Hydroxyalkylation of Cyclic Imides with Oxiranes Part I . Kinetics of Reaction in Presence of Triethylamine as Catalyst

Literature describes kinetics of reactions of alcohols, phenols, carboxylic acids, amines and amides with oxiranes such as ethylene oxide and propylene oxide. However, there is no information regarding kinetic of reaction of imides with oxiranes. In this article the kinetics of the reaction of cyclic monoimides: succinimide, phtalimide, and glutarimide, with ethylene and propylene oxides in presence of triethylamine in aprotic solvent was studied. The rate laws for those processes were established based upon on dilatometric measurements. I was said that cyclic monoimides react with oxiranes in presence of triethylamine to give N-(2-hydroxyalkyl)imides as major product. This product react further with oxiranes in consecutive reaction. The kinetics of the reaction of cyclic mono-imides with oxiranes obey the following rate law: V = k1/2 . Based upon kinetic data the following orders of reactivity of imides and oxiranes were obtained: phtalimide ≥ succinimide > glutarimide and ethylene oxide > propylene oxide. The solvent (DMF, DMSO and dioxane) effect was also studied. From temperature dependences the thermodynamic parameters: activation energy, enthalpy and entropy from linear Eyring plots were obtained. 1/2 cat c 3/2 imide c 1/2 oxirane c

In this paper the kinetics and products of the reaction of cyclic monoimides with oxiranes in presence of triethylamine (TEA) as catalyst were studied in order to establish the mechanism of the reaction.These processes can provide the kinetic models mimicking initial steps of the synthesis of polyetherols from oxiranes and azacyclic substrates containing few imide groups, for example isocyanuric acid [29], barbituric acid [30] or 6-amino-uracyl J. LUBCZAK 89 [31].These azacyclic substrates are sparingly soluble in organic solvents and therefore their reactivities cannot be studied in this class of solvents.
The method used for following the progress of the reac 2. Experimental ials rich) were recrystallized from

Kinetic Measurements
placed in 50 cm 3 volucentration on the rate of re re nce was obtained for the syste tions studied in this work deserves a short comment.Our preliminary results indicated that determination of epoxide groups based on alkalimmetric titration of excess hydrochloric acid in dioxane in these systems failed due to the presence of amine catalyst.Therefore we have used the dilatometric method [32] by measuring the volume contraction that occurs with advancing reaction.Our preliminary studies indicated that the contraction of volume was proportional to the consumption of oxirane (c ox ).This was achieved by the kinetic studies of the reaction of imide with variable amount of EO or PO at the level of 0.2 to 1.0 equivalent and determination of the final volume of contraction (V ∞ ) to give a linear V ∞ = f(c ox ).

ydroxymethyl)succinimi
The HMSI was synthesized according col [9].The crude product precipitated from reaction mixture, which was further purified by recrystallization from benzene.Weighed sample of imide was metric flask and dissolved in ca 30 cm 3 DMSO.After dissolving of imide and TEA as catalyst the temperature of the solution was raised to 25˚C, 30˚C, 35˚C, 40˚C, 45˚C or 50˚C (±0.05˚C).Oxiranes (EO or PO) as liquids stored in refrigerator were transferred into the flask, which was immediately closed to avoid evaporation of boiling oxirane and weighed.The concentration of oxirane was calculated based on mass of substrates used.The solutions prepared in that way were then rapidly trans-ferred into thermostated dilatometer.The influence of catalyst con action was studied at constant concentration of imide and EO (1.0 mol/dm 3 ).The TEA concentration was varied within 0.1 -0.5 mol/dm 3 region with 0.1 mol/dm 3 step.The influence of imide concentration on the rate of reaction was studied at fixed concentration of EO (0.25 mol/dm 3 ) and constant concentration of catalyst (0.5 mol/dm 3 ) with variable imide concentration within 1 -3 mol/dm 3 region.The change of rate constant in the course of reaction and order for oxirane, calculated for the system with excess of imide is exemplified in Table 1.
The influence of oxirane concentration on the rate of action was studied varying its concentration within 0.25 -1.00 mol/dm 3 with 0.25 mol/dm 3 step at constant concentration of imide (1.0 mol/dm 3 ) and TEA (0.5 mol/dm 3 ).Additionally, the kinetics was studied at variable concentration of imide (at 0.25 -1.00 mol/dm 3 ) at fixed EO concentration (1.0 mol/dm 3 ).The results are given in Tables 3 and 4.
The temperature depende m with 1.0 mol/dm 3 imide and oxirane and 0.5 mol/dm 3 TEA concentrations within 30˚C -45˚C for the reaction with EO and 35˚C -50˚C for the reaction with PO temperature range with 5˚C step.The temperature range for kinetic measurements is limited from lower side by long reaction time (2 -4 weeks for EO to 2 -3 months for PO), which precludes the good accuracy of measurement, and from the higher side by boiling of oxiranes in reaction mixture (above 40˚ for EO and 50˚ for PO).
DMSO, DMF and dioxane were used as solvents.Due to lume of ca.45 cm 3 w e concentration of oxirane (b) was defined as: low solubility of some imides and catalyst in dioxane, the imide and oxirane concentrations were 0.5 mol/dm 3 and that of TEA was 0.25 mol/dm 3 .
The glass dilatometer of total vo as used equipped with 40 cm long capillary of inner diameter 1 mm.Before filled in it was kept at reaction temperature for at least 15 min.The reaction time was measured from the moment of mixing the reactants.The first reading of meniscus level was usually made after 10 min.needed to fill in the dilatometer and bringing the mixture to reaction temperature.The next readings were made after constantly increasing time intervals until the meniscus level stabilized.The initial level of meniscus was calculated by extrapolating the readouts to the time t = 0. Further reading were made in increasing time intervals until constant level of meniscus in capillary was attained. Relativ where: c 0B , c B -corresponding to initial and instantanebstitution m ng the system with EO as standard, the relativ ous concentrations of oxirane, mol/dm 3 .l -l (t) -the meniscus level at time t and l  and o l are the levels at the end and beginning of expe ent, respectively.
In order to establish the rate low, the su rim ethod was applied.The values of relative concentration of oxirane, b obtained for 0.2 -0.8 region with 0.05 step and reaction time were put into the kinetic law equations of various orders (Table 1) to establish the order of reaction for oxirane.Standard deviation of k (determined 13 -15 times in single kinetic run) and reproducibility (every determination of k was obtained in three independent kinetic runs) were within 6% -7%, in some cases they were 10%.
Consideri e reactivity of PO (r PO ) was used as the ratio of the rate constants at the same temperature: where: k PO, k EO -rate constants for PO and EO, respecte reactivity of imides was analyzed in analogo om [33]: tively.Relativ us manner using succinimide as standard.
The activation energy (ΔG ≠ ) was calculated fr where: R-gas constant [J/mol versus T a straight line ith activation enthalpy and entropy of was obtained w the reaction as the coefficients of the line [33]:

Product Analysis
The products of reactions were isolated after full kinetic led off under reduced pressure.imides were extracted and puri-  In some cases the post-reaction mixtures were analyzed chromatographically (KB 5901 chromatograph, COR-RABID, m, 6 atm pressure, room temperature, steel columns of 100 mm length and 4 mm diameter, silicagel, 60%/40% v/v hexane/dioxane, 1 cm/min eluent flow.The system was calibrated using the imides and appropriate N-(hydroxyalkyl)imides, namely: HESI, HEPI, HPSI, HPPI, HEGI, HPGI as standards.The standards were synthesized from succinic or phthalic anhydrides and ethanoloamine (HESI, HEPI) [3] or SI, PI or GI and appropriate oxiranes or alkylene carbonates according to the procedures used before [4,5], and purified by recrystallization from benzene or ethanol and identified by elemental analysis and NMR spectroscopy.
Some reactions of imides with EO or PO in acetone in presence of TEA were also performed and the products were analyzed in order to isolate and dete ely the products in solvent of different polarity.IR spectra of N-(hydroxyalkyl)imides were recorded in KBr pellets or in CHCl 3 , while the spectra of post-reaction mixtures were taken in films (spectrometer PARA 00 FT IR, Perkin Elmer). 1 H-NMR spectra of (N-hydroxyalkyl)imides were recorded on a spectrometer B-S586A, 80 MHz, TESLA, Czechoslovakia (in d 6 -DMSO or in CHCl 3 ) with HMDS internal standard.

Kinetics
The rate law for the reaction of S by substituting t concentration, b netic equations of various orders (Table 1).In presence of 12-fold excess of imide (as well as lower excess of imide) the following rate law was found: where c B is instantaneous concentration of oxirane, mol/dm 3 .
In columns 3 and the rate constants calculated for 3 (zero order fit) the rate constant constantly de closest orders (zero and first) are shawn for comparison.In column creased, while in column 5 (first order fit) it increases constantly, therefore the best fit was for 1/2 order.The rate constant depends on the catalyst concentration (c k ) according to the formula: the n ≈ 0.5 was obtained (Figure 1 The reaction of monoimide w of : the reaction of im and the reaction of hydroxyalkyl groups w 3 3 ith oxirane is composed two steps ide groups AH with oxirane (reaction 1), ith remaining oxirane: (8) In order to avoid the second process, the reaction was performed at the excess of reactive groups of imide.This imposed .In such a condition the dependence: (9) operates.According to kinetic measurements the power of the imide group concentration was p ≈ 3/2 (Figure 2).Thus the kinetic law describing the reaction of SI with EO is: Other imides (PI, GI) and also HMSI react with EO and PO in similar manner indicating 1/2 order on oxiran k 1/2 " should stay in agreement with the k calculated from the formula (11), in which the constant concentration of imide groups AH is not assumed, but the real momentary concentration co e concentration (Table 2).
If the Equation ( 10) is valid, the is nsidered, according to the formula: where: c k  c 0 -initial concentration of imide or oxirane, mol/dm 3 , x-the decrease of oxirane concentration due to reaction with imide (consumption of oxirane due to side reactions was neglected), where: Integration of (11) led to: This equation can be used only if oxirane reacts with imide groups, i.e. when excess imide is present.The results on rate constants obtained from ( 10) and ( 11) are collected in Table 3. Satisfactory a c greement was obtained when at least 6-fold imide excess rela was applied.
The involvement of consecutive reactions was importance for presented kinetic studies of reaction of imide group with oxirane.The hydroxyl groups of N-(hydr with oxiranes were oils, while N-hydroxyalkyl derivatives of studied imides (SI, PI) are solids of defined melting point.The N-hydroxyalkyl derivatives crystallized from post-reaction mixtures upon long time.The consecutive products are oily compounds.These were isolated and identified in the reaction of imides with oxiranes used in twice excess.From the comparison of the IR spectra of reaction mixtures with those of pure N-(hydroxyalkyl) imides one can see the difference in the intensity of the band centered at ca 1050 cm -1 , which is slightly higher for mixtures.This indicates the formation of ether bonds, i.e. the consecutive products.Moreover, the bands attributed to carbonyl groups (1700 cm -1 ), and hydroxyl critical oxyalkyl)imide was involved in consecutive reaction with oxirane.The latter react as soon as the product of the first step appears.Thus, both reactions compete in further progress of reaction.Therefore the products were analyzed carefully.
1) It has been found the products of reaction of imide p = 1.43 ± 0.04 ted to oxirane oxiranes with imides in presence of TEA catalyst.groups (3400, 1250 and 1100 cm -1 ) were also observed.
2) Chromatographic analysis of oily product mixtures indicated that crude product consists of 75% -80% N-(hydroxyalkyl)imides and 5% -15% unreacted imide.The yields, elemental analysis and IR and 1 H-NMR spectra of products isolated after kinetic measurements are specified in Product analysis section.
3) For comparison the products were also readily isolated from the post-reaction mixtures for the systems with SI or PI with oxiranes in acetone.N-(hydroxyal-kinetic data for 20% -60% progress of the process were kyl)im ethano imides.The analysis of mixtures combined indicated that consecutive prod cts were form with the yield not larger than 5% -10 ght.  2.

Reactivity of Monoimides and Oxiranes
Considering the electronic structure of oxiranes, the reactivity of these towards monoimides in DMSO should depend on the electropositive character of carbon  1).Analogous dependence was observed for the reactions of PI, GI, and HMSI with oxiranes.Those deviations were less than 10%.However, in order to avoid the consecutive reactions which might influence the accuracy of obtained rate constants for N-hydroxyalkylation reaction, only the Higher reactivity of EO than PO was expected, because electrophilicity of C-1 in PO is weakened due to electro-donation from methyl group.Indeed, rate constants (Table 2) for EO with all studied imides were higher than those for PO.
Comparing relative reactivity of imides (Table 2) one r r of imide group reacted with oxiranes.Its reactivity was therefore lower.In order to compare the relative reaclarge nd out than SI (   Activation energy (ΔG ≠ ) of the reactions were obtained from linear Eyring plots obtained from temperature dependence (Figure 3).Mean values of enthalpy and entropy of activation are collected in Table 6.From the values of entropy of activation one can deduce that the transition state in case of the reaction of PI and GI with PO is more ordered, probably because of rotational free-  dom in case of the GI and PO system in comparison with other systems.This will be discussed later in Part II.
The discussed reactions were conducted in DMSO.The same reactions were also performed in DMF and dioxane and the results of kinetic measurements are collected in Table 7.The 1/2 order of concentration of EO and PO were also found in these solvents, which suggested the same mechanism operating in these solvents.The decreasing values of rate constants were found in the order: DMF > DMSO > dioxane.

Summary and Conclusions
1) Cyclic monoimides react with oxiranes in presence of triethylamine to give N-(2-hydroxyalkyl)imides as major product.This product react further with oxiranes in consecutive reaction.
2) The kinetics of the reaction of cyclic monoimides where c K, c AH and c B stay for concentrations of catalyst, imide, and oxirane, respectively.
3) In polar solvents the reactivity of imides and oxiranes decreases in the raw: PI ≥ SI > GI > HMSI and EO > PO.
4) The decreasing values of rate constants were found in the order: DMF > DMSO > dioxane.The highest reaction rate in DMF is presumably consistent with its basic character related to the presence of nitrogen atom, which might play the same catalytic role as TEA, i.e. enabling the proton transfer from imide into oxirane The least pothe reactions of imides with oxiranes.

e [K]
∆G # [kJ/mol] lar dioxane cannot be involved in this step and therefore the rates of reactions were the least.
5) The reaction of N-(hydroxyalkyl)imides with oxiranes was first order for oxirane.In contrary the order of reaction was 1/2 for oxirane and imide both at excess imide and equimolar conditions.If excess of oxirane was used, the order of reaction changed to two for oxirane indicating clearly the change of a mechanism.In such condition the alcohol product prevails and the rate of oxirane consumption is determined by addition of alcohol to oxirane. [

Table 1 . Rate constants for the reaction of SI with EO at 313 K (40˚C) calculated for various kinetic laws.
n = 0.46 ± 0.01 Figure 1.The plot of rate constant versus catalyst concentration for the reaction of SI with EO.