Synthesis of stable phosphorus ylides from 3, 5-dimethylpyrazole and kinetic investigation of the reactions by UV spectrophotometry

Stable crystalline phosphorus ylides were obtained in excellent yields from the 1:1:1 addition reactions between triphenylphosphine and dialkyl acetylenedicarboxylates, in the presence of a NH-acid, such as 3,5-dimethylpyrazole. These stable ylides exist in solution as a mixture of two geometrical isomers as a result of restricted rotation around the carbon-carbon partial double bond resulting from conjugation of the ylide moiety with the adjacent carbonyl group. To determine the kinetic parameters of the reactions, they were monitored by UV spectrophotometery. The second order fits were drawn and the values of the second order rate constant (k 2 ) were calculated automatically using standard equations within the program. At the temperature range studied, the dependence of the second order rate constant (Ln k 2 ) on reciprocal temperature was in good agreement with the Arrhenius equation. This provided the relevant plots to calculate the activation energy of all reactions. Furthermore, useful information was obtained from studies of the effect of solvent, structure of reactants (different alkyl groups within the dialkyl acetylenedicarboxylates) and also concentration of reactants on the rate of reactions. The proposed mechanism was confirmed according to the obtained results and steady state approximation and first step (k 2 ) of reaction was recognized as a rate-determining step on the basis of experimental data.


Introduction
Development of simple synthetic routes for widely used organic compounds from readily available reagents is one of the major tasks in organic chemistry. 1Phosphorus ylides are reactive systems, which take part in many valuable reactions in organic synthesis [2][3][4][5][6][7][8][9][10][11][12] and are most often prepared by treatment of a phosphonium salt with a base.Most of the phosphonium salts are prepared from the reaction of a phosphine and an alkyl halide [3][4][5][6][7] though they can be obtained also by Michael addition of phosphorus nucleophiles to activated olefins. 2,3Herein, we wish to describe an efficient synthetic route to, and kinetic study of, stable phosphorus ylides using triphenylphosphine, dialkyl acetylenedicarboxylates and a NH-acid such as 3,5dimethylpyrazole.As noted earlier, the 3,5-dimethylpyrazole moiety and its derivatives are widely used in the preparation of medicines. 13With respect to the importance of this purpose, the present work was undertaken for generation of stable phosphoranes.Accordingly, the reaction of triphenylphosphine 1 with dialkylacetylene dicarboxylates 2 (2a or 2b) in the presence of a NH-acid 3 led to the corresponding stable heterocyclic phosphorus ylides 4 (4a or 4b) in excellent yields (see Figure 1).

Results and Discussion
The reactions of 3,5-dimethylpyrazole 3 with dialkyl acetylenedicarboxylates 2 in the presence of triphenylphosphine 1 were carried out in diethyl ether solvent at room temperature and were complete within a few hours.The 1 H and 13 C NMR spectra of the crude products indicated clearly the formation of stable phosphorus ylides 4. Any products other than 4 could not be detected by NMR spectroscopy.The structures of compounds 4a,b were deduced from their IR, 1 H, 13 C and 31 P NMR spectra.Their mass spectra displayed molecular ion peaks at appropriate m/z values and initial fragmentations involved the loss of parts or the complete side chains and scission of the heterocyclic ring system.The 1 H, 13 C, and 31 P NMR spectra of ylides 4a are consistent with the presence of two isomers.The ylide moiety of these compounds is strongly conjugated with the adjacent carbonyl group and rotation around the partial double bond in (E)-4 and (Z)-4 geometrical isomers is slow on the NMR timescale at ambient temperature (see Figure 2).Selected 1 H, 13 C, and 31 P NMR chemical shifts and coupling constants in compound 4b and the major (M) and minor (m) geometrical isomers of compound 4a are shown in Table 1.As can be seen, only one geometrical isomer was observed for 4b, presumably, because of the bulky tert-butyl groups.The 1 H NMR spectra (500 MHz) of compound 4a displayed four triplets (at δ= 0.42 and 1.19, 3 J HH =7.2 Hz, δ=1.24 and 1.31, 3 J HH =7.1 Hz,) arising from methyl protons along with signals for methine protons at δ= 5.45 and 5.14 ppm, which appear as two doublets ( 3 J PH =19.4 Hz) and ( 3 J PH =19.1 Hz), respectively, for the major and minor geometrical isomers.The 13 C NMR spectrum of 4a exhibited 34 distinct resonances that are in a good agreement with the mixture of two rotational isomers.Although the presence of the 31 P nucleus complicates both the 1 H and 13 C NMR spectra of 4a, it helps in assignment of the signals by long-range couplings with the 1 H and 13 C nuclei (see Experimental section).The 1 H and 13 C NMR spectra of 4b are similar to those of 4a, except for the ester groups, which exhibited characteristic resonances with appropriate chemical shifts (see Experimental section).The 1 H and 13 C NMR spectral data for compounds 4a,b are consistent with the geometrical isomers.The structural assignments made on the basis of the 1 H and 13 C NMR spectra of compounds 4a,b were supported by the IR spectra.The carbonyl region of the spectra exhibited two distinct absorption bands for each compound (see Experimental section).Of special interest is the ester absorption at 1726-1631cm - 1 for these compounds.Conjugation with the negative charge appears to be a plausible factor in the reduction of the wave numbers of the carbonyl absorption bands.On the basis of the well established chemistry of trivalent phosphorus nucleophiles, [2][3][4][5][6] it is reasonable to assume that phosphorus ylide 4 results from the initial addition of triphenylphosphine to the acetylenic ester 2 (rate constant k 2 ) and subsequent protonation of the 1:1 adduct by the NH-acid 3 ( rate constant k 3 ) to form phosphoranes 4 (rate constant k 4 ) (see Figure 3).

Kinetic studies
The synthetic reactions between triphenylphosphine 1, dialkyl acetylenedicarboxylates 2 and CH, SH or NH-acids have been reported earlier [14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32] but the kinetic study of these reactions has not been investigated yet.Herein, we wish to describe a kinetic investigation of the current synthesis.6][37][38][39][40][41][42][43][44][45][46] In order to gain further insight into the reaction mechanism, a kinetic study of the reactions was undertaken by the UV spectrophotometric technique.First, it was necessary to find the appropriate wavelength in order to follow the kinetic study of the reaction.For this purpose, in the first experiment, a 3×10 -3 M solution of compounds 1, 2b and 3 was prepared in 1,4-dioxane as solvent.An approximately 3 mL aliquot from each reactant was pipetted into a 10 mm light path quartz spectrophotometer cell and the relevant spectra were recorded over the wavelength range 190-400 nm.Figures 4, 5 and 6 show the ultraviolet spectra of compounds 1, 2b and 3, respectively.In a second experiment, a 1mL aliquot from the 3×10 -3 M solutions of each compound 1 and 3 was pipetted first into a quartz spectrophotometer cell (as there is no reaction between them).Later, a 1mL aliquot of the 3×10 - 3 M solution of reactant 2b was added to the mixture and the reaction monitored by recording scans of the entire spectrum every 13 min over the whole reaction time at ambient temperature.The ultraviolet spectra shown in Figure 7 are typical.From these, the appropriate wavelength was found to be 330 nm (corresponding mainly to product 4b) since at this wavelength, compounds 1, 2b and 3 have relatively no absorbance value.This, then, provided the opportunity to fully investigate the kinetics of the reaction between triphenylphosphine 1, di-tert-butyl acetylenedicarboxylate 2b and 3,5-dimethylpyrazole 3 at 330 nm in the presence of 1,4-dioxane as solvent.Since the spectrophotometer cell of the UV equipment had a 10 mm light-path cuvette, the UV/Vis spectra of compound 4b were measured over the concentration range (2×10 - 4 M ≤ M 4b ≤10 -3 M) to check for a linear relationship between absorbance values and     For each kinetic experiment, first a 1mL aliquot from each freshly made 3×10 -3 M solution of compounds 1 and 3 in 1, 4-dioxane was pipetted into a quartz cell, and then a 1mL aliquot of the 3×10 -3 M solution of reactant 2b was added to the mixture, keeping the temperature at 15.0ºC.The reaction kinetics were followed by plotting UV absorbance against time.Figure 8 shows the absorbance change versus time for the 1:1:1 addition reaction between compounds 1, 2b and 3 at 15.0ºC.The infinity absorbance (A ∞ ), that is the absorbance at reaction completion, can be obtained from Figure 8 at t =395 min.With respect to this value, zero, first or second curve fitting could be drawn automatically for the reaction by the software 47 associated with the UV instrument.Using the original experimental absorbance versus time data provided a second order fit curve (full line) that fits exactly the experimental curve (dotted line) as shown in Figure 9. Thus, the reaction between triphenylphosphine 1, di-tert-butyl acetylenedicarboxylate 2b and 3,5-dimethylpyrazole 3 follows second-order kinetics.The second-order rate constant (k 2 ) is then automatically calculated using a standard equation 37 within the program at 15.0ºC.It is reported in Table 2. Further kinetic studies were carried out using the same concentration of each reactant in a series of experiments with concentrations of 5×10 -3 M and 7×10 -3 M, respectively.As expected, the second order rate constant was independent of concentration and its value was the same as in the previous experiment.In addition, the overall order of reaction was also 2.

Effect of solvents and temperature
To determine the effect of change in temperature and solvent environment on the rate of reaction, it was elected to perform various experiments at different temperatures and solvent polarities, but otherwise under the same conditions as for the previous experiment.For this purpose, ethyl acetate and 1,2-dichloroethane, with dielectric constants of 6 and 10, respectively, were chosen as suitable solvents since they not only dissolved all compounds but also did not react with them.The effects of solvents and temperature on the rate constant are given in Table 2.The results show that the rate of reaction in each solvent was increased at higher temperature.In addition, the rate of reaction between 1, 2b and 3 was accelerated in a higher dielectric constant environment (1,2-dichloroethane) in comparison with a lower dielectric constant environment (1,4-dioxane) at all temperatures investigated.In the temperature range studied, the dependence of the second-order rate constant (Ln k 2 ) of the reactions on reciprocal temperature is consistent with the Arrhenius equation, giving activation energies of reaction between 1, 2b and 3 (26.4kJ/mole) from the slope of Figure.

Effect of concentration
To determine reaction order with respect to triphenylphosphine 1 and dialkyl acetylenedicarboxylate 2 (2b), in a series of separate experiments, all kinetic studies were carried out in the presence of excess 3.Under this condition, the rate equation may therefore be expressed as: In this case (3×10 -2 M of 3 instead of 3×10 -3 M) using the original experimental absorbance versus time data provides a second order fit curve (full line) against time at 330 nm which exactly fits the experimental curve.The value of the rate constant was the same as that obtained from the previous experiment (3×10 -3 M).Repetition of the experiments with 5×10 -2 M and 7×10 - 2 M concentrations of 3 gave, separately, the same fit curve and rate constant.In fact, the experimental data indicated that the observed pseudo second order rate constant (k obs ) was equal to the second order rate constant (k 2 ); this is possible when γ is zero in equation (I).It appears, therefore, that the reaction is zero and second order with respect to 3 (NH-acid) and the sum of 1 and 2 (2b) (α + β = 2), respectively.
To determine the reaction order with respect to dialkyl acetylenedicarboxylate 2 (2b), a separate experiment was performed in the presence of excess of 1 (rate = k΄ obs [3] γ [2] β , k΄ obs = k 2 [1] α (II)).The original experimental absorbance versus time data are shown in Figure 10.These data provide a pseudo first order fit curve at 330 nm, which exactly fits the experimental curve (dotted line) as shown in Figure 11.As a result since γ = 0 (as determined previously), it is reasonable to accept that the reaction is first order with respect to compound 2 (2b) (β = 1).Because the overall order of reaction is two (α + β + γ = 2) it is obvious that α = 1 and the order of triphenylphosphine 1 must be equal to one.This observation was obtained also for reaction between 1, 2a and 3. Based on the above results, a simplified proposed reaction mechanism (Figure 3) is shown in Figure 12.The experimental results indicate that the third step (rate constant k 3 ) is possibly fast.In contrast, it may be assumed that the third step is the rate-determining step for the proposed mechanism.In this case the rate law can be expressed as follows: ] The steady state assumption can be employed for [I 1 ] which generates the following equation, The value of [I 1 ] can be replaced in equation (III) to obtain this equation: Since it was assumed that k 3 is relevant to the rate determining step, it is reasonable to make the following assumption : k -2 >> k 3 [3]   so the rate equation becomes: The final equation indicates that the overall reaction order is three, which is not compatible with the experimental overall order of reaction (= two).In addition, according to this equation the order of reaction with respect to 3,5-dimethylpyrazole 3 is one, whereas it was actually shown to be equal to zero.For this reason, it appeared that the third step is fast.If we assume that the fourth step (rate constant k 4 ) is the rate-determining step for the proposed mechanism, in this case, there are two ionic species to consider in the rate-determining step, namely phosphonium ion (I 2 ) and pyrazole ion (N -).The phosphonium and pyrazole ions, as we see in Figure 3, have full positive and negative charges and form very powerful ion-dipole bonds to the 1,2dichloroethane, the high dielectric constant solvent.However, the transition state for the reaction between the two ions carries a dispersed charge, which here is divided between the attacking pyrazole and the phosphonium ions.Bonding of solvent (1,2-dichloroethane) to this dispersed charge would be much weaker than to the concentrated charge of pyrazole and phosphonium ions.The solvent thus would stabilize the species ions more than it would the transition state, and therefore E a would be higher, slowing down the reaction.However, in practice, 1,2dichloroethane speeds up the reaction and, for this reason, the fourth step, which is independent of the change in the solvent medium, could not be the rate-determining step.Furthermore, the rate law of formation of the product (fourth step) for a proposed reaction mechanism with application of the steady state assumption can be expressed by: By application of steady state for [I -] and [N -], and replacement of their values in the above equation, the following equation is obtained: ] This equation is independent of the rate constant for the fourth step (k 4 ) and shows why the fourth step would not be affected by a change in the solvent medium.9] If the first step (rate constant k 2 ) were the rate determining step, in this case, two reactants (triphenylphosphine 1 and dialkyl acetylenedicarboxylate 2), as we see (Figure 3), have no charge and could not form strong ion-dipole bonds to the high dielectric constant solvent, 1,2dichloroethane.However, the transition state carries a dispersed charge which here is divided between the attacking 1 and 2 and, hence, bonding of solvent to this dispersed charge is much stronger than to the reactants, which lack charge.The solvent thus stabilizes the transition state more than it does the reactants and, therefore, E a is reduced which speeds up the reaction.Our experimental results show that the solvent with the higher dielectric constant exerts a power full effect on the rate of reaction (in fact, the first step has rate constant k 2 in the proposed mechanism) but the opposite occurs with the solvent of lower dielectric constant (1,4-dioxane, see tables 2 and 3).The results of the current work (effects of solvent and concentration of compounds) have provided useful evidence for steps 1 (k 2 ), 3 (k 3 ) and 4 (k 4 ) of the reactions between triphenylphosphine 1, dialkyl acetylenedicarboxylate 2 (2a or 2b) and 3,5dimethylpyrazole 3. Two steps involving 3 and 4 are not determining, although the discussed effects, taken altogether, are compatible with the first step (k 2 ) of the proposed mechanism and would allow it to be the rate-determining step.However, a good kinetic description of the experimental result using a mechanistic scheme based upon the steady state approximation is frequently taken as evidence of its validity.By application of this, the rate formation of product 4 from the reaction mechanism (Figure 12) is given by: We can apply the steady-state approximation to [I 1 ] and [ To obtain a suitable expression for [I 2 ] to put into equation (V) we can assume that, after an initial brief period, the concentrations of [I 1 ] and [I 2 ] achieve a steady state with their rates of formation and rates of disappearance just balanced.Therefore, We can now replace [I 1 ] in the equation (VI) to obtain this equation: The value of [I 2 ] can be put into equation (V) to obtain the rate equation (VIII) for the proposed mechanism: Since experimental data have indicated that steps 3 (k 3 ) and 4 (k 4 ) are fast but step 1 (k 2 ) is slow, it is therefore reasonable to make the following assumption: k 3 [3] >> k -2 so the rate equation becomes: This equation, which was obtained from a mechanistic scheme (shown in Figure 12) by applying the steady-state approximation, is compatible with the results obtained by UV spectrophotometry.

Further kinetic investigations Effect of structure of dialkyl acetylenedicarboxylates
To confirm the above observations, further experiments were performed with diethyl acetylenedicarboxylate 2a under the same conditions used in the previous experiments.The values of the second-order rate constant (k 2 ) for the reaction between 1, 2a and 3 are reported in Table 3 for all solvents and temperatures investigated.The original experimental absorbance curves (dotted line) and the second order fit curve (full line), which exactly fits the experimental curve (dotted line), are shown in Figures 13 and 14     As can be seen from Table 3 the behavior of diethyl acetylenedicarboxylate 2a is the same as for di-tert-butyl acetylenedicarboxylate 2b (Table 2) with respect to the reaction with triphenylphosphine 1 and 3,5-dimethylpyrazole 3. The rate of the former reaction was also accelerated in a higher dielectric constant environment and with higher temperatures, however, this rate under the same conditions is approximately 5 to 6 times greater than for the reaction with di-tert-butyl acetylenedicarboxylate 2b (see Tables 2 and 3).It seems that both inductive ISSN 1424-6376 Page 182 © ARKAT and steric factors for the bulky alkyl groups in 2b tend to reduce the overall reaction rate (see equation IX).In the case of diethyl acetylenedicarboxylate 2a, the lower steric and inductive effects of the diethyl groups exert a powerful effect on the rate of reaction.

Conclusions
Briefly, we have prepared novel phosphorus ylides using a one-pot reaction between triphenylphosphine and dialkyl acetylenedicarboxylates in the presence of a NH-acid such as 3,5 dimethylpyrazole.The present method carries the advantage that, not only is the reaction performed under neutral conditions, but also the substances can be mixed without any activation or modifications.The stable phosphorus ylides 4 (4a and 4b) are potentially useful synthetic intermediates and the procedure described here may be acceptable for the preparation of phosphoranes with variable functionalities.In addition, kinetic investigations of these reactions were undertaken using UV spectrophotometry.The results can be summarized as follows; 1.The appropriate wavelengths and concentrations were determined in order to follow the reaction kinetics.2. The overall reaction order followed second-order kinetics and the reaction orders with respect to triphenylphosphine, dialkyl acetylenedicarboxylate and 3,5-dimethylpyrazole were one,one and zero, respectively. 3. The values of the second-order rate constants for all reactions were calculated automatically with respect to the standard equation using the software associated with the Cary-300 UV instrument.4. The rates of all reactions were accelerated at high temperature.Under the same conditions, the activation energy of the reaction with di-tertbutyl acetylenedicarboxylate 2b (33.9 kJ/mol) was higher than that for the reaction with diethyl acetylenedicarboxylate 2a (18.9 kJ/mol) in 1,4-dioxane.5.The rates of all reactions were increased in solvents of higher dielectric constant and this can be related to differences in stabilization by the solvent of the reactants and the activated complex in the transition state.6. Increased steric bulk in the alkyl groups of the dialkyl acetylenedicarboxylate, accompanied by the correspondingly greater inductive effect, reduced the overall reaction rate. 7.With respect to the experimental data, the first step of proposed mechanism was recognized as a rate determining step (k 2 ) and this was confirmed based upon the steady state approximation.
a suitable concentration range and wavelength identified, the following procedure was employed.

Figure 7 .
Figure 7.The UV spectra of the reaction between 1, 2b and 3 with 3×10 -3 M concentration of each compound as reaction proceeds in 1, 4-dioxane with10 mm light-path cell.

Figure 8 .
Figure 8.The experimental absorbance changes (dotted line) against time at 330 nm for the reaction between compounds 1, 2b and 3 at 15ºC.

Figure 9 .
Figure 9. Second order fit curve (full line) accompanied by the original experimental curve (dotted line) for the reaction between compounds 1, 2b and 3 at 330 nm and 15.0ºC in 1,4dioxane.

Figure 10 .Figure 11 .
Figure 10.The experimental absorbance change versus time for the reaction between 2b and 3 in the presence of excess 1 (10 -2 M) at 330 nm and 15.0˚C in 1,4-dioxane.

Figure 12 .
Figure 12.The simplified Scheme for the proposed reaction mechanism.
at 15.0ºC and 330 nm.

Figure 13 .
Figure 13.The experimental absorbance change (dotted line) versus time for the reaction between 1, 2b and 3 at 330 nm and 15.0˚C in 1,4-dioxane.

Figure 14 .
Figure 14.Second order fit curve (full line) accompanied by the original experimental curve (dotted line) for the reaction between 1, 2b and 3 at 330 nm and 15.0˚C in 1,4-dioxane.

Table 2 .
The values of overall second order rate constant for the reaction between 1, 2b and 3 in the presence of solvents such as 1,4-dioxane and 1,2-dichloroethane at all temperatures investigated k 2 .M -1 .min-1

Table 3 .
The values of overall second order rate constant for the reaction between 1, 2a and 3 in the presence of solvents such as, 1,4-dioxane and 1,2-dichloroethane at all temperatures investigated k 2 .M -1 .min-1