Significant Substituent Effects on Pyridinolysis of Aryl Ethyl Chlorophosphates in Acetonitrile

Abstract

The substituent effects on the pyridinolysis (XC5H4N) of Y-aryl ethyl chlorophosphates are investigated in acetonitrile at $35.0^{\circ}C$. The two strong ${\pi}$-acceptor substituents, X = 4-Ac and 4-CN in the X-pyridines, exhibit large positive deviations from the Hammett plots but little positive deviations from the Br$\ddot{o}$nsted plots. The substituent Y effects on the rates are really significant and the Hammett plots for substituent Y variations in the substrates invariably change from biphasic concave downwards via isokinetic at X = H to biphasic concave upwards with a break point at Y = 3-Me as the pyridine becomes less basic. These are interpreted to indicate a mechanistic change at the break point from a stepwise mechanism with a rate-limiting bond formation (${\rho}_{XY}$ = -6.26) for Y = (4-MeO, 4-Me, 3-Me) to with a rate-limiting leaving group expulsion from the intermediate (${\rho}_{XY}$ = +5.47) for Y = (4-Me, H, 3-MeO). The exceptionally large magnitudes of ${\rho}_{XY}$ values imply frontside nucleophilic attack transition state.

Introduction

The cross-interaction constants (CICs) are one of the strong tools to clarify the reaction mechanism by means of the substituent effects on the rates.1 The CIC, ρij, is defined as Eqs. (1) and (2) where i and j represent the substituent X in the nucleophiles and/or Y in the substrates and/or Z in the leaving groups, respectively. A Taylor series expansion of log kij around σi = σj = 0 (i = j = H) leads to Eq. (1). Pure second- (e.g., ρXXσX2 or ρYYσY2 or ρZZσZ2), third- (e.g., ρXXYσX2σY or ρXYYσXσY2, etc) and higher-derivative terms are not considered because they are normally too small to be taken into account. The magnitude of ρij is inversely proportional to the distance between i and j through the reaction center.1

Eq. (1) can be written in a different form of Eq. (3):

When the second term of the right-side of Eq. (3) is equal to zero, ρjH + ρijσi = 0 with σi = σi,iso = –ρjH/ρij, and these relations are substituted into Eq. (3), the result is the below Eq. (4).

Then the value of kij becomes constant, i.e., kij,iso where the subscript ‘iso’ indicates isokinetic, for all j because the values of ρiH, ρjH and ρij are constant.2 When the interaction between the substituent i and j is so strong that the magnitude of the ρij value is great, the signs of the selectivity parameters (ρX, ρY, or ρZ) are sometimes changing consistently from positive (ρj ＞ 0) via null (ρj = 0; isokinetic) to negative (ρj ＜ 0) as the substituent i becomes more electronwithdrawing (or -donating) group.3 The reactions of some substrates with X-anilines gave isokinetic phenomena, ρj = 0 at σi,iso, as follows: (i) 1-Y-aryl ethyl chloride with ρXY = –2.05, giving ρX= 0 at σY,iso = –0.23;4 (ii) Y-benzhydryl chlorides with ρXY = –1.46, giving ρX = 0 at σY,iso = 0.22;5 (iii) Y-benzoyl bromides with ρXY = –0.62, giving ρY = 0 at σX,iso = 0.94;6 (iv) cumyl Z-arenesulfonates with ρXZ = –0.75, giving ρZ = 0 at σX,iso = 0.83.7

In the present work, the nucleophilic substitution reactions of Y-aryl ethyl chlorophosphates (1) with X-pyridines are studied kinetically in acetonitrile (MeCN) at 35.0 ± 0.1 oC (Scheme 1). The aim of this work is to gain further information on the substituent effects of the nucleophiles and substrates on the reaction mechanism mainly based on the CICs, free energy relationships and strong π-acceptor parasubstituents in the pyridines.

Scheme 1.Pyridinolysis of Y-aryl ethyl chlorophosphates (1) in MeCN at 35.0 ℃.

 

Results and Discussion

Tables 1-3 list the second-order rate constants (k2/M–1 s–1),Hammett (ρX) and Brönsted (βX) coefficients with X, and Hammett coefficients (ρY) with Y in MeCN at 35.0 ℃. For convenience, henceforth, the substituent X in the nucleophiles and Y in the substrates are divided into two blocks, respectively, as follows: (i) s-block with X = (4-MeO, 4-Me, 3-Me); (ii) w-block with X = (3-Ph, 3-Cl, 4-Ac, 3-CN, 4- CN); (iii) l-block with Y = (4-MeO, 4-Me, 3-Me); and (iv) r-block with Y = (3-Me, H, 3-MeO).8 The rate is faster with a stronger nucleophile which is compatible with a typical nucleophilic substitution reaction with partial positive charge development at the nucleophile N atom in the transition state (TS). The two strong π-acceptor substituents (X = 4-Ac, 4-CN), however, exhibit great positive deviations from the Hammett plots (Fig. 1) while little positive deviation from the Brönsted plots (Fig. 2). The rate with Y is not consistent with a typical nucleophilic substitution reaction. The Hammett plots (Fig. 3) for substituent Y variationse show a break point at Y = 3-Me, and gradually change from biphasic concave downwards with s-block, via linear (almost isokinetic) with X = H, to biphasic concave upwards with w-block. These phenomena with both biphasic concave downward and upward free energy correlations depending on the substituents are unprecedented one, showing the surprising substituent effects on the kinetics and mechanism. Note that (i) isokinetic phenomena are observed with both l- and r-block, ρY ≈ 0, ρY = 0.04 of l-block and ρY = –0.03 of r-block, at σX,iso ≈ 0 or X = H; and (ii) thus, unusual negative values of ρY (＜ 0) are obtained with both w,l- and s,r-block as seen in Figure 3. The negative value of ρY indicates partial positive charge development at the oxygen atom of the phenoxy ligand in the TS, contrary to a typical nucleophilic substitution reaction.

Table 1.aThe sequence of the X-pyridines in the first column is followed the order of the corresponding σX value, neither second-order rate constant nor pKa value. The order of the pKa values of the X-pyridines is as follows: X = 4-MeO ＞ 4-Me ＞ 3-Me ＞ H ＞ 3-Ph ＞ 3-Cl ＞ 4-Ac ＞ 4-CN ＞ 3-CN. bThe sequence of the Y-substrates in the first row is followed the order of the corresponding σY value.

Table 2.aTwo strong π-acceptor, X = (4-Ac, 4-CN), are excluded. bAll X.

Table 3.aY = (4-MeO, 4-Me, 3-Me). bY = (3-Me, H, 3-MeO).

Figure 1.Hammett plots with X of the reactions of 1 with XC5H4N in MeCN at 35.0 ℃.

Figure 2.Brönsted plots with X of the reactions of 1 with XC5H4N in MeCN at 35.0 ℃.

Figure 3.Hammett plots with Y of the reactions of 1 with XC5H4N in MeCN at 35.0 ℃.

The two strong π-acceptor para-substituents, X = (4-Ac, 4-CN), exhibit great positive deviations from the Hammett plots for substituent X variations regardless of the nature of Y as seen in Figure 1. This behavior indicates that the two π-acceptor substituents yield exalted reactivity. The exalted basicity (or enhanced nucleophilicity) of the strong π-acceptor groups would be owing to the weak π-donor effects.9 The Hammett σp values of the π-acceptor substituents represent the inductive and π-electron-withdrawing effects. However, the experimental pKa value only represents the inductive effect of X, because protonation/deprotonation takes place at the σ lone pair on N which is orthogonal to the ring π- system.9a As a result, the protonation/deprotonation does not disturb the ring π-system, but the positive charge center in the conjugate acid, naturally, attracts π-electrons inductively without through-conjugation between the σ-lone pair and the π-acceptor para-substituent. Thus, the pKa values of π- acceptor substituents correctly reflect the substituent effects when the N atom of pyridine becomes positively charged in the TS because the determination of pKa involves a positive charge on N atom (azonium type).

As observed in the present work, the two π-acceptor substituents exhibited positive deviations from the Hammett plots, while little deviations from the Brönsted plots, for the pyridinolyses of (i) methyl chloroformate in MeCN9a and water;10 (ii) Y-benzenesulfonyl chlorides in MeOH;11 (iii) Y-benzyl bromides in DMSO;12 and (iv) Y-phenacyl bromides in MeCN.13 These indicate that the N atom of pyridine becomes positively charged in the TS, and that the degree of the bond formation is considerably extensive. On the contrary, the two π-acceptor substituents did not exhibit deviations from either the Hammett or Brönsted plots for the pyridinolysis of Y-aryl phenyl chlorophosphates.14 No positive deviations for the π-acceptor in both plots were rationalized by the early TS with little positive charge development on the N atom of pyridine. The early TS, in which the extent of both the bond formation and leaving group departure is small, was supported by the small magnitudes of Brönsted coefficients and CIC: βX = 0.16-0.18 and ρXY = –0.15.14

Figure 4 shows the determination of ρXY according to Eq. (2), ρXY = ∂ρY/∂σX = ∂ρX/∂σY, giving the great magnitudes of CICs: ρXY = –6.26 with l-block and ρXY = +5.47 with r-block.15 It is the suggestion of the authors that the reaction proceeds through a stepwise process with a rate-limiting bond formation with l-block while through a stepwise process with a rate-limiting leaving group departure from the intermediate with r-block, based on the sign of ρXY, negative with l-block while positive with r-block.16 Isokinetic phenomena are observed, ρY ≈ 0 at X = H for both l-and r-block, due to the great magnitudes of CICs. The values of σX,iso and kXY,iso can be calculated from Eq. (4);17 these values for lblock are as follows;

Figure 4.Plots of ρY (or ρX) vs σX (or σY) to calculate the ρXY values of the reactions of 1 with X-pyridines in MeCN at 35.0 ℃. The obtained ρXY values by multiple regressions are: (a) ρXY = –6.26 ± 0.13 (r = 0.995) with l-block; (b) ρXY = +5.47 ± 0.15 (r = 0.994) with r-block. Note that the two strong π-acceptor X = (4-Ac, 4-CN) are not considered to calculate the ρXY values.

σX,iso = –ρXHρYH/ρXY = –(0.04)/(–6.26) = 0.0064(≈ 0; X = H); kXY,iso = 32.3 × 10–3/M–1 s–1

and these values for r-block are as follows;

σX,iso = –ρXHρYH/ρXY = –(0.03)/(–5.47) = 0.0055(≈ 0; X = H); kXY,iso = 32.7 × 10–3/M–1 s–1

The magnitudes of ρXY for both l- and r-block are exceptionally great. The obtained magnitudes of ρXY (= –6.26 and +5.47) are the unprecedented large values for the phosphoryl transfer reactions studied in this lab.15 The unusual large magnitudes of ρXY imply that the nucleophile and substrate are in close enough proximity to interact ‘very’ strongly. In other words, the degree of the bond formation is really extensive in the TS for both l- and r-block. This suggestion is in agreement with the results of the behavior of the two strong π-acceptor para-substituents, X = 4-Ac and 4-CN in X-pyridines which is indicative of the ‘very’ extensive bond formation and positive charge development on N atom in the TS. The equatorial nucleophilic attack should lead to a tighter P–N bond in the trigonal bipyramidal pentacoordinate (TBP-5C) structure,18 because the equatorial bonds are shorter than the apical bonds.19 Hence a larger magnitude of ρXY is obtained compared to the apical nucleophilic attack. Thus the authors propose the frontside equatorial attack TS (Scheme 2) based on the large magnitudes of ρXY for both land r-block, and the behavior of the two strong π-acceptor para-substituents, X = 4-Ac and 4-CN.

As mentioned earlier: (i) the nitrogen atom of the pyridine becomes considerably positively charged in the TS based on the behavior of the two strong π-acceptor para-substituents; (ii) very extensive bond formation occurs in the TS based on the large magnitudes of ρXY for both l-and r-block; and (iii) partial positive charge develops on the oxygen atom of the phenoxy ligand with both w,l- and s,r-block in the TS based on the negative ρY value. Accordingly, the TS structures and charge distribution with four-blocks [(i) s,l-; (ii) s,r-; (iii) w,l-; and (iv) w,r-block] are described in Scheme 2. It should be noted that the description of the charge distribution in the TS is nothing but qualitative, never quantitative, to achieve the electronic balance. The negative ρY values with w,l- and s,r-block are observed when one substituent (X or Y) is electron-donating and the other (X or Y) is electron-withdrawing group. On the contrary, the positive ρY values with s,l- and w,r-block are observed when both substituent, X and Y, are either electron-donating or electron-withdrawing groups.

Scheme 2.Proposed frontside equatorial attack TS structures with: (i) s,l-; (ii) s,r-; (iii) w,l-; and (iv) w,r-block.

 

Experimental Section

Materials. Y-aryl ethyl chlorophosphates were prepared as previously described.20 The physical constants of Y = (4- MeO, 4-Me, H, 3-MeO) were reported earlier20 and those of ethyl 3-methylphenyl chlorophosphate were as follows (supporting information):

(C2H5O)(3-CH3-C6H4O)P(=O)Cl: Colorless oily liquid. 1H NMR (200 MHz, CDCl3) δ 1.46 (t, 3H), 2.34 (s, 3H), 4.34-4.43 (m, 2H), 7.03-7.24 (m 4H), 13C NMR (100 MHz, CDCl3) δ 15.67, 21.17, 66.58, 116.82-149.68; νmax (neat), 3060-2979 (Arom. Str.), 2919-2865 (Alph. Str.), 1616, 1583, 1496, 1306 (P=O str.), 1154 (P-O-Ph Str.), 790 (P-Cl str.); EI-MS m/z 234 (M).

Kinetic Measurements. The second-order rate constants and selectivity parameters were obtained as reported earlier.13,14 For the present work, the concentrations of [substrate] = 3 × 10–3 M and [X-pyridine] = (0.1-0.3) M were used.

Product Analysis. Ethyl 4-methoxyphenyl chlorophosphate was refluxed with equimolar amount of 4-acetylpyridine for more than 15 half-lives in MeCN at 35.0 ℃. Solvent was evaporated under reduced pressure. Then 5 mL 50% ethylacetate/n-hexane mixed solution was added to it for washing. Several attempts were taken for this purpose. Solvent was then removed under oil-diffusion pump to finalize reddish-brown oily liquid product. The physical constants of product were as follows (supporting information):

[(C2H5O)(4-CH3O-C6H4O)P(=O)(4-CH3CO-C5H4N)]+Cl–: Reddish-brown oil. 1H NMR (400 MHz, CD3CN) δ 1.24 (t, 3H, CH3), 2.61 (s, 3H, CH3), 3.71 (s, 3H, OCH3), 4.05 (m, 2H, CH2), 6.71-7.10 (m, 4H, Arom.), 7.96 (d, 2H, Pydn. J = 1.6 Hz), 8.75 (d, 2H, Pydn. J = 1.7 Hz); 13C NMR (100 MHz, CD3CN) δ 16.64, 27.47, 56.27, 64.22, 115.47-154.44, 122.22, 126.25, 130.42, 148.19, 197.77; 31P NMR (162 MHz, CDCl3) d 4.80 (1P, s, P=O). LC-MS m/z 388 (M+).