Volodymyr Shibanov INTERACTION BETWEEN MOLECULE FRAGMENTS UNBOUND BY CHEMICAL BONDS THROUGH THE SPACE DURING NMR 1 H SPECTRA REGISTRATION

The chemical shifts in NMR H spectra of the para-substituted propylbenzens of the general formula: pX-C6H4-CRRCH2CH3 (where R, R = H, CH3) have been analyzed. The presence or absence of the aryl fragment influence on the methyl end-group was observed by the value of the basic spectral parameters – the chemical shifts of methyl protons (δСН3) in comparison with analogous data of corresponding alkanes. The specific criteria for identifying such effect were developed and validated. We make the overall conclusion about high probability of the reciprocal intramolecular interactions between unbound fragments of the molecule in tert-amylbenzene and tertamylphenol (R = R = CH3).


Introduction
While analyzing the peculiarities of NMR 1 H and 13 C spectra of different classes of organic compounds we suppose that under recording spectra conditions the intramolecular interactions between unbound fragments of the molecule may take place through the space.The existence of the mentioned interactions leads, to our mind, to the observed changes in spectra compared with anticipated (expected) values.
The schematic drawing of investigated molecule containing fragments "K-L-M" is represented in Fig. 1.The arbitrary division into the fragments is in accordance with functional principle and depends upon the formulated aim.The aim is the investigation of NMR spectral parameters of the fragment "M" depending upon the structure of the fragment "K".The absence of chemical bonds between atoms of the fragments "K" and "M" is an indispensable condition.Both fragments are bound by chemical bonds only with "medium" fragment "L", with its opposite sides.In the linear conformation 1 the interaction of unbound fragments "K" and "M" is absent.It is possible in the curved conformation 2. To our mind, the deviation of spectral parameters of the fragment "M" from the anticipated values reveals this.
The transition states with curved conformations, the same as 2, are well-known in the synthetic chemistry, e.g. in the reactions of electrophylic cyclization or in substitution reactions proceeding with the transfer of reaction centre.Similar transition state was given in an article [1] describing NMR 17 O spectra of crowded alcohols, where authors postulated through-space interaction in CH 3 ---O.
We propose the following explanation of the observed phenomenon in accordance with the postulate: "The possible interaction between two energy states occurs always under given conditions if it leads to the decrease of the system total energy".
The changes of energy levels of two-component system taking place during their interaction are represented in Fig. 2. At the initial state "A" the interaction between energy levels of its components (A 1 and A 2 ) does not occur.The result is the system transition to its new state "B".In this connection two new energy levels (B 1 and B 2 ) are formed and a new system occupies a lower energy level (state B 1 ).
Lviv Polytechnic National University Institutional Repository http://ena.lp.edu.uaIt is logical to suppose that intramolecular interaction of unbound fragments of the molecule through the space in the conformation 2 leads to some energy gain (the decrease of the total energy of the system).It is the reason that the interaction occurs.
The supposed interaction in the structures "K-L-M" is expressed in the change of expected values of the "M" fragment spectral parameters.It was postulated by us earlier [2,3] for para-substituted alkylbenzenes by the general formula 3 (including monoalkylbenzenes, where X=H).In the molecules of mentioned compounds the fragment "K" is the alkyl group (Alk) and the fragment "M" includes nuclei of atoms H-2 (H-6), H-3 (H-5), С-1, C-2 (C-6) and C-3 (C-5) of the phenyl ring.
In the second example -phenylsulfones 4functional groups X are the fragments "K" and protons H-2, H-3 and H-4 [4,5] are the fragments "M".One more example shows aryl-containing methyl and ethyl ethers 5, where protons of alkoxyl groups are the fragments "M" and aryl groups (Ar) are the fragments "K" [6,7].
We also observed the same but not described earlier deviations from the anticipated values of δ H parameters in the NMR 1 H spectra for ethers, acetales, alcohols, saturated alkanes, etc.The variety of compounds for which the described phenomenon takes place allows us to assume that the observed results are widely spread and may be general for all classes of compounds.
As a result of interaction between "K" and "M" molecule fragments in the conformation 2 the changes in spectral parameters of the fragment "K" should be expected beside the described changes in spectral parameters of the fragment "M".Hitherto we did not observe such changes.There are several reasons for this fact.One of them is insufficient resolution of spectral lines in available NMR 1 H spectra obtained at lowfrequency instruments (see below).Thus we were not able to attribute the obtained signals.
But the main reason was the choice of logicallyfounded "expected" [7] spectral parameter necessary for its comparison with the experimental value.In this paper we describe the mentioned problems and ways of their solving as well as achieving the obtained results.
To prove the existence of changes in spectral parameters of the fragment "K" the structures of parasubstituted propylbenzenes 6-16 (where R, R 1 = H, CH 3 ) are the most suitable to our mind.These compounds are the part of wider class of alkylbenzenes according to the general formula 3.

Experimental
Selection of spectral data sources is based on their reliability and compatibility, criteria of which are discussed in [8].We used the values of basic spectral parameters δ CH3 H,N [9] and δ CH3 H,N [10] obtained in deuterochloroform as a solvent and taken from informational sources [9] and [10].The use of CDCl 3 as a solvent is grounded earlier [3,4], therefore spectra obtained in other solvents (for example CCl 4 or DMSOd 6 ) are not discussed here.
Signals attribution in NMR 1 H spectra.In all cases in [9] there is the author's attribution of spectral signals to the corresponding values of δ CH3 H,N [9] .Since in [10] the attribution of triplet signals of corresponding methyl groups is absent, we attributed them by ourselves.Usually the values of δ CH3 H,N [9] and δ CH3 H,N [10] parameters are in good agreement between each other and the difference between them is less than 0.020 ppm.The latter is accep-ted by us as the average possible experimental error (measurements accuracy), i.e. accuracy of δ CH3 H,N parameters determination.
Basic spectral parameters δ CH3 H,N [9] given in [9] were obtained using instruments with different frequency: low-frequency instrument (90 MHz) and high-frequency instrument (300 or 400 MHz).In those cases when two different values of δ CH3 H,N [9] are given in [9], we used the value obtained on high-frequency instrument.To our mind the most reliable are basic spectral parameters δ CH3 H,N [10] obtained on the instrument with frequency of 300 MHz and δ CH3 H,N [9] parameters obtained on the instrument with frequency of 400 MHz.The values of parameters of both types (experimental basic δ CH3 H,N and calculated differential ∆δ CH3 H,N , see below) are given with the accuracy of 0.001 ppm.
Besides basic spectral parameters δ CH3 H,N [10] and δ CH3 H,N [9] so called "experimental" differential spectral parameters [7] are represented in the Table .They were calculated in accordance with the formula given below.Also for the methyl group there are "expected values" of basic spectral parameter W and "expected values" of differential spectral parameters ∆W [9] and ∆W [10] .The definitions of all parameters are given below.

Results and Discussion
The comparison of basic spectral parameters δ Me H,N of methyl groups protons (given by italic as CH 3 in the formulas presented in the Table ) was carried out by two ways.In the first one (more obvious but less strict case) we partially used virtual values W. In the second one (more strict but less obvious case) we used only experimental data.

Introduced Terms
The definite systematic deviations of experimental values (δ exp H ) from their anticipated (expected) values δ ant H = W are the most obvious criterion of the presence of assumed interaction between unbound fragments "K" and "M" in the molecules of investigated alkylbenzenes 6-16.For quantitative comparison of such deviations we introduced virtual differential parameters To our mind, another set of experimental differential spectral parameters is more rigid.We denoted them as ∆δ CH3 H,N .Such parameters are calculated as the difference between experimental values: . The latter value is the corresponding basic spectral parameter δ Me Н,stand of the compound taken by us as a standard.The advantage of this set of differential parameters is using for calculations only experimental values.The main disadvantage is the absence of evident physical meaning.Differential "experimental" spectral parameters ∆δ CH3 H,N . As an example we calculate ∆δ CH3 Н for propylbenzene (6): ∆δ СН3 Н6 [10] = δ СН3 Н6 [10] δ СН3 Н17[10] = = 0.930 -0.880 = +0.050ppm.The subtrahend is the corresponding "standard" compound which is n-pentane (17) here.To calculate the differential "experimental" parameters of other compounds it is advisable to use other substances as standards depending on their structures."Anticipated values" of basic spectral parameters are virtual evaluation values.We denote them as capital Latin letters W N , e.g.W 6 .The numerical values of W N parameters are approximate and debatable.They are equal to those assumed values of methyl groups protons chemical shifts which would be in a case of absence of unbound fragments "K" and "M" interaction in the molecules of investigated compounds.Therefore under the term of "anticipated value" we mean logically grounded virtual value of the basic spectral parameter W N , i.e. the non-existent value of the signal we are interested in and which we would expect to see in NMR 1 H spectrum.
"Anticipated values" of differential spectral parameters, which we also call as virtual parameters, are calculated evaluation values.They are denoted by the symbol "∆W", e.g.∆W N .It is the difference between experimental basic parameter δ CH3 H and virtual anticipated parameter W, e.g.∆W 6 = δ Н6 -W 6 .

Criteria of the Presence of Supposed Interaction between Unbound
Fragments "K" and "M" in the Molecules of Investigated Compounds Earlier [7] we selected three main criteria: 1. Negative values of virtual differential parameters ∆W.The larger values the stronger interaction.
2. If the absolute values of ∆W parameters are close to zero, the founded assumptions about the presence or absence of the mentioned interaction are made with difficulty.
3. The negative values of "experimental" differential parameters ∆δ CH3 Н,N .The larger absolute value of the negative parameter the stronger the supposed interaction.
Therefore, negative values of differential parameters ∆W and ∆δ CH3 Н,N are given in the Table by bold (greater size).

Substantiation of Standard Compounds Choice
In above-mentioned example of "experimental" differential parameter calculation for propylbenzene 6 pentane 17 is chosen as a standard compound.Such a choice is explained by the following: 1. To compare NMR 1 H spectra of alkylbenzenes 6-16 and 24 we choose just the simplest alkanes as a standard compound, absorption of which takes place in the highest field.
The simplest alkane containing n-propyl radical -CH 2 -CH 2 -CH 3 is propane which is gaseous under usual conditions.Its formula differs from the formulas of investigated propylbenzenes 6, 9 and 12 by hydrogen atom instead of aryl fragment.The next homolog (n-butane, which is also gaseous compound) contains methyl group instead of aryl group.The NMR 1 H spectra of both compounds in CDCl 3 are absent in [9,10], therefore we could not use them as standard compounds.The next homolog is liquid n-pentane 17 which contains Lviv Polytechnic National University Institutional Repository http://ena.lp.edu.uaethyl group instead of aryl one attached to the propyl radical.Its NMR 1 H spectrum is given in both [9] and [10], therefore we can use it as a standard.In accordance with above-mentioned isopentane 18 and 3-methylpenatne 19 may be standard compounds for sec-butylbenzenes 7, 10 and 13-16 and 2,2-dimethylbutane 21 and 3,3-dimethylpentane 22 -for tert-amylbenzenes.

Statistic substantiation of chosen standard values δ CH3 H,st.
We chose compounds 17, 19 and 22 as standard compounds.In all of them the investigated radical is attached to the ethyl group.The same logic was used for "statistic substantiation" of standard value for tert-amylcontaining compounds 8 and 11.However here we met unanticipated problems.In the literature sources [9, 10] we found 3,3dimethyloctane 23 is a single homolog of 3,3dimethylpentane 22 from the direction of "large" radicals.Its δ CH3 Н,23 [9] value equals to 0. Therefore the choice of δ СН3 Н,st.= 0.790 ppm is substantiated by two arguments one of which we used earlier [11]."Short-chain" methyl group is the first in the homologous row of alkyl groups and that is why it may differ from "typical long-chain" alkyl groups (n-butyl and higher).Ethyl and n-propyl groups are intermediate ones between "short-chain" methyl groups and "long-chain" alkyl groups.But the spectral properties investigated in [11] allow to suppose that "intermediate" ethyl and n-propyl groups are closer to "long-chain" groups.As it is shown above, the same situation is observed for tert- The second argument.In our opinion the replacement of "short-chain" methyl group for aryl fragment (i.e. the transfer from the compound 21 to the compounds 8 and 11) would lead to more essential changes than replacement of "middle-chain" ethyl group in 22 and especially of "long-chain" pentyl group in 23.Therefore the value δ СН3 Н,st.= 0.790 ppm is more preferable than δ СН3 Н,st.= 0.830 ppm as a standard parameter for tert-amylcontaining compounds.

Comparison of "Experimental
Therefore due to the third criterion of the presence or absence of intramolecular interaction between unbound fragments it is probable that the interaction between methyl group of propyl fragment and hydrogen atoms of Lviv Polytechnic National University Institutional Repository http://ena.lp.edu.uaaryl fragment does not exist or has a small value in the investigated n-propylcontaining compounds 6, 9 and 12.
In contrast to n-propylaryl compounds 6, 9 and 12, the presence of negative by sign differential parameters ∆δ СН3 Н,N in the isobutyl fragment -CH(CH 3 )-CH 2 -CH 3 assumes the interaction between ethyl group of sec-butyl fragment and aryl ring in the molecules of compounds 7, 10 and 13-16, as well as 24 (which also may be attributed to this group of compounds).
The presence of considerable by value and negative by sign differential parameters ∆δ СН3 Н,N calculated for the protons of methyl end-group in -C(CH 3 ) 2 -CH 2 -CH 3 allow to sustain that the interaction between methyl group of tert-amyl fragment and aryl ring occurs in the molecules of the compounds 8 and 11.

Evaluation of Anticipated (Virtual) Basic Spectral Parameters W
We had a complicated task -to substantiate logically the suggested virtual values of W parameters for the compounds 6-16.
Earlier [7] we showed that values of W parameters mainly depend on the presence or absence of functional groups in the molecule structure.The presence of electron-attractive substituents shifts the W parameter toward a low field and vice versa.Comparing the experimental values of basic spectral parameters δ СН3 Н,N in substituted and unsubstituted compounds for three types of alkylbenzenes: propylbenzenes 6, 9, 12; secbutylbenzenes 7, 10, 13-16 and tert-amylbenzenes 8, 11 one can see that in all cases the substituents act according to above-mentioned principle.For example, in most cases of sec-butylbenzenes the electron-donating substituents (OH in 10, OMe in 15, OAc in 16 and NH 2 in 13) cause the methyl group absorption in a higher field and electronaccepting nitrogroup -in a lower field compared with unsubstituted sec-butylbenzene 7. The maximal difference between values of basic spectral parameters is 0.033 ppm (δ СН3 Н,N,14 = 0.833 ppm against δ СН3 Н,N,13 = 0.800 ppm).
Therefore it is necessary to take into account the influence of substituent in a phenyl ring while choosing the values of virtual parameters W.
The value of W parameter shows how much (in our opinion) the absorption of methyl end-group in alkylbenzenes 6-16 would be changed compared with that in n-alkanes 17, 19, 22 (which were chosen as standard compounds) while exchanging of the ethyl group for the aryl one.Obviously, due to the stronger electronaccepting influence of aryl group (compared with ethyl group) in alkylbenzenes 6-16 the shift of δ СН3 Н,N values toward the low field should take place.However we have no grounded suggestions concerning the value of such shift.Earlier in [7] we assumed that electron-accepting action of phenyl fragment is comparable with the action of iodine or bromine atoms as substituents.Therefore we investigated the values δ СН3 Н,N given in [9,10] for the row of 1-haloidpropanes, 2-haloidbutanes and 2-haloid-2methylbutanes, where bromine, iodine and chlorine atoms were used as haloids.Regardless of the type of haloid atom and structure of alkyl radical in haloid alkyl, the values of all founded parameters δ Ме Н are within the range from 0.90 to 1.10 ppm.The values δ Ме Н in the corresponding alcohols, esters and ethers have the same order of magnitude.Thus, it was advisable to accept the value of W parameters equal to 1.00 ppm for all compounds 6-16 irrespective of the aryl group structure.Taking into account the greater uncertainty committed while the choosing the value W, we do not take into account the less by value differences concerning the influence of substituents in the phenyl ring, as well as the presence (or absence) of methyl groups in the n-propyl fragment of these compounds.
Taking into account the above-mentioned peculiarities of the choice of virtual parameters W we extended the uncertainty interval for ∆W by sign and value from -0.100 ppm to +0.100 ppm.In spite of the considerable negative values of differential parameters ∆W N for n-propylcontaining compounds, they found themselves in this interval.Hence, parameters ∆W N cannot be con-Lviv Polytechnic National University Institutional Repository http://ena.lp.edu.uasidered applicable to determine the absence or presence of intramolecular interaction between unbound molecule fragments in n-propylcontaining compounds.The cautious conclusion may be done that the interaction between methyl group of propyl fragment and aryl fragment atoms in the compounds 6, 9, 12 does not occur or it is very small.
Therefore, in contrast to n-propyl compounds 6, 9, 12, the presence of considerable negative differential parameters ∆δ СН3 Н,N and virtual parameters ∆W calculated for the protons of methyl end-group in -CH(CH 3 )-CH 2 -CH 3 allows to assume that the interaction between methyl endgroup of sec-butyl fragment and aryl fragment atoms occurs in the compounds 7, 10, 13-16.More impressive results were obtained during investigations of virtual differential parameters ∆W for tert-amylbenzenes 8 and 11.Very large negative values of ∆W parameters are represented in the Table.They exceed 0.300 ppm: ∆W 8[9] = -0.310ppm for tert-amylbenzene 8; as well as ∆W 11[10] = -0.340ppm and ∆W 11 [9] = -0.320ppm for para-tert-amylphenol 11.
The presence of very large negative differential parameters ∆δ СН3 Н,N and ∆W N , calculated for the protons in methyl end-group in -C(CH 3 ) 2 -CH 2 -CH 3 allow to assert that the strong interaction between methyl group of tertamyl fragment and aryl ring occurs in the compounds 8 and 11.Thus, both "experimental" and "virtual" differential parameters of alkylbenzenes 7, 10 and 13-16 containing sec-butyl alkyl group and particularly compounds 8 and 11 containing tert-amyl fragment meet all criteria concerning the presence of intramolecular influence of phenyl ring through the space on methyl endgroup of alkyl fragment.
Taking into account that we postulated earlier [2, 3] the same influence of alkyl groups on phenyl ring (on its ortho-protons and carbon atoms C-1 and C-21 in particular), we may point to reciprocal influence of molecule fragments which are unbound by chemical bonds between each other.The circle is enclosed.The reciprocal influence most likely may be realized through the space, for example in "bent" conformation of tert-amylbenzene 8 (or tert-amylphenol 11) represented in Fig. 4. The conformations 1 and 2 (where X = H, Alk = CRR 1 CH 2 CH 3 and R 1 , R 2 = H or CH 3 ) are schematically represented in Figs. 3 and 4 respectively.Just by "partial" 2 influence of the system consisting of 6 annular π-electrons of phenyl ring on the distant methyl group of tert-amyl fragment (hydrogen atoms of which are inside the circles) we explain the unique by value shift of its protons toward the high yield (till value δ СН3 Н < 0.70 ppm).