Protonation of alkanolamines and cyclic amines in water at temperatures from 293.15 to 373.15 K

Chemicals

The alkanolamines studied in this work are the 2-aminoethanol, 2,2′-iminodiethanol, 2-[bis(2-hydroxyethyl)amino]ethanol, 2-amino-2-methylpropan-1-ol, and 2,2′-(methylimino)diethanol. Cyclic amines are morpholine, 4-methylmorpholine, pyridine, 1- and 2-methylpiperidine, and 2,6-dimethylpiperidine. Suppliers, purities, CAS numbers, and formulas are given in Table 1. All amines were used without further purification.

Table 1

List of the molecules studied with suppliers and purities.

Name	Supplier	Purity	CAS number	Formula
2-Aminoethanol	Riedel-de Haën	99 %	141-43-5
2,2′-Iminodiethanol	Riedel-de Haën	99 %	111-42-2
2-[bis(2-Hydroxyethyl)amino]ethanol	Riedel-de Haën	99 %	102-71-6
2,2′-(Methylimino)diethanol	Riedel-de Haën	98.5 %	105-59-9
2-Amino-2-methylpropan-1-ol	Fluka, Steinheim, Germany	97 %	124-68-5
Morpholine	Sigma Aldrich, Steinheim, Germany	99.5 %	110-91-8
4-Methylmorpholine	Sigma Aldrich	99.5 %	109-02-4
Pyridine	Sigma Aldrich	99 %	110-86-1
1-Methylpiperidine	Acros Organics, Geel, Belgium	99 %	626-67-5
2-Methylpiperidine	Acros Organics	99 %	109-05-7
2,6-Dimethylpiperidine	TCI Europe, Zwijndrecht, Belgium	>99 %	766-17-6

Aqueous hydrochloric acid (HCl) solutions were prepared from 1 M volumetric solution purchased from Fischer Scientific, Illkirch, France. Water for solutions was distilled and degassed under vacuum in order to obtain free CO₂ solutions.

Buffers solutions used for calibration of the pH electrode were from Riedel-de Haën, Seelze, Germany for pH = 4.00, 7.00, 10.01, 12.00 at 298.15 K, and from Fisher Bioblock Scientific, Illkirch, France for pH = 7.02 and 9.18 at 298.15 K. All the buffer solutions are NIST certified.

For potentiometric measurements, the composition of the amine and HCl solutions are similar to those reported in previous paper [6]. Amine solution is about 0.028 mol·kg^–1 while the composition of HCl solution is closed to 0.100 mol·kg^–1. For calorimetric measurements, the solutions are prepared by mass. Average composition of amine and hydrochloric acid solutions are about 0.200 mol·kg^–1. All solutions were maintained under nitrogen atmosphere to avoid CO₂ absorption during storage.

Potentiometric technique

The experimental procedure previously presented [6] consists in measuring the pH of the amine solution during titration of the amine by HCl following chemical eq. 1. The equilibrium constant of amine protonation K is expressed as function of pH and activities of amine and protonated amine (a_Am and a_AmH+), following eq. 2.

log K is calculated from pH data recorded as function of HCl additions (0.05 mL). The activities are determined by solving a system of equations including charge equilibrium, amine conservation, and water dissociation as described previously [6]; activity coefficients were calculated using Debye–Hückel model. The log K is determined for each addition of HCl, in the domain that precedes the equivalent point; for one experiment, this method leads to calculation of approximately 100 values. The uncertainty on log K, estimated from experimental reproducibility tests, is lower than 0.05 pH units.

Calorimetric technique

The enthalpy of protonation of amines in aqueous solution is determined from the heat power recorded when mixing amine solution with HCl solution. The heat power of mixing is measured using a Calvet-type calorimeter, Setaram BT2.15 model. The calorimeter is equipped with a mixing unit developed in the laboratory for experiments in dynamic mode [3]. The experimental installation is schematically represented in Fig. 1. The amine and HCl solutions flow through a mixing cell located inside the calorimeter where the heat power is detected by a thermopile. The solutions are initially introduced inside two injection loops made of hastelloy tube (1.59 mm outside and 0.80 mm inside diameters), of 30 mL total volume. Each loop is connected to a syringe pump filled with water. The loops and the pumps are thermally regulated at the same temperature (303.15 K). In this way, the amine and HCl solutions are pushed by water to the mixing point at controlled volumic flow rate. This procedure has been chosen to prevent corrosion problems inside the syringe pumps.

Fig. 1

Schematic representation of the flow technique used for heat of mixing determination.

The enthalpy of mixing (Δ_mH) is obtained by dividing the heat power of mixing by the molar flow rate of HCl or of amine, as shown in eq. 3.

Molar flow rates are obtained from volumic flow rate, composition and volumic mass of each solution. The relative uncertainty on the enthalpy is estimated to 5 %.

An example of experimental enthalpy results is presented in Fig. 2. The enthalpy of mixing, expressed per mole of HCl (open symbols) or per mole of amine (full symbols), is represented as function of HCl/amine molar flow rate ratios; this ratio represents the level of amine neutralization. The enthalpy expressed per mole of amine increases up to total neutralization of amine. The enthalpy of mixing per mole of HCl, is observed to be constant, within experimental uncertainty, up to neutralization ratio 0.6–0.8. This plateau and, the slope of Δ_mH per mole of amine vs. HCl/amine ratio, are identical. The Δ_mH value corresponds to the enthalpy of protonation (Δ_pH). The average value of the plateau will be considerate as the enthalpies of protonation Δ_pH at a given temperature. Due to technical problems to control volume flow rates at atmospheric pressure, the measurements were carried out at 0.5 MPa. Experiments were also realized at pressure 1.1 MPa in order to check pressure effect; the results obtained at 0.5 and 1.1 MPa show no significant differences (Fig. 2) in this pressure domain, within experimental uncertainty. Neutralization experiments were performed with identical concentrations for the HCl and amine solutions. Two concentrations, 0.15 and 0.30 mol·kg^–1, were tested and no significant differences were observed between the enthalpies values (Fig. 2). Because of too small heat effects, it was not possible to investigate lower concentrations. According to our pressure and composition effect tests, we assume the measured enthalpies (Δ_pH) to be representative of standard values.

Fig. 2

Enthalpy of mixing at 313.15 K of {water + 2,2′-(methylimino)diethanol} with {water + HCl} expressed per mole of HCl (open symbols) and per mole of 2,2′-(methylimino)diethanol (full symbols). (●,○) p = 0.5 MPa and m = 0.15 mol‧kg^–1, (▲, ∆) p = 1.1 MPa and m = 0.15 mol‧kg^–1, (■, □) p = 0.5 MPa and m = 0.30 mol‧kg^–1, (♦, ◊) p = 1.1 MPa and m = 0.30 mol‧kg^–1.

Equilibrium constant of protonation (K) as function of temperature

The enthalpies of protonation (Δ_pH) were determined at four temperatures between 313.15 and 373.15 K. The experimental values were fitted to linear eq. 4.

The equilibrium constant of protonation ln K was obtained by integration of eq. 5. The constant of integration is given by a value determined by potentiometric technique, usually at ambient conditions.

Validation of the calorimetric method

The procedure was validated with 2,2′-(methylimino)diethanol, for which log K (or pK_a) data as function of temperature are available in literature [6–9]. The values of Perez-Salado Kamps and Maurer [8] and those of Hamborg et al. [9] were obtained at temperature ranging from 278 to 368 K. The equilibrium constants of protonation were obtained by the authors from measurements of electromotive forces in aqueous solutions of amine and hydrochloric acid, using a hydrogen electrode. Oscarson et al. [7] used a method similar to that presented here, based on calorimetric determination of enthalpies of protonation to get temperature dependence of K from 299.9 to 422.1 K.

Our enthalpies of protonation (Δ_pH) were determined at four different temperatures: 313.15, 333.15, 353.15, and 373.15 K. The equilibrium constants of protonation were obtained by integration of the enthalpies vs. temperature, using a K reference value obtained from potentiometric technique [6]. Experimental Δ_pH values are given in Table 2 and represented as function of temperature in Fig. 3. As reported by Oscarson et al. [7], the enthalpy of protonation Δ_pH was observed to increase linearly with temperature. Experimental values were fitted to linear function of temperature, and the parameters of this linear correlation are reported in Table 2. The deviations between fitted and experimental values are smaller than the experimental uncertainty.

Fig. 3

Enthalpy of protonation of alkanolamines vs. temperature. (♦) 2-aminoethanol; (■) 2,2′-iminodiethanol; (□) 2,2′-iminodiethanol from Oscarson et al. [7]; (▲) 2-[bis(2-hydroxyethyl)amino]ethanol; (●) 2,2′-(methylimino)diethanol; (○) 2,2′-(methylimino)diethanol from Oscarson et al. [7]; (▼) 2-amino-2-methylpropan-1-ol.

Our experimental enthalpies of protonation are in good agreement with the values of Oscarson et al. [7] on all ranges of temperatures studied (Fig. 3).

The ln K values at given temperature were calculated using eq. 6, where a and b are the parameters determined from linear regression of enthalpies (eq. 4). The value at 298.15 K was initially determined by potentiometry [7]. The ln K results are reported in Table 3 and represented as function of temperature in Fig. 4, together with reference literature values [7–9].

Fig. 4

Equilibrium constant of protonation of 2,2′-(methylimino)diethanol in water (ln K) vs. temperature; (▲) potentiometric results [6]; (●) calorimetric results; (○) Perez-Salado Kamps et al. [8]; (□) Oscarson et al. [7]; (▽) Hamborg et al. [9].

The equilibrium constants of protonation of 2,2′-(methylimino)diethanol in water (ln K) obtained by calorimetry are in very good agreement with both the value initially determined by potentiometry [6] at temperature up to 353 K and with literature data [7–9]. Our data were correlated to temperature using eq. 7; adjustable parameters A, B, and C are reported in Table 6. The correlation (eq. 7) of ln K values was previously determined from potentiometric measurements up to 353.15 K [6]. The correlation was observed to correctly represent the new data, up to 373 K.

Alkanolamines

After validation with 2,2′-(methylimino)diethanol, the experimental procedure combining potentiometric and calorimetric techniques was applied to 2-aminoethanol, 2,2′-iminodiethanol, 2-[bis(2-hydroxyethyl)amino]ethanol, and 2-amino-2-methylpropan-1-ol. The enthalpies of protonation (Δ_pH) are reported in Table 2 and represented as function of temperature in Fig. 3. As observed for 2,2′-(methylimino)diethanol, the dependency of Δ_pH with temperature is linear. The parameters of the linear regression of experimental data (eq. 4) are reported in Table 2. The enthalpies of protonation of 2,2′-iminodiethanol are in good agreement with the values reported by Oscarson et al. [7]. No literature reference was found for the other alkanolamines.

Equilibrium constants of protonation (ln K) for 2-aminoethanol, 2,2′-iminodiethanol, 2-[bis(2-hydroxyethyl)amino]ethanol, and 2-amino-2-methylpropan-1-ol, calculated using eq. 6, were represented as function of temperature in Fig. 5. The correlations (eq. 7) reported for these amines by Simond et al. [6], which were adjusted from potentiometric data up to 353 K, can still correctly represent ln K values up to 373.15 K. However, this agreement was not observed for 2-amino-2-methylpropan-1-ol. As shown in Fig. 6, the data obtained from values of enthalpy of protonation are not well described by the previous parameters [6] of eq. 7 when increasing temperature. This deviation is due to the value obtained by potentiometry at the highest temperature (343.15 K) that seems to be underestimated. The new parameters for calculated the equilibrium constant of protonation of 2-amino-2-methylpropan-1-ol are given in Table 6. However, the potentiometric and calorimetric techniques lead to consistent values.

Fig. 5

Equilibrium constant of protonation (ln K) of alkanolamines vs. temperature from calorimetric technique (full symbols) and potentiometric results [6] (open symbols). (♦, ◊) 2-aminoethanol; (■, □) 2,2′-iminodiethanol; (▲, Δ) 2-[bis(2-hydroxyethyl)amino]ethanol; (●, ○) 2,2′-(methylimino)diethanol; lines: regression proposed in ref. [6] using only potentiometric data.

Fig. 6

Equilibrium constant of protonation (ln K) of 2-amino-2-methylpropan-1-ol vs. temperature from calorimetric technique (▼) and potentiometric results [6] (▽). - - - : regression from Simond et al. [6]; ___ regression using revised parameters.

Cyclic amines

The constants of protonation (K) have been determined for morpholine, 4-methylmorpholine, and pyridine using both calorimetric and potentiometric methods. The potentiometric experiments were performed up to 343.15 K, and K results are given in Table 4. The enthalpies of protonation (Table 5) were determined from 313.15 to 373.15 K and used to calculate (eq. 6) equilibrium constants of protonation (ln K). Both experimental data for ln K, from potentiometry and calorimetry, are represented in Fig. 7. The experimental ln K data were fitted to eq. 7; the adjustable parameters were reported in Table 6. Literature data of ln K as function of temperature were found, up to 333.15 K, only for morpholine [10]. The comparison (Fig. 7) shows a good agreement with our data. Comparing morpholine solutes, one can assume that the presence of the methyl group increases slightly the value of ln K_a (i.e., –ln K) on the range of temperature studied. The difference is stronger at low temperature.

Fig. 7

Equilibrium constants of protonation (ln K) for cyclic amines vs. temperature; open symbols: potentiometric results; full symbols: calorimetric results; (□, ■): pyridine; (Δ,▲): 4-methylmorpholine; (○, ●): morpholine; (▽): morpholine [10].

The study was extended to substituted demixing piperidines [11]: 1- and 2-methylpiperidine, and 2,6-dimethylpiperidine. In the case of those substituted piperidines, the potentiometric measurements were not possible because of high values of pK_a (close to 11 at 298.15 K). The problem is probably due to the performance of the “Mettler Toledo InLab^® Routine combined pH” electrode used for experiments, which provides nonreproducible results in high pH alkaline conditions. The order of magnitude of the data collected was correct but the uncertainty on the value was too large to be considered. For those three amines, only calorimetric measurements of enthalpy of protonation were performed. The enthalpies of protonation of substituted piperidine are reported in Table 5. The reference value (ln K_{298.15 K}) needed in eq. 6 to calculate ln K as a function of temperature was taken from literature data [12, 13]. These references were found from the “pK_a Data Compiled by R. Williams” and available at Scribd (world’s digital library). The correlation parameters obtained when fitting experimental equilibrium constants of protonation ln K to eq. 7 are given in Table 6. Experimental and fitted ln K values are represented vs. temperature in Fig. 8.

Fig. 8

Equilibrium constant of protonation (ln K) for methylpiperidines vs. temperature. Data obtained from calorimetric technique; (■): 1-methylpiperidine; (▲): 2-methylpiperidine; (●): 2,6-dimethylpiperidine (□, Δ, ○) [12, 13].