh F, Miranda-Quintana RA, Verstraelen T, Bultinck P, Ayers PW (2016) When is the Fukui function not normalized? The danger of inconsistent energy interpolation models in density functional theory. J Chem Theor Comp 12:5777–5787
81. Tsunada T Reconsideration of chemical indices in conceptual density functional theory Theor Chem Acc 2023 142 26 10.1007/s00214-023-02962-5
Tsunada T (2023) Reconsideration of chemical indices in conceptual density functional theory. Theor Chem Acc 142:26
82. Levebvre C Khartabil H Henon E New insight into atomic-level interpretation of interactions in molecules and reacting systems Phys Chem Chem Phys 2023 25 16 11398 11409 10.1039/D2CP02839E 37021406
Levebvre C, Khartabil H, Henon E (2023) New insight into atomic-level interpretation of interactions in molecules and reacting systems. Phys Chem Chem Phys 25(16):11398–1140937021406
83. Liu H-P Hunt KLC Molecular softness, hypersoftness, infrared absorption, and variational Raman scattering: new relations derived from nonlocal polarizability J Chem Phys 1995 103 24 10597 10604 10.1063/1.469844
Liu H-P, Hunt KLC (1995) Molecular softness, hypersoftness, infrared absorption, and variational Raman scattering: new relations derived from nonlocal polarizability. J Chem Phys 103(24):10597–10604
84. Capitani JF Nalewajski RF Parr RG Non-Born-Oppenheimer density functional theory of molecular systems J Chem Phys 1982 76 568 573 10.1063/1.442703
Capitani JF, Nalewajski RF, Parr RG (1982) Non-Born-Oppenheimer density functional theory of molecular systems. J Chem Phys 76:568–573
85. Munoz M Cardenas C How predictive could alchemical derivatives be? Phys Chem Chem Phys 2017 19 24 16003 16012 10.1039/C7CP02755A 28594013
Munoz M, Cardenas C (2017) How predictive could alchemical derivatives be? Phys Chem Chem Phys 19(24):16003–1601228594013
86. Cardenas C Tiznado W Ayers PW Fuentealba P The Fukui potential and the capacity of charge and the global hardness of atoms J Phys Chem A 2011 115 2325 2331 10.1021/jp109955q 21351729
Cardenas C, Tiznado W, Ayers PW, Fuentealba P (2011) The Fukui potential and the capacity of charge and the global hardness of atoms. J Phys Chem A 115:2325–233121351729
87. Geerlings P De Proft F External fields in conceptual density functional theory J Comput Chem 2022 44 442 10.1002/jcc.26978 36054623
Geerlings P, De Proft F (2022) External fields in conceptual density functional theory. J Comput Chem 44:44236054623
88. Lipiński J Sum rules for nonlinear optical properties of molecules Chem Phys Lett 2004 394 397 399 10.1016/j.cplett.2004.07.006
Lipiński J (2004) Sum rules for nonlinear optical properties of molecules. Chem Phys Lett 394:397–399
89. Lipiński J On the consequences of the violation of the Hellmann-Feynman theorem in calculations of electric properties of molecules Chem Phys Lett 2002 363 313 318 10.1016/S0009-2614(02)01186-7
Lipiński J (2002) On the consequences of the violation of the Hellmann-Feynman theorem in calculations of electric properties of molecules. Chem Phys Lett 363:313–318
90. Geerlings P Fias S Boisdenghien Z De Proft F Conceptual DFT: chemistry from the linear response function Chem Soc Rev 2014 43 4989 5008 10.1039/c3cs60456j 24531142
Geerlings P, Fias S, Boisdenghien Z, De Proft F (2014) Conceptual DFT: chemistry from the linear response function. Chem Soc Rev 43:4989–500824531142
91. Wang B Geerlings P Liu S De Proft F Extending the scope of conceptual density functional theory with second order analytical methodologies J Chem Theor Comp 2024 20 3 1169 1184 10.1021/acs.jctc.3c01184
Wang B, Geerlings P, Liu S, De Proft F (2024) Extending the scope of conceptual density functional theory with second order analytical methodologies. J Chem Theor Comp 20(3):1169–1184
92. Ordon. P. Effect of molecular deformations on the chemical DFT indices, Ph.D. Thesis, Wrocław 2003, www.ordon.edu.pl
