In this paper new solubility and potentiometric measurements are reported to model the behaviour ofmelamine in aqueous NaCl and (CH3)4NCl ionic media at different ionic strengths (0.1 ≤ I/mol L−1 ≤ 3.8)and temperatures (283.15 ≤ T/K ≤ 318.15). For this purpose, some literature data were used together withexperimental data. In NaCl solutions, the solubility of melamine decreases with increasing ionic strengthand increases with increasing temperature (�H = 30.5 kJ mol−1 at I = 0 mol L−1), whereas in (CH3)4NClsolution increases with increasing both ionic strength and temperature. In NaCl, at T = 298.15 K, themelamine solubility is 26.1 and 16.3 mmol L−1 at I = 0.104 and 2.304 mol L−1, respectively; whereas in(CH3)4NCl it is 29.9 and 33.6 at I = 0.111 and 1.058 mol L−1, respectively, at T = 298.15 K. Values at infinitedilution are provided together with solubility values of neutral species calculated at different tem-peratures and ionic strengths. As an example, the solubility of melamine is 0.0271 mmol L−1 in purewater. From solubility data, the Setschenow and the activity coefficients were also determined. Theprotonation constant are reported in condition similar to the solubility measurements, and their depend-ence on temperature shows that the proton binding is exothermic (�H = −26.6 kJ mol−1 at I = 0 mol L−1).The entropic contribution is low (T�S = 2.4 kJ mol−1 at I = 0 mol L−1) and increases with increasing ionicstrength in NaCl, whereas in (CH3)4NCl it remains almost constant. The ionic strength dependence
was modelled by means of the extended Debye–Hückel and the Specific ion Interaction Theory (SIT)and data at infinite dilution are calculated. Finally, comparing the protonation constants in differentionic media, the formation of two weak complexes was noticed between the protonated melaminespecies, AH+ and Cl− and between the tetramethyammonium cation, (CH3)4N+ and the deprotonatedmelamine species (A). At infinite diluition (T = 298.15 K) it was found K(AH+ + Cl−) = 0.45 ± 0.05 L mol−1
and K(A + (CH3)4N+) = 0.63 ± 0.03 L mol−1, in agreement with previous findings for amine ligands.
. Introduction
Melamine (2,4,6-triamino-1,3,5-triazine) (see Fig. 1) is a chem-cal intermediate that in its natural state is a white crystallineowder. It was commercially produced in the 1930s from theommon substance urea, which is distilled to produce melamine.elamine is used for a wide variety of applications, including plas-
ics, adhesives, laminates, paints, permanent-press fabrics, flameetardants, textile finishes. In factories, it is mixed in large vats to
roduce a resin; which is commonly used in the manufacturing ofarticleboard [1]. Melamine is a monomer in the manufacturing oflastic materials (melamine–formaldehyde plastics) used to make
tableware products, suitable for food contact applications becauseof its hardness, heat resistance and general stability. These supe-rior characteristics enable the use of melamine-based tableware forthis purpose; however, repeated use can increase the possibility ofmelamine migration into food [2]. Melamine is also added to plas-tic foams to increase density and durability. Melamine and othertriazine compounds are used as a nitrogen source in slow releaseurea-based fertilizer mixtures. Their accumulation and persistencein the environment are well known [3]. Considering that melamineis almost tasteless and that it is very rich in nitrogen (67% by mass),it is added to foods to inflate the apparent protein levels; this adul-teration caused the death of hundreds of pets and, more recently,
in China the adulteration of milk for infant, caused renal failurebecause the formation of insoluble melamine–cyanurate crystalsin the kidney [4,5]. According to the Environmental ProtectionAgency Toxic Chemical Release Inventory, until 1987 [6], 82 000 kg
elamine were released into the air, 240 000 kg were dischargednto water, and the exposure to melamine in the environment
as considered low, but few quantitative data were available. Fol-owing the incidents of China, the competent authorities (US-FDA,
HO, CONTAM and EFSA,) reevaluated health aspects of melamine,ence a number of risk assessments have been performed, and aecent review of Dorne et al. [7] provided an overview of thesessessments. In the same paper, many toxicity data are reportedor melamine in laboratory animals and in livestock, and a para-raph is devoted to the nephrotoxicity of melamine in humans.s already reported, all such uses may lead to a massive pres-nce of melamine in food and environment, for this reason manytudies on the hazard of melamine were found in the literature8–10] (epidemiological and toxicological studies), whereas littlettention has been focused on its acid–base properties. Among thevailable literature data, many studies regard analytical methodsor the melamine trace determination, whilst few data are presentor solution properties. Jang et al. [11] calculated the acid disso-iation constants in water from density functional theory; Dudley12] in the 1951 potentiometrically determined the dissociationonstants of variously substituted melamines and related triazines,nding pK = 5 for melamine. Morales [13] in his paper described theotentiometric titration of melamine in dimethylsulfoxide.
As already explained above, melamine is important in the envi-onmental, industrial and biological fields, and therefore the studyf a speciation model to better understand the interaction of thisolecule in natural media is fundamental. In particular, in this
aper the solubility and the acid–base properties of melamine inifferent media, NaCl and (CH3)4NCl, were studied at different tem-eratures (283.15–318.15 K) and ionic strengths.
. Materials and methods
.1. Chemicals
All chemicals were purchased from Sigma Aldrich or Fluka andsed without further purification, except for the tetramethylam-onium chloride ((CH3)4NCl), which was purified from methanol
s described by Perrin et al. [14]. Sodium hydroxide, tetram-thylammonium hydroxide and hydrochloric acid solutions wererepared from concentrated standard solution, and they were stan-ardized against potassium hydrogen phthalate (for bases) andodium carbonate (for acid), previously dried in oven at 383.15 Kor 2 h. Sodium chloride (analytical grade) solutions were pre-ared weighing the solid previously dried in oven at 383.15 for
h. The solutions were freshly prepared, using grade A glasswarend twice-distilled water (R ≥ 18 M�).
.2. Apparatus
To avoid systematic errors two different potentiometric appa-atus were used to measure the free hydrogen ion concentration,
a) model 809 MetrohmTitrando, equipped with a combined pHlass electrode (from Methrom 6.032.100); (b) model 713 Metrohmotentiometer connected to a Metrohm 665 automatic burette ando a model 8101 Ross type Orion electrode, coupled with a standard
298.15 0.10–2.60318.15 0.10–2.70
calomel electrode. For both systems the estimated precision was±0.20 mV and ±0.001 mL for e.m.f. and titrant volume readings,respectively. A PC was connected to the apparatus and automatictitrations were performed using the MetrohmTiAMO 1.0 softwareto control titrant delivery and data acquisition.
2.3. Procedure for potentiometric and solubility measurements
The preparation of the solutions for the potentiometric mea-surements consisted in different amounts of melamine, dissolvedin the desired ionic medium, sodium chloride or tetramethylam-monium chloride. The melamine concentration in the experimentsranged between 3 and 15 mmol L−1 in NaCl and between 5 and20 mmol L−1 in (CH3)4NCl (see Table 1). During the titrations, thesolutions were magnetically stirred and N2(g) was bubbled throughthe solution to debar O2(g)and CO2(g) inside. A volume of 25 mLof aqueous solution, containing melamine and the ionic mediumused (NaCl or (CH3)4NCl) at different ionic strengths, was titratedwith sodium hydroxide or tetramethylammonium hydroxide upto pH ∼ 7.5. Before each experiment, independent titrations of HClsolutions with standard sodium hydroxide (or (CH3)4NOH) wereperformed to determine the formal electrode potential in the sameexperimental conditions (temperature and ionic strength) of thesystems under investigation. The free hydrogen ion concentrationscale was used (pH −log[H+]).
Solubility measurements were performed as follows: saturatedsolutions were prepared in thermostatted vessels adding an excessof melamine to NaCl or (CH3)4NCl aqueous solutions at fixed ionicstrength values (0.1–3.0 mol L−1) (see Table 1). The solutions werestirred at fixed temperatures (298.15 or 310.15 K) for 18–24 h. Pre-liminary conductivity tests showed that longer stirring times areunnecessary, and a time of 4–6 h is sufficient. After the stirring, thesolutions were filtered with MFMillipore (MCEmembrane) filters0.45 �m. To minimize the systematic errors, several independentexperiments were carried out for each ionic strength. The titrationson the supernatant were carried out by potentiometry using NaOHor (CH3)4NOH standard as titrant, as previously reported.
2.4. Calculations
All the computer programs used in this work were reviewedelsewhere [15].
1 Equilibria 355 (2013) 104– 113
llnclptcd
3
3
mds
A
S
waA(s(
S
aspsst(ttichtdta(
l
l
w(aaStmas0c
Table 2Experimental solubilities in NaCl and in (CH3)4NCl at different temperatures andionic strengths.
All the parameters of the potentiometric titrations (E◦, Kw, ana-ytical concentration of reagents) were refined using the non lineareast square computer program ESAB2M. The melamine proto-ation constants were determined using both STACO and BSTAComputer programs. The LIANA computer program was used forinear and non linear analysis and to fit different equations. In thisaper, all the data are reported in both molar and molal concentra-ion scales; the relationship for the conversion from molar to molaloncentration scale is reported elsewhere [16]. All experimentalata were corrected for temperature-induced volume changes.
. Results and discussion
.1. Solubility measurements
In previous papers [17,18] the theoretical aspects of solubilityeasurements were already described. Briefly, owing to proton
issociation or association (Eq. (1), melamine is indicated as A forimplicity)0 + H+ = HA+ (1)
The total solubility of melamine can be expressed according to:
T = [A0] + [HA+] (2)
here ST indicates total solubility, A0 and HA+ indicate neutralnd protonated species of melamine, respectively. If we indicate0 as the solubility of the neutral species S0 rearranging the Eq.
2) and considering the protonation constants calculated in theame experimental conditions of solubility measurements (KH
1 )see hereafter), we have:
T = S0 · (1 + KH1 · [H+]) (3)
nd then we can calculate the solubility of neutral species mea-uring the total solubility. The analysis of the data obtained fromotentiometric measurements, with BSTAC computer program,howed that at the solubility pH, deprotonated melamine repre-ents 99.1% of the total concentration, and then we can assumehat log ST = log S0. In Table 2, melamine experimental solubilitiesST ≈ S0) are reported in NaCl and in (CH3)4NCl at different concen-ration and at T = 298.15 K, 310.15 K and 318.15 K. One can observehat the solubility in sodium chloride decreases with increasingonic medium concentration, whereas in tetramethylammoniumhloride, the values slightly increase. In both ionic media, weave, as expected, that the solubility decreases with increasingemperature. In many papers ([17,18] and reference therein) theependence of the solubility of neutral species on the salt concen-ration (for 1:1 supporting electrolytes, I ∼ cMX), has been modelledccording to Long and McDevit [19] using the following equationsSetschenow equation [20]):
og y = logS0
0
S0= kc · cMX (4)
og � = logS0
0
S0= km · mMX (5)
here y and � are the activity coefficients expressed in the molarc) and the molal (m) concentration scales, respectively; S0
0 and S0
re the solubilities of the neutral species at infinite dilution andt different salt concentrations, respectively, and kc and km are theetschenow constant (in both molar and molal concentration scale)hat correlates the solubility with the concentration of the ionic
edium, Eqs. (4) and (5) are valid when the activity coefficients
re equal to the unity (i.e., y0 = �0 = 1) and then when the totalolubility is low (ST < 0.05 mol L−1). If the solubility is higher than.05 mol L−1, one must consider the ligand self-interaction, but thisorrection is not necessary in the case under examination. Our
a ±95%.b log S = log ST ≈ log S0, see text.
experimental data were analyzed with a Setschenow type equation,but taking also into account the temperature variation togetherwith variation of the ionic medium concentration, namely
where � = is the reference temperature (in our case � = 298.15 K),whilst T = 298.15, 310.15 and 318.15 K, 52.23 is 1/(R ln10) inkJ mol−1 and 19.145 is (R ln10) in J−1 mol−1. Applying the aboveequation the solubility of melamine can be determined at differ-ent temperatures and ionic strengths. The Setschenow coefficientswere analyzed considering a non linear variation on the salt con-
centration for NaCl and on temperature:
k(c,m) = k(c,m)∞ + k(c,m)0− k(c,m)∞
(c, m)MX + 1+ AT(c,m) · (T − 298.15) NaCl
(7)
C. Bretti et al. / Fluid Phase Equilibria 355 (2013) 104– 113 107
Table 3Solubility of melamine in pure water and temperature dependence parameters.
(CH3)4NCl), see Table 1. The data analysis allowed us to deter-mine the protonation constant values at different temperaturesand ionic strengths, and the experimental values are reported in
0 1 2 3 4 5
-0.2
0.0
0.2
0.4
0.6
log γ
±95%.b As kc,1, parameter of Eq. (7a).c As km,1, parameter of Eq. (7a).
hereas in (CH3)4NCl we considered a single parameter non linearependence
here the Setschenow coefficient (k(c,m)) was expressed, in termsf the parameters k(c,m)0 and k(c,m)∞ valid for (c, m)MX → 0 and (c,)MX → ∞, respectively, k(c,m)1 is the parameter in (CH3)4NCl at
= 1 mol L−1 and AT(c,m) represents the temperature dependence ofhe Setschenow coefficient. Eqs. (7)–(7a) can be used both for the
olar (c) and the molal (m) concentration scales. The solubility ofeutral species of melamine at different concentrations of ionicedium was fitted with Eqs. (6)–(7a) and the results are tabulated
n Table 3. The smoothed values are reported in Tables 4 and 5 forhe molar and molal concentration scales, respectively. Although
elamine is a dangerous chemical for environment, few data arenown about fundamental properties, such as the salt effect andhe temperature dependence on aqueous solubility. Some of theseata in pure water at different temperatures and determined withifferent methods are reported by Yalkowsky and He [21], and areabulated in Table 6. Literature data together with our experimentalata in pure water were studied with the Eq. (6), without the termc,m · (c, m)MX, and the results are: log S0
moothed values of solubility are reported in Table 7 both inhe molar and in the molal concentration scales. From the anal-sis of the data in Table 2 with Eqs. (6)–(7a) it was possibleo determine the Setschenow coefficient. The value in NaCl ism = (0.123 + (0.024 − 0.123))/(mMX + 1) + 0.0011 · (T − 298.15),hereas the value in (CH3)4NCl is
m = (−0.282)/(mMX + 1) + 0.0024 · (T − 298.15). The calculationf the Setschenow coefficients allows us to determine the activityoefficients of the neutral species at different temperatures andalt concentration. In Table 8, the activity coefficients of theeutral species are reported at T = 298.15 and 310.15 K, for NaClnd (CH3)4NCl. In Fig. 2, it can be observed that the trend of theelamine activity coefficients in NaCl is opposite than that in
CH3)4NCl, in fact the log � values increase with increasing mMXn NaCl, whereas decrease with increasing mMX in (CH3)4NCl.his behaviour was already expounded by Sugunan et al. [22], theuthors considered that small highly charged ions confers rigidityo the water structure, obstructing the entry of solute moleculend causing salting out (decreasing of solubility with increasingf the concentration solute). On the contrary, according to theheories reported by Bockris et al. [23] and Long–McDevit [24]
etramethylammonium chloride produces salting in (increasing ofolubility with increasing solute concentration), attributing thisehaviour to the capacity of the ion to disperse forces between
ons and neutral solute and solvent molecules.
a ±95%.b log S = log ST ≈ log S0, see text.
3.2. Protonation constants
The protonation constant of melamine were determined poten-tiometrically in NaCl and in (CH3)4NCl. The measurements wereperformed at different ionic strengths (0.106 ≤ I/mol L−1 ≤ 3.366in NaCl and 0.155 ≤ I/mol L−1 ≤ 2.755 in (CH3)4NCl) and tempera-tures (283.15 ≤ T/K ≤ 318.15 in NaCl and 298.15 ≤ T/K ≤ 318.15 in
Table 7Literature data together with our experimental data in pure water were smoothedand are reported both in the molar and in the molal concentration scales.
Table 8Activity coefficients calculated by Eqs. (6)–(7a) for melamine in NaCl and (CH3)4NCl,at T = 298.15 and 310.15 K in the molal concentration scale.
ables 9 and 10 for NaCl and (CH3)4NCl, respectively. The pro-onation constant values linearly increase with increasing ionictrength and decrease with increasing temperature. In Fig. 3, the
onic strength dependence of the melamine protonation constantn NaCl is also reported at different temperatures. The protona-ion constants in NaCl solutions are higher than the correspondingalues in (CH3)4NCl solutions, due to the interaction between the
able 6iterature aqueous solubility reported in Yalkowsky and He [21].
eprotonated melamine and the cation (CH3)4N+ of the suppor-ing electrolyte [25,26]. This interaction does not occur with Na+
ation. For this reason, the difference between the protonation con-tant in NaCl and in (CH3)4NCl is higher at high ionic strengthalues than at low ionic strength values. At the same time, aeak interaction was noticed between the protonated melamine
pecies HA+ and the anion of the supporting electrolytes, Cl−
n both cases. To our knowledge, few papers reporting protona-ion constant values of melamine have been published [12,13]. Inig. 4, the ionic strength dependence of the melamine protona-ion constant in different ionic media is reported at T = 310.15 K.
able 10xperimental values of protonation constants in the molar concentration scale inCH3)4NCl at different temperatures and ionic strengths.
The ionic strength dependence was modelled by means of theextended Debye–Hückel and the Specific ion Interaction Theory(SIT) approaches [27–30]. The general equation used to fit the pro-tonation constant values in NaCl and (CH3)4NCl at different ionicstrengths with respect to the ionic strength is reported in Eq. (8).
log KH� = log KH0
� − A · z∗√
I
1 + 1.5 · √I
+ L(I) (8)
where log KH0�
is the infinite dilution value, � is thereference temperature 298.15 K, z∗ =
∑z2
react −∑
z2prod,
A = 0.51 + (0.856 · (T − 298.15) + 0.00385 · (T − 298.15)2)/1000 inaqueous solutions and L(I) is a linear function of ionic strengththat can be formulated in different ways. The simplest expressionfor this linear term is L(I) = C·I, extended Debye–Hückel equation,where C is a semi-empirical parameter. If we use the molal con-centration scale for both ionic strength and protonation constantvalues, Eq. (8) becomes the Specific ion Interaction Theory (SIT)equations, and L(I) = �ε·I, where ε is the specific interactioncoefficient. In general, the protonation constant of melamine canbe expressed as a function of activity coefficients as follows
where km is the Setschenow constant [20]. C, ε and km parame-ters can be true constants or vary with ionic strength, generallywhen I < 3 mol L−1, and when the supporting electrolyte is a 1:1alkali metal salt. Recently [31] we proposed Eq. (11).
P = p∞ + p0 − p∞I + 1
(11)
where P can be C, ε and km, and p0 ∼ p at I → 0 and p∞ ∼ p at I → ∞.Note that Eq. (12) is similar to Eq. (6), substituting log S0 withlogKH0, log S with log KH and kc,m with C or �ε.
Considering that different potentiometric measurements wereperformed at different temperatures, we also determined the tem-perature dependence parameters of the interaction studied in thiswork. The equation used for this purpose is the well known Clarkeand Glew equation [32]. In the case of melamine, the equation usedfor both ionic strength and temperature dependence is as follows:
log KH� = log KH0
� + �H · F1(T) + �cp · F2(T) + �ε · I (12)
where F1(T) and F2(T) are already described above in the Eqs. (6a)and (6b).
110 C. Bretti et al. / Fluid Phase Equilibria 355 (2013) 104– 113
Table 11Ionic strength and temperature dependence parameters of the melamine protonation constants in NaCl and (CH3)4NCl ionic media by Eq. (12) at T = 298.15 K.
Protonation constants values at infinite dilution are reportedn Table 11, together with the ionic strength and temperatureependence parameters, whereas the estimated protonation con-tant values at different temperatures and ionic strengths areeported in Tables 12 and 13 for NaCl and (CH3)4NCl, respectively.n Table 14, the thermodynamic parameters for the protona-ion of melamine are also reported. Compared with the enthalpicontribution (�H = −26.6 kJ mol−1 at I = 0 mol L−1), the entropicontribution in the protonation reaction is low (T�S = 2.4 kJ mol−1
t I = 0 mol L−1) and increases with increasing ionic strength in
aCl, whereas in (CH3)4NCl it remains almost constant, there-
ore the entropic contribution is more important in NaCl than inCH3)4NCl.
able 12moothed values of protonation constants in both molar and molal concentrationcales in NaCl.
a In the molar concentration scale.b In the molal concentration scale.c ±95%.
3.3. Weak interaction with the ionic media
The protonation constants of amines, measured in aqueous solu-tions containing a supporting electrolyte, must be considered, inmost cases, as conditional constants because, in conjunction withthe protonation, we must take into consideration also the interac-tions with the anionic and cationic components of the supportingelectrolyte. As an example, the measurements performed in NaClare affected only by the interaction of chloride with the protonatedamine species, as the sodium cation does not interact significantly
with amines. On the contrary, in the case of (CH3)4NCl, we havesignificant interaction of the unprotonated amine (A) with thetetraalkylammonium cation [25,26]. If we take into account the
Table 13Smoothed values of protonation constants in both molar and molal concentrationscales in (CH3)4NCl.
he simultaneous analysis of conditional protonation constants inifferent ionic media at different ionic strengths allows us to obtainhe formation constants for the weak species by minimizing thequares sum of the differences p̄ − p̄∗ [33].
2
(p̄ − p̄∗) (22)
In this analysis the protonation data in NaCl and in (CH3)4NClere included in the ionic strength range 0 < I/mol L−1 ≤ 1.0 at
3 4 3 4
a ±95%.b kJ mol−1.
298.15 K. According to the model which takes into account theformation of weak complexes, the ionic strength dependence of ageneric equilibrium constants is calculated using the equation [33]
log K� = log K0� − z∗ ·
( √I
2 + 3 · √I
− 0.1 ·√
I3
)+ C · I (23)
in several investigations dealing with the formation of weak com-plexes we selected the range 0 < I/mol L−1 ≤ (1–1.5), consideringthat at I/mol L−1 > 1, activity effects can be comparable with weakinteraction effects. Therefore, we performed these calculations tak-ing into account only data in the above ionic strength range andat T = 298.15 K when more experiments are available. Results aresummarized in Table 15. Moreover, Eq. (21) can be rearranged togive
log KH∗ = log KH + log(1 + KCl · [Cl−])
− log(1 + K (CH3)4N · [(CH3)4N+]) (24)
and the parameters of Eq. (24) can be refined using a simple non-linear least squares computer program. In this case we consideredaltogether the protonation data in the two ionic media at differ-ent temperatures. The dependence on temperature and on ionicstrength was taken into account by the Eqs. (23) and (24), respec-tively. Also the results of these calculations are summarized inTable 15. It is noticeable that both log KCl and log K (CH3)4N togetherwith their dependence on ionic strength calculated by the two dif-ferent methods are quite comparable. This, in turn, implies thatthe model that accounts for the formation of weak species isvalid in the whole experimental ranges for temperature and ionicstrength.
4. Discussion
Solubility of melamine was determined in different ionic media,ionic strengths and temperatures. As a general trend, the solubilityincreases with increasing temperature, and the values in (CH3)4NClare systematically higher than the values in NaCl. The backgroundsalt concentration has a different effect on the melamine solubility,in fact in NaCl a “salting out” effect is observed and the solubil-ity decreases markedly with increasing NaCl concentration at bothtemperatures (298.15 and 310.15), whereas in (CH3)4NCl thereis a different trend. At T = 298.15 K, we observe “salting in” anda slight solubility increasing with increasing ionic medium con-centration and a slight decrease of the solubility at T = 318.15 (seeTable 2).
Considering that the models used for the determination of theionic strength and temperature dependence of the melamine solu-bility are different for NaCl and (CH3)4NCl aqueous solutions, it is
difficult to compare the results between these two ionic media, inparticular as regards the Setschenow contants k(c, m). In any case,in NaCl we obtained positive values for Setschenow coefficients(Table 3), which produces an increase of the activity coefficient of
1 Equili
nvssc[sP
dNas(flCii[awls
eeaKewra
e−
sti(aoottpm
enb
5
1
[
[
[[[
[
[
[
[
[[[
[[
[[
[
[[
[[[[
12 C. Bretti et al. / Fluid Phase
eutral species (Fig. 2), whereas in (CH3)4NCl we obtained negativealues and a decreasing of the activity coefficient of the neutralpecies with increasing the ionic strength. Numerical compari-on with Setschenow coefficient of other cyclic amines revealedomparable values for the k(c,m) values in NaCl. In Bretti et al.18] positive Setschenow coefficient values were reported andimilar log � behaviour for neutral species of 2,2′-Bipyridyl, 1,10-henanthroline and 2,2′,6′,2′′-Terpyridine in NaCl.
As shown in Tables 9 and 10, the protonation constantsetermined in (CH3)4NCl are lower than those obtained inaCl, the differences in log KH can be explained by taking intoccount the formation of weak species. In this paper we con-idered the interactions of Cl− with protonated amines and ofCH3)4N+ with unprotonated amine. The possibility of complexormation between tetraalkylammonium cations and N-donorigands was already found in previous papers; for exampleapone et al. [34] studied the behaviour of pyridine in various
onic media and interpreted the differences between the variousonic media in terms of formation of weak complexes, such as(CH3)4N(py)]+ or [(C2H5)4N(py)]+. In the same paper the associ-tion of Cl− with protonated pyridine has been studied. In thatork they found that the association of chloride with [H(py)Cl0]
ead to an increasing effect for the apparent protonation con-tants.
In another paper, De Stefano et al. [25] reported similar consid-rations for three linear unsubstituted polyamines. They found anmpirical relationship for the interaction of Cl− or I− with proton-ted amines and the calculated value for the HACl species is logCl = −0.29; in this paper log KCl = −0.30 ± 0.02 was found. In Casalet al. [35], the authors determined log KCl = −0.2 for polyaminesith general formula C2n−2NnH5n−2. Analogously, Crea et al. [36]
eported log KCl = −0.16 and log KCl = 0.19 for several unsubstitutednd N-alkyl substituted polyamines, respectively.
For (CH3)4N+ complexes with unprotonated amines, De Stefanot al. [25] found log K(CH3)4N = −0.36, compared with log K (CH3)4N =0.2 found in this paper.
In the paper of Bretti et al. [26], the authors have studiedome open-chain polyamines and the differences in the protona-ion constants log KH (NaCl and (CH3)4NCl or (C2H5)4NI) have beennterpreted in terms of formation of weak interactions betweenCH3)4N+ or (C2H5)4N+ and unprotonated or partially proton-ted neglecting interaction with the chloride anion. The valuesf the formation constants of the amines specie with (CH3)4N+
r (C2H5)4N+ cannot be neglected, and for the amines studiedhey found a semiempirical relationship (I = 0 mol L−1) betweenhe number of amino groups and the stability of the weak com-lexes. Appling this relationships, we have logK (CH3)4N = −0.44 forelamine.Negative protonation enthalpy values are in agreement with lit-
rature findings on the amine thermodynamic properties and theumerical values are in a fairly good agreement with those reportedy Bretti et al. [37] for other polyamines.
. Concluding remarks
The findings of this work can be summarized as follows:
. The solubility, the acid base and the thermodynamic propertiesof melamine were studied in NaCl and (CH3)4NCl at differenttemperatures and ionic strengths. The melamine solubility is
generally higher in (CH3)4NCl than in NaCl, and this differenceincreases with increasing both temperature and ionic strength.The solubility enthalpy is endothermic (�H = 30.5 kJ mol−1 atI = 0 mol L−1 and T = 298.15 K).
[
[
[
bria 355 (2013) 104– 113
2. The activity coefficient of the neutral species increases withincreasing NaCl concentration, whereas decreases with increas-ing ionic strength in (CH3)4NCl.
3. As reported for many amines the protonation constant in NaCl ishigher than in (CH3)4NCl, due to the interaction of the (CH3)4N+
cation with deprotonated amine species.4. As reported in the literature for amines, protonation enthalpy
is exothermic (�H = −26.4 kJ mol−1 at I = 0 mol L−1 andT = 298.15 K) and the entropic contribution is quite low.
5. Two weak species were evidenced, AHCl and (CH3)4NA+, andtheir stability is in agreement with previous findings for amineligands.
Acknowledgement
We thank the University of Messina for partial financial support.
References
[1] R. Antikainen, R. Haapanen, S. Rekolainen, J. Clean. Prod. 12 (2004) 919–934.[2] E.L. Bradley, V. Boughtflower, T.L. Smith, D.R. Speck, L. Castle, Food Addit. Con-
tam. 22 (2005) 597–606.[3] S.K. Widmer, R.F. Spalding, J. Environ. Qual. 24 (1995) 445–453.[4] E.Y. Chan, S.M. Griffiths, C.W. Chan, Lancet 372 (2008) 1444–1445.[5] M.S. Filigenzi, B. Puschner, L.S. Aston, R.H. Poppenga, J. Agric. Food Chem. 56
(2008) 7593–7599.[6] IARC, Monographs on the Evaluation of Carcinogenic Risks to Humans, Overall
Evaluations of Carcinogenicity: An Updating of IARC Monographs, World HealthOrganization International Agency for Research on Cancer, Lyon (France),1987.
[7] J.L. Dorne, D.R. Doerge, M. Vandenbroeck, M.J. Fink-Gremmels, W. Mennes, H.K.Knutsen, F. Vernazza, L. Castle, L. Edler, D. Benford, Toxicol. Appl. Pharmacol.270 (2013) 218–219.
11] Y.H. Jang, S.H.S.B. Chang, J. Ku, D.S. Chung, J. Phys. Chem. A 113 (2009)13036–13040.
12] J.R. Dudley, J. Am. Chem. Soc. 73 (1951) 3007–3008.13] R. Morales, Anal. Chem. 70 (1968) 1148–1149.14] D.D. Perrin, W.L.F. Armarego, Purification of Laboratory Chemicals, 3rd ed.,
Pergamon Press, Oxford (UK), 1988.15] C. De Stefano, S. Sammartano, P. Mineo, C. Rigano, Computer tools for the spe-
ciation of natural fluids, in: A. Gianguzza, E. Pelizzetti, S. Sammartano (Eds.),Marine Chemistry – An Environmental Analytical Chemistry Approach, KluwerAcademic Publishers, Amsterdam, 1997, pp. 71–83.
16] F. Crea, C. De Stefano, A. Gianguzza, D. Piazzese, S. Sammartano, Talanta 68(2006) 1102–1112.
17] C. Bretti, R.M. Cigala, F. Crea, C. Foti, S. Sammartano, Fluid Phase Equilib. 263(2008) 43–54.
18] C. Bretti, F. Crea, C. De Stefano, S. Sammartano, Fluid Phase Equilib. 272 (2008)47–52.
19] F.A. Long, W.F. McDevit, Chem. Rev. 51 (1952) 119–169.20] J.Z. Setschenow, Z. Physik. Chem. 4 (1889) 117–125.21] S.H. Yalkowsky, Y. He, Handbook of Aqueous Solubility Data: An Extensive
Compilation of Aqueous Solubility Data for Organic Compounds Extracted fromAquasol Database, CRC Press, Boca Raton, FL, 2003.
22] S. Sugunan, B. Thomas, Ind. J. Chem. 34A (1995) 134–136.23] J.O.M. Bockris, J. Bowler-Reed, J.A. Kitchener, Trans. Faraday Soc. 47 (1951)
184–192.24] F.A. Long, W.F. McDevit, J. Am. Chem. Soc. 74 (1952) 1773–1777.25] C. De Stefano, O. Giuffrè, S. Sammartano, J. Chem. Eng. Data 50 (2005)
1917–1923.26] C. Bretti, F. Crea, O. Giuffrè, S. Sammartano, J. Solution Chem. 37 (2008)
183–201.27] J.N. Brönsted, J. Am. Chem. Soc. 44 (1922) 887–898.28] I. Grenthe, I. Puigdomenech, Modelling in Aquatic Chemistry, OECD-NEA, Paris,
France, 1997.29] E.A. Guggenheim, J.C. Turgeon, Trans. Faraday Soc. 51 (1955) 747–761.30] G. Scatchard, Chem. Rev. 19 (1936) 309–327.31] C. Bretti, C. Foti, S. Sammartano, Chem. Spec. Bioavail. 16 (2004) 105–110.32] E.C.W. Clarke, D.N. Glew, Trans. Faraday Soc. 62 (1966) 539–547.
33] A. De Robertis, C. De Stefano, S. Sammartano, C. Rigano, Talanta 34 (1987)
933–938.34] S. Capone, A. Casale, A. Currò, A. De Robertis, C. De Stefano, S. Sammartano, R.
Scarcella, Ann. Chim. 76 (1986) 441–472.35] A. Casale, A. De Robertis, F. Licastro, Thermochim. Acta 143 (1989) 289–298.
