The reactions of 4-chloro-2-butanol and 3-chloro-1-butanol with aqueous sodium hydroxide, and 1-chloro-2-propanol and 2-chloro-1-propanol with isopropyl amine

The total reaction of 4-chloro-2-butanol 1 with NaOH(aq) is dominated (74%) by intramolecular substitution (S N i), besides which bimolecular substitution (S N 2, 12%) and 1,4-elimination (i.e. fragmentation, contrary to earlier arguments) exhibit a significant contribution (11%). The total reaction of 3-chloro-1-butanol 2 instead is dominated by 1,4-(72%) and 1,2-elimination (25%), the substitution reactions being just observable (S N i 2% and S N 2 1%). In 1 both the +I-effect and the conformational factors in the intermediate  chloroalkoxy anion favour the S N i-reaction, whereas in 2 the situation is opposite and the location of Cl on a secondary carbon also makes the S N i-reaction less favourable. The relative proportions of 1,4-and 1,2-eliminations for 2 can be explained by thermodynamic basis since the consequent products are more stable than the corresponding products from 1 . 1-chloro-2-propanol 3 and 2-chloro-1-propanol 4 both react with isopropyl amine giving the same product, namely 1-isopropylamino-2-propanol, which indicates that the reaction proceeds through the propylene oxide intermediate. Compound 1 also reacted with isopropyl amine predominantly via S N i-reaction, giving first 2-methyloxetane which then further gave 4-isopropylamino-2-butanol, whereas 2 gave 3-isopropylamino-1-butanol through a direct S N 2-reaction.


Introduction
In alkaline media 1,3-halohydrins can decompose in the following ways: [1][2][3] (1) intramolecular substitution (SNi) leading to the formation of an oxetane; (2) 1,4-elimination (fragmentation) which leads to an alkene and an oxo compound in the consequent cleavage; (3) 1,2-elimination leading to an ,or a ,-unsaturated alcohol; (4) bimolecular substitution (SN2) leading to a 1,3-diol.With isopropyl amine the reaction appears to proceed according to routes (1) or (4). 4 The type of reaction depends on the structure of the substrate and the reaction conditions.1a,2 Decomposition can take place simultaneously by more than one of the above four routes.In the present case the first alternative is the most interesting: 1,2 According to Bartók et al. 1 and Searles et al. 2 in the first step a -chloroalkoxy anion is formed and then the chlorine is replaced by the oxygen atom and an oxetane is formed (route 1).Alternatively, the -chloroalkoxy anion can release the halogen and split into an alkene and oxo compound via 1,4-elimination (route 2).
Only a limited number of studies on the alkaline dehydrochlorination of 1,3-chlorohydrins have been made in rather variable conditions.Bartók and coworkers 1 studied the reactions of several 2-1b and 2,2-disubstituted 1c and 1-1d and 1,1-disubstituted 3-chloro-1-propanols 1e in aqueous 1,4-dioxane (3:7) using Ba(OH)2 as a base.They were obliged to use very high temperatures (even as high as 125 °C).Usually it is recommended that one should follow the studied reactions up till 3 half-lives.However, Bartók et al. 1 stopped even after less than 20-30% conversion.
Searles et al. 2 were mainly interested in the syntheses of oxetanes which they tried to prepare by treating several 2-or 2,2-disubstituted 3-bromo-1-propanols with 15 or 50% aqueous KOH.They state that at either base concentrations, substituents at carbon 2 favour the 1,4-elimination process (fragmentation) over intramolecular substitution.However, their own results prove the opposite, i.e. in 50 % aqueous KOH the latter reaction was favoured over the former by a factor of 1.78.Similarly, Gaylord et al. 5 tried to prepare oxetanes by treating 2-chloro-4-hexanol and 1-phenyl-3chloro-1-butanol with powdered KOH.In the former case they obtained 25% of 2-ethyl-4methyloxetane but in the latter none.
Richardson et al. 3 studied the basic decomposition of 3-chloro-1-propanol and its 2,2-dimethyl derivative as well as that of 4-chloro-2-methyl-1-butanol in 40% aqueous methanol.It is not easy to understand how their work was really done since in their kinetic experiments the initial substrate to NaOH ratio varied from 2.4 to 11.In other words all of base was consumed after the reaction had proceeded even less than 10%.When determining the products the substrate to base ratio was 0.5, 11 and 0.6, respectively.In the case of 3-chloro-1-propanol and 4-chloro-2-methyl-1-butanol they determined the % yields of different products after 28 and 39% conversion but for 2,2-dimethyl-3chloro-1-propanol after 10-20 half-lives.In fact in the latter case the outcome would have been the same after ca 10% conversion since all base was then already consumed.

Reactions of 4-chloro-2-(1) and 3-chloro-1-butanols (2) with aqueous NaOH
The reactions of 1 and 2 were carried out in water at four different base concentrations (0.1, 0.2, 0.4 and 1.00 M NaOH).The base concentration had no systematic effect on the second-order rate coefficients obtained (Table 1).According to Richardson et al. 3 temperature had no remarkable effect on the product ratios either.For determining the activation parameters the reactions were carried out in 0.1 M NaOH(aq) each at 6 different temperatures (Table 2).The activation parameters (Ea, A, H  and S  ; Table 3) were calculated according to the normal procedure.The rate constants determined by Bartók and Bozóki-Bartók 1e at 80 °C for 1 and by Forsberg 6 at 80.07 °C for 1 and 2 are relatively close to those determined in this work (Table 2).However, the activation parameters determined by the former authors are far different from those in this work (Table 4) probably due to the different solvent and relatively mild base.Forsberg, 6 however, determined the decomposition rates for 1 and 2 only at 80.07 °C.The decomposition of 1 was predominated by the SNi reaction leading to 2-methyloxetane (Scheme 1 and Table 3) which can be explained by the most favoured conformation of the chloroalkoxy anion in this case (A, Scheme 2).Opposite to the statements of Richardson et al. 3 and Forsberg 6 that 1,4-elimination can occur only if a substituted alkene is formed, 1 clearly gave ethene (+acetaldehyde, Tables 3 and 4). 2 gave only very small amount of 2-methyloxetane (SNi), the main products being 1-propene (+CH2=O, 1,4-E/fragmentation) and 2-buten-1-ol blanco (1,2-E) besides a small amount of butane-1,3-diol (SN2, Tables 3 and 4) which is again in agreement with its favoured conformation (B, Scheme 2).In general, it can be mentioned that the product analyses have usually been far from complete.The few more comprehensive examples are collected in Table 4.For instance, Bartók and coworkers 1 were also mostly interested in oxetane formation although in some cases they tried to explain also the presence of other products.Table 3. Kinetic parameters for the total (T) and the partial reactions (SNi, 1,4-E, and 1,2-E) in 0.100 N NaOH (aq) and the average distribution of final reaction products at various base concentrations for 4-chloro-2-butanol 1 and 3-chloro-1-butanol 2 1 2 T: A, dm 3 mol 1 s 1 12.2 x 10 11 (1.7 x 10 10 ) a 7.0 x 10 11 T: Ea, kJmol 1 102.9 ± 3.2 (89.6 ± 3.3) a 94.9 ± 1.2 22.9 ± 9.5 (-58.9 ± 9.2) a 27.3 ± 3.6 SNi: A, dm 3 a Recalculated from the rate constants by Bartók et   The reactions of 1,3-chlorohydrins with base fall into the category of parallel second-order reactions 3 , where the overall rate coefficient (kr) equals the sum (ki) of the second-order rate coefficients for the SNi, 1,4-E (fragmentation), SN2 and 1,2-E reactions, i.e. ki = kr(% yield of i/100).
Taking this into account together with the fact that according to Richardson et al. 3 the percentage yield of various products does not seem to depend much on temperature we were able to evaluate the approximate activation parameters for the partial main reactions of 1 and 2 (Table 3).Scheme 1. Different routes for alkaline dehydrochlorination of 1 and 2.

Scheme 2. Conformational illustration of -chloroalkoxy anion,
When inspecting the activation parameters shown in Table 3 it appears that the faster dehydrochlorination of 2 is mainly due to the 8 kJ mol 1 lower activation enthalpy.The activation parameters (Ea and H  ) for the partial reactions (Table 3) of 1 or 2 do not differ very much from those for the total reaction except the A and entropy terms.When comparing our activation parameters (Ea, H  and S  ) for the partial reactions of 1 they are very close to those determined by Richardson et al. 3 for the same type of partial reactions of 3-chloro-1-propanol and 4-chloro-2methyl-2-butanol despite the fact that the latter reactions were carried out in 40% aqueous methanol.
As to the reactions of chlorohydrins in mere water or in acid solution it has been shown especially with 1,2-chlorohydrins that practically no reaction occurs or at least they are extremely slow. 7© ARKAT-USA, Inc.
Both the reaction of 1-chloro-2-propanol 3 and that of 2-chloro-1-propanol 4 with isopropyl amine gave principally the same product, namely 1-isopropylamino-2-propanol 8 the rate ratio being 7:1, 9 the first-order rate coefficients being (9.8 ± 0.7)x10 6 and (1.3 ± 0.1)x10 6 for 3 and 4, respectively.Since the reaction of propylene oxide with isopropyl amine 6,8 gave the same end product we could conclude that the reaction of 3 and 4 first gives the epoxide with SNi-mechanism, which then reacts further with another molecule of isopropyl amine with SN2-mechanism leading to compound 8 in agreement with the results mentioned above.Literature results indicate that even conformational factors favour the reaction proceeding via the epoxide. 4,10

Reactions of 4-chloro-2-(1) and 3-chloro-1-butanols (2) with isopropylamine
In relation to the reactions of chlorides 1 and 2 with NaOH (aq) it was also interesting to study their reactions with isopropyl amine since based on their reactions with NaOH(aq) one could expect that they would give different products as indeed was the case (see experimental).As to 1 the reaction can proceed first with SNi-mechanism leading to 2-methyloxetane which reacts then with another molecule of isopropylamine leading to the product.9a Alternatively 1 can react directly to 4isopropylamino-2-butanol 9 by SN2-mechanism (Scheme 3) which is however quite unlikely.In the case of 2 the product analyses proved 9a that it gave 3-isopropylamino-1-butanol 10, with a direct SN2-mechanism (Scheme 3).Obviously the secondary chloride reacts much slower since 1 reacts 80-times faster than 2 with isopropyl amine.The first-order rate constant was (45.0 ± 1.3) x 10 6 s 1 for 1 and (0.55 ± 0.02) x 10 6 s 1 for 2, the rate ratio being 80:1.When comparing the reactions of 3 and 4 with those of 1 and 2 with isopropyl amine 3 reacts ca 5 times slower than 1 whereas 4 reacts ca. 2 times faster than 2. © ARKAT-USA, Inc.

Kinetic measurements
A weighed amount of 4-chloro-2-1 and 3-chloro-1-butanols 2 (+ internal standard n-amyl alcohol if used) and the base (NaOH) solution were thermostated at least 1h in the measuring temperature The base solution was poured quickly with mixing into the reaction vessel containing the substrate, the initial concentration of which was then obtained as co.Thereafter a sample was withdrawn immediately to determine gas chromatographically the peak area (Ao) of the substrate and that (Ao IS ) of the internal standard (if used).The substrate concentrations of samples withdrawn at suitable intervals were obtained based on the gas chromatographic analyses as follows: Method with internal standard: ct = (At/At IS )/(Ao/Ao IS )co Method with external standard a constant amount injected: ct = (At/Ao)co ct is the substrate concentration (mol dm 3 ) at time t and co that when t = 0.At is the peak area (mVs) of the substrate at time t and Ao that at time t = 0.At IS is the peak area (mVs) of the internal standard at time t and Ao IS that at time t = 0.
Second order rate constants kr for the reactions in NaOH(aq) were calculated from the equation  4).The base concentration appears to have no effect on kr (Table 4).The apparent fist-order rate constants for the reactions of 14 with isopropyl amine were easy to determine since the reactions were carried out in over 16fold excess of the latter.The prepared mixtures as stated above were divided in 10-15 tightly closed © ARKAT-USA, Inc.
screw cap test tubes and placed in an oil bath at 80 °C.The samples withdrawn at suitable intervals were analysed with a Perkin-Elmer F11 gas chromatograph equipped with flame ionization detector and two SE-30 columns.The column temperature was raised by 20 °C/min from 50-175 °C.The following rate (kt) equation was applied kt = (1/t)ln [(A1+A2)/A1] where t = time, A1 the peak area of the substrate and A2 the peak area of the product.

Table 1 .
The

Table 5 .
Response factors determined for the initial reactants and reaction products