Next Article in Journal
Dopamine-Imprinted Polymers: Template-Monomer Interactions, Analysis of Template Removal and Application to Solid Phase Extraction
Previous Article in Journal
On Use of the Variable Zagreb vM2 Index in QSPR: Boiling Points of Benzenoid Hydrocarbons
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Valence Topological Charge-Transfer Indices for Dipole Moments: Percutaneous Enhancers

Institut Universitari de Ciència Molecular, Universitat de València, Dr. Moliner 50, E-46100 Burjassot, Spain
Molecules 2004, 9(12), 1222-1235; https://doi.org/10.3390/91201222
Submission received: 2 June 2004 / Accepted: 18 June 2004 / Published: 31 December 2004

Abstract

:
Valence topological charge-transfer (CT) indices are applied to the calculation of dipole moments. The algebraic and vector semisum CT indices are defined. The combination of CT indices allows the estimation of the dipole moments. The model is generalized for molecules with heteroatoms. The ability of the indices for the description of the molecular charge distribution is established by comparing them with the dipole moments of homologous series of percutaneous enhancers (phenyl alcohols and 4‑alkylanilines). Linear and quadratic correlation models are obtained. CT indices improve the multivariable quadratic regression equations for the dipole moment. The variance decreases 97% (4‑alkylanilines). No superposition of the corresponding Gk–Jk and GkV–JkV pairs is observed in the fits, which diminishes the risk of co‑linearity. The inclusion of the heteroatom in the π‑electron system is beneficial for the description of the dipole moment, owing to either the role of the additional p orbitals provided by the heteroatom or the role of steric factors in the π‑electron conjugation. Inclusion of a conjugated double bond in the alkyl chain lends to more rigid structures with dipole moment variations <1%.

Introduction

Passive drug penetration across absorbent membranes involves both the intrinsic resistance of the lipidic barriers themselves and that of the adjacent aqueous boundary layers, which can provide substantial additional strength against the passage of penetrants [1,2]. When adjacent to the lipoidal membrane and its luminal side, aqueous boundary layer thickness can be modified by agitation and, for in vitro experiments, the importance of suitable stirring conditions to obtain reproducible results was recognized [3]. In comparative studies, the existence of the aqueous boundary layers, although acknowledged, was sometimes virtually ignored because of its uniformity throughout a strictly designed and controlled steady-state experimentation work [4,5,6,7].
Díez-Sales et al. performed in vitro permeability studies using a series of weakly basic xenobiotics of long-term toxicological interest across a homogeneous lipoidal artificial membrane under several working conditions, including the presence and the partial or total absence of aqueous boundary layers [8]. Synthetic surfactants, at either the critical micelle or supramicellar concentration, were reported to disrupt the in vivo aqueous stagnant diffusion layers adjacent to the membrane at its luminal side, as well as to increase the intrinsic membrane polarity through a direct penetration effect [9,10,11,12], and it was thought to evaluate these properties on in vitro artificial membranes. To achieve this, Díez-Sales et al. established permeability–lipophilicity correlations, and compared them for each condition according to previously reported discriminative equations [13,14,15,16,17].
Percutaneous penetration depends essentially on the lipophilicity of the tested substances [18]. The use of homologous series of compounds made it possible to establish behaviour models, which allow predicting the percutaneous absorption capacity of chemically related substances [19,20]. Much of the theoretical background information on percutaneous absorption was developed from studies of non‑electrolytic permeating species [21,22], and only a little is available on the effect of pH and pKa on permeation of ionizable drugs through the skin. The vehicle’s pH may have a profound influence upon the percutaneous delivery from topical products [23]. Depending on the pKa of the compound and on the pH of the vehicle, an equilibrium mixture of ionized and unionized species would be present in the immediate vicinity of the skin [24]. According to the pH–partition simple hypothesis, only the non‑ionized forms of drugs are able to pass through lipoidal membranes in significant amounts. However, the permeation of ionized drugs through the skin is possible and therefore must be taken into account, although ionization reduces topical availability [25]. The mechanism by which percutaneous penetration enhancers operate is not fully understood, but the structured lipids within the intercellular channels play an important role in controlling absorption. Penetration enhancers may act by either interacting with the highly ordered lipid structure, or modifying the partitioning of the drug into the tissue [26,27].
In earlier publications, topological charge-transfer (CT) indices were applied to the calculation of the molecular dipole moment of hydrocarbons [28], valence-isoelectronic series of benzene, styrene [29,30] and cyclopentadiene [31], and phenyl alcohols [32]. An index inspired by biological plastic evolution [33,34] and fractal dimension analysis [35,36] improved the results. In the present study, the valence CT indices have been applied to the calculation of the dipole moment of phenyl alcohols, 4‑alkylanilines and aliphatic amines. The next section presents the CT indices. Following that, the indices are generalized for heteroatoms. Next, the calculation results are discussed. The last section summarizes the conclusions.

Materials and methods

The most important matrices that delineate the labelled chemical graph are the adjacency (A) [37] and the distance (D) matrices, wherein Dij = ⎕ij if i = j, “0” otherwise; ⎕ij is the shortest edge count between vertices i and j [38]. In A, Aij = 1 if vertices i and j are adjacent, “0” otherwise. The D[‑2] matrix is that whose elements are the squares of the reciprocal distances Dij‑2. The intermediate matrix M is defined as the matrix product of A by D[‑2]:
M = A D [ 2 ]
The CT matrix C is defined as C = MMT, where MT is the transpose of M [39]. By agreement Cii = Mii. For ij, the Cij terms represent a measure of the intramolecular net charge transferred from atom j to i. The topological CT indices Gk are described as the sum of absolute values of the Cij terms defined for the vertices i,j placed at a topological distance Dij equal to k:
G k = i = 1 N 1 j = i + 1 N | C i j | δ ( k , D i j )
where N is the number of vertices in the graph, Dij are the entries of the D matrix, and δ is the Kronecker δ function, being δ = 1 for i = j and δ = 0 for ij. Gk represents the sum of all the Cij terms, for every pair of vertices i and j at topological distance k. Other topological CT index, Jk, is defined as:
J k = G k N 1
This index represents the mean value of CT for each edge, since the number of edges for acyclic compounds is N – 1.
The algebraic semisum CT index μalg is defined as
μ a lg = 1 2 i = 1 N 1 j = i + 1 N A i j | C i j | = 1 2 e = 1 m | C e |
where Ce is Cij index for vertices i and j connected by edge e [28]. The sum extends for all pairs of adjacent vertices in the molecular graph and μalg is a graph invariant. An edge analysis suggests that each dipole moment μe, of edge e connecting vertices i and j can be evaluated from its edge Ce index as:
μ e = 1 2 C e
Each edge dipole can be associated with a vector μe in space. μe has magnitude |μe|, lies in the edge e connecting vertices i and j, and its direction is ji. The molecular dipole vector μ results as the vector sum of the edge dipoles as:
μ = e = 1 m μ e = 1 2 e = 1 m C e
summed for all the m edges in the molecular graph. The vector semisum CT index μvec is defined as the module of μ:
μ vec = N ( μ ) = ( μ x 2 + μ y 2 + μ z 2 ) 1 / 2
Therefore, μvec is a graph invariant.

Valence charge-transfer indices for heteroatoms

When heteroatoms are present, some way of discriminating atoms of different kinds needs to be considered [40]. In valence CT indices terms, the presence of each heteroatom is taken into account by introducing its electronegativity value in the corresponding entry of the main diagonal of the adjacency matrix A. For each heteroatom X, its entry Aii is redefined as:
A u V = 2.2 ( χ X χ C )
to give the valence adjacency AV matrix, where χX and χC are the electronegativities of heteroatom X and carbon, respectively, in Pauling units. The subtractive term keeps AiiV = 0 for the C atom, and the factor gives AiiV = 2.2 for O, which was taken as standard. From AV, MV and CV, GkV, JkV, μalgV and μvecV are calculated following the former procedure with AV. CiiV, GkV, JkV, μalgV and μvecV are graph invariants. μvec is sensitive only to the steric effect of heteroatoms, while μvecV is sensitive to both electronic and steric effects.

Calculation Rresults and Discussion

The molecular CT indices Gk, Jk, GkV and JkV (with k < 6) are reported in Table 1 for a series of 13 phenyl alcohols (eight form a homologous series and five are congeneric) and 14 amines (seven homologous 4‑alkylanilines and seven congeneric). In particular, for the homologous phenyl alcohols, G1 and G2 are sensitive to the presence of the alkyl chain, and G3, G4 and G5 indicate, respectively, the occurrence of at least 2, 3 and 4 C atoms in this chain. For the 4‑alkylanilines, G1 is a sign of the incidence of the alkyl chain, and G2G5 signify the existence of at least 2–5 C atoms in this chain. For both series, G1V denotes the presence of the alkyl chain, and G2V–G5V imply the occurrence of at least 2–5 C atoms in this chain.
Table 1. Values of the Gk and Jk charge-transfer indices up to fifth order for phenyl alcohols or amines
Table 1. Values of the Gk and Jk charge-transfer indices up to fifth order for phenyl alcohols or amines
MoleculeNG1 G2 G3 G4 G5
phenol72.00000.88890.37500.22220.0000
benzyl alcohol81.25006.77780.81250.41330.1250
2‑phenyl‑1‑ethanol91.25006.77780.87500.56440.2431
3‑phenyl‑1‑propanol101.25006.77780.87500.60440.3333
4‑phenyl‑1‑butanol111.25006.77780.87500.60440.3611
5‑phenyl‑1‑pentanol121.25006.77780.87500.60440.3611
6‑phenyl‑1‑hexanol131.25006.77780.87500.60440.3611
7‑phenyl‑1‑heptanol141.25006.77780.87500.60440.3611
1‑phenyl‑2‑propanol102.25006.77780.93750.71560.3611
2‑phenyl‑2‑propanol102.75007.22221.56250.79560.3750
3‑phenyl‑2‑propen‑1‑ol101.25008.22220.87500.48440.2708
1‑phenyl‑1‑pentanol121.75007.11111.31250.87560.4861
1‑phenyl‑2‑pentanol122.25007.00001.06250.75560.4792
aniline71.25006.66670.50000.22220.0000
4‑methylaniline82.50008.00000.87500.52440.0000
4‑ethylaniline92.50008.11111.18750.71560.1806
4‑propylaniline102.50008.11111.25000.86670.2986
4‑butylaniline112.50008.11111.25000.90670.3889
4‑pentylaniline122.50008.11111.25000.90670.4167
4‑hexylaniline132.50008.11111.25000.90670.4167
2‑methylaniline82.00008.22221.00000.44440.0000
4‑isopropylaniline103.00008.22221.50000.90670.3611
4‑(1’‑propenyl)aniline102.50009.44441.00000.74670.2639
1‑aminonaphthalene111.750012.22221.68750.98220.1250
2‑aminonaphthalene112.250012.00001.50000.99110.1736
4‑cyclohexylaniline132.50008.22221.56251.05780.5208
4‑aminobiphenyl132.250013.77781.25000.79110.4167
Molecule J1 J2 J3 J4 J5
phenol0.33330.14810.06250.03700.0000
benzyl alcohol0.17860.96830.11610.05900.0179
2‑phenyl‑1‑ethanol0.15630.84720.10940.07060.0304
3‑phenyl‑1‑propanol0.13890.75310.09720.06720.0370
4‑phenyl‑1‑butanol0.12500.67780.08750.06040.0361
5‑phenyl‑1‑pentanol0.11360.61620.07950.05490.0328
6‑phenyl‑1‑hexanol0.10420.56480.07290.05040.0301
7‑phenyl‑1‑heptanol0.09620.52140.06730.04650.0278
1‑phenyl‑2‑propanol0.25000.75310.10420.07950.0401
2‑phenyl‑2‑propanol0.30560.80250.17360.08840.0417
3‑phenyl‑2‑propen‑1‑ol0.13890.91360.09720.05380.0301
1‑phenyl‑1‑pentanol0.15910.64650.11930.07960.0442
1‑phenyl‑2‑pentanol0.20450.63640.09660.06870.0436
aniline0.20831.11110.08330.03700.0000
4‑methylaniline0.35711.14290.12500.07490.0000
4‑ethylaniline0.31251.01390.14840.08940.0226
4‑propylaniline0.27780.90120.13890.09630.0332
4‑butylaniline0.25000.81110.12500.09070.0389
4‑pentylaniline0.22730.73740.11360.08240.0379
4‑hexylaniline0.20830.67590.10420.07560.0347
2‑methylaniline0.28571.17460.14290.06350.0000
4‑isopropylaniline0.33330.91360.16670.10070.0401
4‑(1’‑propenyl)aniline0.27781.04940.11110.08300.0293
1‑aminonaphthalene0.17501.22220.16880.09820.0125
2‑aminonaphthalene0.22501.20000.15000.09910.0174
4‑cyclohexylaniline0.20830.68520.13020.08810.0434
4‑aminobiphenyl0.18751.14810.10420.06590.0347
MoleculeG1VG2VG3VG4VG5V
phenol2.20001.10000.36390.08470.0000
benzyl alcohol2.95006.66110.56250.25330.0370
2‑phenyl‑1‑ethanol2.95007.10560.74440.40440.1319
3‑phenyl‑1‑propanol2.95007.10560.99440.50190.2222
4‑phenyl‑1‑butanol2.95007.10560.99440.66190.2824
5‑phenyl‑1‑pentanol2.95007.10560.99440.66190.3936
6‑phenyl‑1‑hexanol2.95007.10560.99440.66190.3936
7‑phenyl‑1‑heptanol2.95007.10560.99440.66190.3936
1‑phenyl‑2‑propanol3.45007.65560.80690.55560.2500
2‑phenyl‑2‑propanol3.45008.20561.31250.63560.2870
3‑phenyl‑2‑propen‑1‑ol2.95008.32780.63060.38190.1597
1‑phenyl‑1‑pentanol2.95007.32221.18190.77310.4861
1‑phenyl‑2‑pentanol3.45007.65561.05140.73310.3681
aniline0.85006.22220.25560.15350.0000
4‑methylaniline2.10007.55560.63060.45570.0440
4‑ethylaniline2.10007.66670.94310.64680.1690
4‑propylaniline2.10007.66671.00560.79790.2871
4‑butylaniline2.10007.66671.00560.83790.3773
4‑pentylaniline2.10007.66671.00560.83790.4051
4‑hexylaniline2.10007.66671.00560.83790.4051
2‑methylaniline1.60007.77780.87780.37570.0000
4‑isopropylaniline2.60007.77781.25560.83790.3171
4‑(1’‑propenyl)aniline2.10009.00000.75560.67790.2199
1‑aminonaphthalene1.350011.67221.32080.77600.0810
2‑aminonaphthalene1.850011.55561.25560.85360.0856
4‑cyclohexylaniline2.10007.77781.31810.98900.4768
4‑aminobiphenyl1.850013.33331.00560.72240.3727
MoleculeJ1VJ2VJ3VJ4VJ5V
phenol0.36370.18330.06060.01410.0000
benzyl alcohol0.42140.95160.08040.03620.0053
2‑phenyl‑1‑ethanol0.36880.88820.09310.05060.0165
3‑phenyl‑1‑propanol0.32780.78950.11050.05580.0247
4‑phenyl‑1‑butanol0.29500.71060.09940.06620.0282
5‑phenyl‑1‑pentanol0.26820.64600.09040.06020.0358
6‑phenyl‑1‑hexanol0.24580.59210.08290.05520.0328
7‑phenyl‑1‑heptanol0.22690.54660.07650.05090.0303
1‑phenyl‑2‑propanol0.38330.85060.08970.06170.0278
2‑phenyl‑2‑propanol0.38330.91170.14580.07060.0319
3‑phenyl‑2‑propen‑1‑ol0.32780.92530.07010.04240.0177
1‑phenyl‑1‑pentanol0.26820.66570.10740.07030.0442
1‑phenyl‑2‑pentanol0.31360.69600.09560.06660.0335
aniline0.14171.03700.04260.02560.0000
4‑methylaniline0.30001.07940.09010.06510.0063
4‑ethylaniline0.26250.95830.11790.08090.0211
4‑propylaniline0.23330.85190.11170.08870.0319
4‑butylaniline0.21000.76670.10060.08380.0377
4‑pentylaniline0.19090.69700.09140.07620.0368
4‑hexylaniline0.17500.63890.08380.06980.0338
2‑methylaniline0.22861.11110.12540.05370.0000
4‑isopropylaniline0.28890.86420.13950.09310.0352
4‑(1’‑propenyl)aniline0.23331.00000.08400.07530.0244
1‑aminonaphthalene0.13501.16720.13210.07760.0081
2‑aminonaphthalene0.18501.15560.12560.08540.0086
4‑cyclohexylaniline0.17500.64810.10980.08240.0397
4‑aminobiphenyl0.15421.11110.08380.06020.0311
The molecular dipole moments μ (experimental, and calculated as vector semisums of Cij and CijV) are listed in Table 2. As experimental values were not available for any whole series, some reference values were computed with program MOPAC-AM1. The reliability of AM1 results was tested with all the entries in Table 2 for which experimental data were available. AM1 calculations adequately reproduced the oscillatory behaviour of the experimental data, mimicking four minima (phenol, 2‑phenyl‑1‑ethanol, aniline and 4‑ethylaniline) and four maxima (benzyl alcohol, 3‑phenyl‑1‑propanol, 4‑methylaniline and 4‑propylaniline). This test suggested that AM1 allows a good approximation, at least for the general performance of both homologous series, and that the error is sufficiently constant throughout the series. The number of compounds in both series was not increased because longer molecules are not percutaneous enhancers owing to their low transdermal penetration.
Table 2. Molecular dipole moment values, μ (D), for alcohols and amines calculated with charge-transfer indices.
Table 2. Molecular dipole moment values, μ (D), for alcohols and amines calculated with charge-transfer indices.
MoleculeNumber of carbon atoms in alkyl chainVector semisumValence vector semisumExperimentala
phenol00.7372.4311.400 (1.233b)
benzyl alcohol10.5892.4871.700 (1.568b)
2‑phenyl‑1‑ethanol20.7002.2571.590 (1.497b)
3‑phenyl‑1‑propanol30.5732.5191.640 (1.597b)
4‑phenyl‑1‑butanol40.7022.2491.345b
5‑phenyl‑1‑pentanol50.5732.5191.626b
6‑phenyl‑1‑hexanol60.7022.2501.346b
7‑phenyl‑1‑heptanol70.5732.5181.634b
1‑phenyl‑2‑propanol30.9232.3471.564b
2‑phenyl‑2‑propanol30.7802.8211.463b
3‑phenyl‑2‑propen‑1‑ol30.5612.4951.591b
1‑phenyl‑1‑pentanol50.4262.7941.746b
1‑phenyl‑2‑pentanol50.8952.3671.496b
aniline00.6550.7921.560 (1.584b)
4‑methylaniline10.0791.5261.640 (1.423b)
4‑ethylaniline20.3851.3391.492b
4‑propylaniline30.0871.5331.502b
4‑butylaniline40.3981.3331.511b
4‑pentylaniline50.0931.5331.489b
4‑hexylaniline60.3951.3321.496b
2‑methylaniline11.1870.7791.590 (1.560b)
4‑isopropylaniline30.6251.4901.450b
4‑(1’‑propenyl)aniline30.1081.5121.512b
1‑aminonaphthalene40.6450.8021.490 (1.690b)
2‑aminonaphthalene40.6550.7921.820 (1.868b)
4‑cyclohexylaniline60.4801.1261.760 (1.447b)
4‑aminobiphenyl60.6510.7951.447b
a Taken from Reference 44.b Calculations carried out with program MOPAC-AM1.
In particular, 3‑phenyl‑2‑propen‑1‑ol was chosen in order to compare the influence of a double bond in the alkyl chain region of a phenyl alcohol. The presence of the enol group (conjugated double bond in β-position with respect to the –OH group) lent 3‑phenyl‑2‑propen‑1‑ol to greater rigidity than for 3‑phenyl‑1‑propanol. However, μAM1 change was <1%. For the three phenyl propanols (primary, secondary and tertiary alcohols), the primary alcohol 3‑phenyl‑1‑propanol showed the greatest μAM1. For the three phenyl pentanols (primary, and α‑ and β‑phenyl –OH secondary alcohols), the primary alcohol 5‑phenyl‑1‑pentanol presented a relatively large μAM1. On going from 4‑propylaniline to 4‑(1’‑propenyl)aniline, the effect of the conjugated double bond on μAM1 was again <1%. For 2‑ and 4‑methylaniline, the para‑(4)‑methylaniline showed the greatest μexperiment but the smallest μAM1. For 4‑propylaniline and 4‑isopropylaniline, the more linear 4‑propylaniline revealed the greatest μAM1. For 4‑butylaniline and both aminonaphthalenes, the rigid 2‑aminonaphthalene displayed the greatest μAM1. For 4‑hexylaniline, 4‑cyclohexylaniline and 4‑aminobiphenyl, the more linear 4‑hexylaniline exhibited the greatest μAM1. Figure 1 illustrates how the dipole moment of both homologous series fluctuates with the number of C atoms in the alkyl chain, n. For either series, the experimental μ and calculated μvecV vary together in an alternate fashion: for odd n, μ is greater than for even n. However, the computed μvec presents the opposite tendency. The corresponding interpretation is that, as μvec is not sensitive to the electronic effect of either heteroatom, the steric (μvec) and electronic (μvecV) factors are antagonistic, and the electronic effect dominates over the steric one. The μexperiment decreases ca. 4% for either n = 0 → 6 or n = 1 → 7.
Figure 1. Dipole moment of phenyl alcohols (empty symbols) and 4‑alkylanilines (filled symbols) as a function of the number of C atoms in alkyl chain.
Figure 1. Dipole moment of phenyl alcohols (empty symbols) and 4‑alkylanilines (filled symbols) as a function of the number of C atoms in alkyl chain.
Molecules 09 01222 g001
The experimental dipole moments and CT indices were correlated for the homologous series. As μexperiment for the 4‑alkylanilines is almost constant (≈1.5D), especially for n > 1, it was represented by only one point in μ space. Therefore, the joint fit of phenyl alcohols and this point was performed. The best linear fit turned out to be:
μ = 1.45 + 0.0906 G 2 9.39 J 4 V
N = 9 r = 0.688  SD = 0.115 F = 2.7  MAPE = 4.54 %  AEV = 0.5269
where MAPE is the mean absolute percentage error and AEV is the approximation error variance. All other models with greater MAPE and AEV were discarded, including those models counting the number of C atoms in the chain. The variables G2 and J4V suggest the importance of electronic density distribution in the different compounds. The best quadratic model for μ turned out to be:
μ = 17.0 12.8 G 1 V + 2.60 ( G 1 V ) 2 97.8 ( J 4 V ) 2
N = 9  MAPE = 4.12 %  AEV = 0.5072
and AEV decreased by 4%. The inclusion of μvec or μvecV did not improve the fit.
As μ4‑alkylanilines is almost constant, a series of ammonia and 10 homologous aliphatic primary amines was chosen in order to compare the dipole moments [41]. Figure 2 shows how the dipole moment of the aliphatic amine series varies with the number of C atoms, n. In general, μexperiment and μvecV change together in an alternate fashion (μodd n > μeven n) while μvec,odd n < μvec,even n. Again, the steric and electronic factors are antagonistic, and the electronic effect dominates. Although μexperiment decreases by 9% for n = 0 → 10, it increases by 21% for n = 1 → 9.
Figure 2. Dipole moment of ammonia and aliphatic amines as a function of the number of C atoms, n. Point n = 9 is AM1 calculation.
Figure 2. Dipole moment of ammonia and aliphatic amines as a function of the number of C atoms, n. Point n = 9 is AM1 calculation.
Molecules 09 01222 g002
As μ4‑alkylanilines hardly varies, it represents only one point in the space of μ and the joint fit of the aliphatic amines and this point was performed. The best linear model for μ turned out to be:
μ = 1.45 0.285 G 1 0.162 J 1 V + 32.6 J 5 V
N = 12 r = 0.879  SD = 0.078 F = 9.1  MAPE = 3.60 %  AEV = 0.2291
and AEV decreased by 57%. The best quadratic model and the inclusion of μvec or μvecV did not improve the correlation.
A comparative study was performed for each homologous series. For ammonia and the aliphatic amines, the best linear model for μ turned out to be:
μ = 1.46 0.305 J 1 + 2.17 J 3 0.162 G 1 V + 30.8 J 5 V
N = 11 r = 0.883  SD = 0.086 F = 5.3  MAPE = 3.59 %  AEV = 0.2226
and AEV decreased by 58% from that in Equation (6).
No superposition of the corresponding Gk–Jk or GkV–JkV pairs is observed in Equations (6–9), which diminishes the risk of co‑linearity in the fits, given the close relationship between each pair Gk–Jk in Equation (2) [42,43].
If the first point in Figure 2 (Equation 9) is omitted the best linear fit turns out to be:
μ = 1.29 + 1.27 G 3 2290 J 4 0.602 G 2 V + 0.584 J 2 V + 4190 J 5 V
N = 10 r = 0.950 r = 0.950  SD = 0.069 F = 7.3  MAPE = 1.92 %  AEV = 0.1004
and AEV decreased 81%. Equation (10) is overfitted but it is given with the only purpose of showing that ammonia (n = 0), which is not an aliphatic amine, acts as a marginal outlier for Equation (9). The best quadratic model and the inclusion of μvec or μvecV did not improve the correlation.
For the homologous phenyl alcohols, the best linear model turned out to be:
μ = 0.799 + 0.364 J 2 V + 0.871 μ vec V
N = 8 r = 0.930  SD = 0.064 F = 16.0  MAPE = 2.59 %  AEV = 0.1351
and AEV decreased by 74%. Equation (11) is overfitted but it is given with the purpose of justifying that μ4‑alkylanilines were represented by only one point in μ space in Equations (6–7). The best quadratic model did not improve the fit.
For the homologous 4‑alkylanilines, the best linear model turned out to be:
μ = 8.87 1.28 G 2 + 1.43 G 1 V
N = 7 r = 0.992  SD = 0.009 F = 118.1  MAPE = 0.32 %  AEV = 0.0167
and AEV decreased by 97%. Equation (12) is overfitted but it is given with the purpose of showing that μphenyl alcohols are difficult to fit due to their oscillating variation with the number of C atoms. The best quadratic model and the inclusion of μvec or μvecV did not improve the correlation.

Conclusions

From the preceding results the following conclusions can be drawn:
  • Inclusion of the heteroatom in the π‑electron system was beneficial for the description of the dipole moment, owing to either the role of the additional p orbitals provided by the heteroatom or the role of steric factors in the π‑electron conjugation. The analysis of the electronic and steric factors in μ caused by the heteroatom showed that both factors are antagonistic, and that the electronic factor dominates over the steric one.
  • The conjugated double bond in the alkyl chain of 3‑phenyl‑2‑propen‑1‑ol and 4‑(1’‑propenyl)-aniline lent to more rigid structures with dipole moment variations <1%.
  • Linear and quadratic correlation models were obtained for the molecular dipole moments of phenyl alcohols, 4‑alkylanilines and aliphatic amines. μvecV improved the multivariable regression equations for μ, diminishing the risk of co‑linearity in the fit. Improvements in correlation suggest the general applicability of this index.

Acknowledgements

The author acknowledges financial support from the Spanish MCT (Plan Nacional I+D+I, Project No. BQU2001‑2935‑C02‑01) and Generalitat Valenciana (DGEUI INF01‑051 and INFRA03‑047, and OCYT GRUPOS03‑173).

References

  1. Stehle, R.G.; Higuchi, W.I. In Vitro Model for Transport of Solutes in Three-Phase System II: Experimental Considerations. J. Pharm. Sci. 1972, 61, 1931–1935. [Google Scholar] [CrossRef]
  2. Lovering, E.G.; Black, D.B. Diffusion Layers Effects on Permeation of Phenylbutazone through Polydimethylsiloxane. J. Pharm. Sci. 1974, 63, 1399–1402. [Google Scholar] [CrossRef]
  3. Barry, B.W. Novel Mechanisms and Devices to Enable Successful Transdermal Drug Delivery. Eur. J. Pharm. Sci. 2001, 14, 101–114. [Google Scholar] [CrossRef]
  4. Sato, S.; Kim, S.W. Macromolecular Diffusion through Polymer Membranes. Int. J. Pharm. 1984, 22, 229–255. [Google Scholar] [CrossRef]
  5. Ackermann, C.; Flynn, G.L.; van Wyk, C.J. Percutaneous Absoption of Urea. Int. J. Cosm. Sci. 1985, 7, 251–264. [Google Scholar] [CrossRef]
  6. Ackermann, C.; Flynn, G.L.; Smith, W.M. Ether-Water Partitioning and Permeability through Nude Mouse Skin in Vitro. II. Hydrocortisone‑n‑alkyl Esters, Alkanols and Hydrophilic Compounds. Int. J. Pharm. 1987, 36, 67–71. [Google Scholar] [CrossRef]
  7. Ackermann, C.; Flynn, G.L. Ether-Water Partitioning and Permeability through Nude Mouse Skin in Vitro. I. Urea, Thiourea, Glycerol and Glucose. Int. J. Pharm. 1987, 36, 61–66. [Google Scholar] [CrossRef]
  8. Díez-Sales, O.; Copoví, A.; Casabó, V.G.; Herráez, M. A Modelistic Approach Showing the Importance of the Stagnant Aqueous Layers in in Vitro Diffusion Studies, and in Vitro-in Vivo Correlations. Int. J. Pharm. 1991, 77, 1–11. [Google Scholar]
  9. Plá-Delfina, J.M.; Pérez-Buendía, M.D.; Casabó, V.G.; Peris-Ribera, J.E.; Sánchez-Moyano, E.; Martín-Villodre, A. Absorption Partition Relationships for True Homologous Series of Xenobiotics as a Possible Approach to Study Mechanisms of Surfactants in Absorption. I. Aromatic Amines in Rat Colon. Int. J. Pharm. 1987, 37, 49–64. [Google Scholar]
  10. Collado, E.F.; Fabra-Campos, S.; Peris-Ribera, J.E.; Casabó, V.G.; Martín-Villodre, A.; Plá-Delfina, J.M. Absorption Partition Relationships for True Homologous Series of Xenobiotics as a Possible Approach to Study Mechanisms of Surfactants in Absorption. II. Aromatic Amines in Rat Small Intestine. Int. J. Pharm. 1988, 44, 187–196. [Google Scholar]
  11. Garrigues, T.M.; Collado, E.F.; Fabra-Campos, S.; Pérez-Buendía, M.D.; Martín-Villodre, A.; Plá-Delfina, J.M. Absorption Partition Relationships for True Homologous Series of Xenobiotics as a Possible Approach to Study Mechanisms of Surfactants in Absorption. III. Aromatic Amines and Cationic Surfactants. Int. J. Pharm. 1989, 57, 189–196. [Google Scholar]
  12. Garrigues, T.M.; Pérez-Varona, A.T.; Climent, E.; Bermejo, M.V.; Martín-Villodre, A.; Plá-Delfina, J.M. Gastric Absorption of Acidic Xenobiotics in the Rat: Biophysical Interpretation of an Apparently Atypical Behaviour. Int. J. Pharm. 1990, 64, 127–138. [Google Scholar]
  13. Díez-Sales, O.; Guzmán, D.; Cano, D.; Martín, A.; Sánchez, E.; Herráez, M. A Comparative in Vitro Study of Permeability with Different Synthetic and Biological Membranes. Eur. J. Drug Metab. Pharmacokinet. 1991, Spec. No. 3. 441–446. [Google Scholar]
  14. Sánchez-Moyano, E.; Seco, C.; Santolaria, A.; Fabra-Campos, S.; Herráez, M.; Matín-Villodre, A. Partition Behavior of Anilines in Bulk-Phase and High-Performance Liquid Chromatographic Systems: Influence on Correlation with Biological Constants. J. Pharm. Sci. 1992, 81, 720–725. [Google Scholar]
  15. Díez-Sales, O.; López-Castellano, A.; Maiques-Lacer, F.J.; Herráez-Domínguez, M. An in Vitro Percutaneous Absorption Study of Non-ionic Compounds across Human Skin. Pharmazie 1993, 48, 684–686. [Google Scholar]
  16. Díez-Sales, O.; Pérez-Sayas, E.; Martín-Villodre, A.; Herráez-Domínguez, M. The Prediction of Percutaneous Absorption: I. Influence of the Dermis on in Vitro Permeation Models. Int. J. Pharm. 1993, 100, 1–7. [Google Scholar]
  17. Díez-Sales, O.; Watkinson, A.C.; Herráez-Domínguez, M.; Javaloyes, C.; Hadgraft, J. A Mechanistic Investigation of the in-Vitro Human Skin Permeation Enhancing Effect of Atone®. Int. J. Pharm. 1996, 129, 33–40. [Google Scholar]
  18. López, A.; Morant, M.J.; Guzmán, D.; Borrás-Blasco, J.; Díez-Sales, O.; Herráez, M. Skin Permeation Model of Phenylalkylcarboxylic Homologous Acids and their Enhancer Effect on Percutaneous Penetration of 5-Fluorouracil. Int. J. Pharm. 1996, 139, 205–213. [Google Scholar]
  19. López, A.; Pellett, M.A.; Llinares, F.; Díez-Sales, O.; Herráez, M.; Hadgraft, J. The Enhancer Effect of Several Phenyl Alcohols on Percutaneous Penetration of 5-Fluorouracil. Pharm. Res. 1997, 14, 681–685. [Google Scholar]
  20. López, A.; Faus, V.; Díez-Sales, O.; Herráez, M. Skin Permeation Model of Phenyl Alcohols: Comparison of Experimental Conditions. Int. J. Pharm. 1998, 173, 183–191. [Google Scholar]
  21. Yalkowsky, S.H.; Flynn, G.L. Transport of Alkyl Homologs across Synthetic and Biological Membranes: A New Model for Chain Length–Activity Relationships. J. Pharm. Sci. 1973, 62, 210–217. [Google Scholar] [CrossRef]
  22. Flynn, G.L.; Yalkowsky, S.H.; Roseman, T.J. Mass Transport Phenomena and Models: Theoretical Concepts. J. Pharm. Sci. 1974, 63, 479–510. [Google Scholar] [CrossRef]
  23. Irwin, W.J.; Sanderson, F. D.; Po, A.L.W. Percutaneous Absorption of Ibuprofen: Vehicle Effects on Transport through Rat Skin. Int. J. Pharm. 1990, 66, 193–200. [Google Scholar] [CrossRef]
  24. Swarbrick, J.; Lee, G.; Brom, J.; Gensmantel, N.P. Drug Permeation through Human Skin II: Permeability of Ionizable Compounds. J. Pharm. Sci. 1984, 73, 1352–1355. [Google Scholar] [CrossRef]
  25. Banerjee, P.S.; Ritschel, W.A. Transdermal Permeation of Vasopressin. I. Influence of pH, Concentration, Shaving and Surfactant on in Vitro Permeation. Int. J. Pharm. 1989, 49, 189–197. [Google Scholar] [CrossRef]
  26. Lewis, D.; Hadgraft, J. Mixed Monolayers of Dipalmitoylphosphatidylcholine with Azone or Oleic Acid at the Air-Water Interface. Int. J. Pharm. 1990, 65, 211–218. [Google Scholar] [CrossRef]
  27. Williams, A.C.; Barry, B.W. Terpenes and the Lipid–Protein Partitioning Theory of Skin Penetration Enhancement. Pharm. Res. 1991, 8, 17–24. [Google Scholar]
  28. Torrens, F. A New Topological Index to Elucidate Apolar Hydrocarbons. J. Comput.-Aided Mol. Des. 2001, 15, 709–719. [Google Scholar] [CrossRef]
  29. Torrens, F. Valence Topological Charge-Transfer Indices for Dipole Moments. J. Mol. Struct. (Theochem) 2003, 621, 37–42. [Google Scholar] [CrossRef]
  30. Torrens, F. Valence Topological Charge-Transfer Indices for Dipole Moments. Mol. Diversity. in press.
  31. Torrens, F. Valence Topological Charge-Transfer Indices for Reflecting Polarity: Correction for Heteromolecules. Molecules. submitted for publication.
  32. Torrens, F. Valence Topological Charge-Transfer Indices for Dipole Moments. Molecules 2003, 8, 169–185. [Google Scholar] [CrossRef]
  33. Torrens, F. A New Chemical Index Inspired by Biological Plastic Evolution. Indian J. Chem., Sect. A 2003, 42, 1258–1263. [Google Scholar]
  34. Torrens, F. A Chemical Index Inspired by Biological Plastic Evolution: Valence-Isoelectronic Series of Aromatics. J. Chem. Inf. Comput. Sci. 2004, 44, 575–581. [Google Scholar] [CrossRef] [PubMed]
  35. Torrens, F. Fractal Dimension of Transdermal-Delivery Drug Models. Leb. Sci. J. in press.
  36. Torrens, F. Fractal Dimension of Transdermal-Delivery Drug Models: 4‑Alkylanilines. Physica A. submitted for publication.
  37. Randić, M. On Characterization of Molecular Branching. J. Am. Chem. Soc. 1975, 97, 6609–6615. [Google Scholar] [CrossRef]
  38. Hosoya, H. Topological Index. A Newly Proposed Quantity Characterizing the Topological Nature of Structural Isomers of Saturated Hydrocarbons. Bull. Chem. Soc. Jpn. 1971, 44, 2332–2339. [Google Scholar] [CrossRef]
  39. Gálvez, J.; García, R.; Salabert, M.T.; Soler, R. Charge Indexes. New Topological Descriptors. J. Chem. Inf. Comput. Sci. 1984, 34, 520–525. [Google Scholar] [CrossRef]
  40. Kier, L.B.; Hall, L.H. Molecular Connectivity VII: Specific Treatment of Heteroatoms. J. Pharm. Sci. 1976, 65, 1806–1809. [Google Scholar] [CrossRef] [PubMed]
  41. Lakard, B. Analysis of Permanent Electric Dipole Moments of Aliphatic Amines. Internet Electron. Conference of Molecular Design 2003. [Google Scholar]
  42. Box, G.E.P.; Hunter, W.G.; MacGregor, J.F.; Erjavec, J. Some Problems Associated with the Analysis of Multiresponse Data. Technometrics 1973, 15, 33–51. [Google Scholar] [CrossRef]
  43. Hocking, R.R. The Analysis and Selection of Variables in Linear Regression. Biometrics 1976, 32, 1–49. [Google Scholar]
  44. McClellan, A.L. Tables of Experimental Dipole Moments; Freeman: San Francisco, 1963. [Google Scholar]

Share and Cite

MDPI and ACS Style

Torrens, F. Valence Topological Charge-Transfer Indices for Dipole Moments: Percutaneous Enhancers. Molecules 2004, 9, 1222-1235. https://doi.org/10.3390/91201222

AMA Style

Torrens F. Valence Topological Charge-Transfer Indices for Dipole Moments: Percutaneous Enhancers. Molecules. 2004; 9(12):1222-1235. https://doi.org/10.3390/91201222

Chicago/Turabian Style

Torrens, Francisco. 2004. "Valence Topological Charge-Transfer Indices for Dipole Moments: Percutaneous Enhancers" Molecules 9, no. 12: 1222-1235. https://doi.org/10.3390/91201222

Article Metrics

Back to TopTop