Hu Wenxiang^* Peng Qingtao
(Institute of Military Medicine, Headquarters of General Equipment, Beijing 100101 )

Received Jun. 14, 2000; Supported by the National Natural Science Foundation of China.

Abstract The partition coefficients of acidic organophosphorus esters in n-octane/water system were determined spectrophotometrically via the formation of complexes with ferric thiocyanates. The dependence of the molecular weight, steric and polar effects of substituent group of organophosphorus esters on the partition coefficients was discussed. At the same time, the pK_a values of acidic organophosphorus esters were also determined and the substituent effects were studied. The relationship between the concentration of the samples and determined errors was first observed.
Keywords Acidic organophosphorus esters, Partition coefficient in oil-water system, Substituent effects

Partition coefficient in oil-water system is a measurment of hydrophobic and lipophilic interaction. It is important for partition coefficient in oil-water system on the study of extraction equilibrium, velocity of chemical reaction, ion selective electrode, vascular sclerois and so on. So the determination and the concern study of partition coefficient in oil-water system has been still developed continuously^[1-3]. Especially, the combination way between medicine molecular and the acceptor is mainly the non-bonded interaction, hydrophobic and lipophilic interaction is one of the most main non-bonded interaction, therefore it is very important for partition coefficient in oil-water system on the study of medicine design, structure and effect relation. Hansch thought that the partition coefficient of substance in 1-octanol/water system could be decomposed to the sum contribution of parent and the substituent group hydrophobic constant (p). Rokker thought that it could be decomposed to the sum contribution of every structure element fragment constant (f) in the molecular. In n-octane/water system, p or f is decreased with the change of substituent alkyl group from linear chain to fork isomerization in the compound. It has not been reported how to change of that in paraffin-water system. The determination method of partition coefficient of acidic organophosphorus esters in n-octane/water system was studied in this paper, which determined spectrophotometrically via the formation of complexes with ferric thiocyanates.
    The pK_a values of acidic organophosphorus esters is one of the important physicochemical constants of organophosphous compounds. It is important to study the structure effect and correlation analysis for the establishment of the substituent effect parameters. The pK_a values of acidic organophosphorus esters were also determined and the substituent effects were studied in this paper.

1 EXPERIMENT
1.1 Determination of partition coefficient in oil-water system
1.1.1 Reagents
Ferric rhodanate solution 0.4mol/L was newly prepared: 150g Fe(NO₃)₃ ^. 9H₂O and 60g NH₄SCN were dissolved in 800mL water, and the pH was adjusted to 2.9 with saturated ammonium acetate, then diluted to 1000mL. The reagents mentioned above and CCl₄ , H₂SO₄ were all A.R., HClO₄ and n-octane were C.P. Ammonium acetate buffer solution and 1mol/L HClO₄ solution were prepared in this lab.
1.1.2 Apparatus
751 ultra-visible spectrophotometer, pH Meter Type PHs-2 with Orzon combined electrode, Type 232 combined electrode, Type 232 Calomel electrode and Type 231 glass electrode, magnetic stirrer, ZD 58 motor oscillator were used for all measurements.
1.1.3 Experiment method
1.1.3.1 Determination of acidic organophosphorus esters^[1]
A certain amount of sample was weighed to a 25mL volumetric bottle, and diluted to the scale with ammonium acetate buffer solution (pH=2.9). A certain amount was taken into a 60mL separatory funnel after vibration, and ammonium acetate buffer solution (pH=2.9) was added in to make the volume to 10mL. 10mL ferric rhodanate solution was added, mixed, then 6mL CCl₄ was added in too. Then oscillated on the oscillator for 3 min. The organic phase was filtered with Whatman No. IPS filter paper after the phase was separated completely. The absorption of the organic phase was determined at 430nm on the ultra-visible spectrophotometer with CCl₄ as the blank.
1.1.3.2 Determination of distribution ratio
A certain amount of acidic organophosphorus esters was weighed to a 10mL volumetric bottle, diluted to the scale with n-octane. Different amount of samples in n-octane solution was taken to four separatory funnels, and made the volume to 5mL with n-octane. 15(or 30 or 45)mL 1mol/L HClO₄ solution was added in, then the phase ratio was 1:3 (or 1:6 or 1:9). Oscillated the funnels on the oscillator for 1h, separated centrifugally. 10mL water phase that had been distributed was taken in a 25mL beaker, adjusted the solution to pH=2.9 with saturated ammonium acetate and diluted H₂SO₄ solution, which use pH meter with magnetic stirring. Then transfered to 60 mL separatory funnels quantitively. The colorimetric analysis below was same as 1.3.1. The separation ratio D = (C_in -nC_aq)/ C_aq. C_in was the initial concentration of the sample, n was the reciprocal of phase ratio, C_aq was the concentration of the sample in water phase.
1.2 Determination of pK_a
1.2.1 Instrument and reagents
Acid-base 636 titration apparatus, 231 glass electrode and 232 calomel standard electrode, and all reagents were A.R. All acidic organophosphorus compounds were synthesized by us^[4].
1.2.2 Determination of pK_a
0.5mmol acidic organophosphorus ester was precisely weighed to a 60mL beaker and 40mL 75% ethanol solution was added. The beaker was placed onto the acid-base 636 titration apparatus, then titrated the pK_a with agitation by using 0.1mol/L NaOH solution. 231 Glass electrode and 232 calomel standard electrode were installed on the titration apparatus. The content of acid was calculated according to MV/W×100%. M was molar concentration of NaOH solution, V was the consumptive volume of NaOH solution for the end point of titration, and W was 0.5mmol which was the molar quantities of acidic organophosphorus ester. According to the ionization equilibrium of acidic organophosphorus esters: HLH⁺ + L^-

pK_a = -log{[H⁺][L^-] /[HL]} = -log{ MV×MV/[W-MV]}

2 RESULTS AND DISCUSSION
2.1 The determination principle and results of partition coefficient of acidic organophosphorus esters
Generally, the below formula of acidic organophosphorus esters in oil-water system is existed^[1]:
DY = K_d + 2K₂ K_d² C_aq/Y (1)^[1]
D is the distribution ratio,Y =1+ K_a /[H⁺]， K₂ is the dimeric constant of organic phase, K_d is the partition coefficient in oil-water system, K_a is the ionization constant in water phase, C_aq is the total concentration of acidic organophosphorus esters in water phase. We determined log K_d of acidic organophosphorus esters in HClO₄ solution which [H⁺]=1.0 mol/L, Y=1，Then the equation of the two phase distribution in high acidity was obtained:
DY = K_d + 2K₂ K_d² C_aq (2)
According to the equation, the complexation reaction between new prepared ferric rhodanate and acidic organophosphorus esters in water phase had been carried out, then the mixture was extracted to CCl₄ organic phase and determined with ultra-visible spectrophotometer. The concentration of acidic organophosphorus esters in water phase C_aq could be obtained. The regression between D and C_aq was conducted, the intercept was K_d, the slope divided by 2K_d² was K₂. The results were given in Table 1.

Table 1 The substituent group parameters and partition coefficient in oil-water system of acidic organophosphorus esters

Notes: X and Y was respectively the two substituent group of acidic organophosphorus esters.Ss_I was the induced polar effect parameter, SE_S^P was the steric effect parameter that we defined, and the data were quoted from reference 1. ^* was quoted from reference 5, ^** was quoted from reference 6, and the two were all in n-heptane/water system; ^*** was quoted from reference 7, and was in n-octane/water system.

    It should be noted that the solubility of acidic organophosphorus esters in water phase would be decreased with the increase of the number of the carbon atoms of the two substituent group X and Y. Otherwise, according to the photometric analysis theory, the relative error of the determined concentration was small comparatively as the absorbance was during 0.2-0.8. The concentration of acidic organophosphorus esters in water phase should be increased in order to increase the absorbance. When the total number of the carbon atom in the molecular were large comparatively (SC>15), it was difficult for this method to determine the partition coefficient. To avoid the unstability influence of the colour of the molysite, the molysite solution that used all should be newly prepared, and be colorimetric determined in the same time after the molysite had been used.
2.2 The substituent group effects of the partition coefficient in n-octane/water system
The polybasic linear regression was carried out among log K_d (see Table 1), molecular weight M of the compound and the substituent group parameter according to formula (3).
logK_d = c + d·logM + rSs_I + dSE^P_S (3)
The regression coefficient could be obtained c -48.32 d 20.72 r -0.78 d 0.097. The linear correlation coefficient r=0.997, the relative standard error s=0.11. T-test was conducted among every regression coefficient, and t could be obtained respectively: t_c -22.31 t_d 18.52 t_r -1.24 t_d 0.29. From the above results, the molecular weight was the most main factor that influenced the partition coefficient in oil-water system. In fact, the nine log K_d in Table 1 were taken to the correlative analysis with the total carbon atom number in the molecular, formula (4) could be obtained:
logK_d = -5.07 + 0.53 SC (4)
r=0.99， n=9
It had been reported that with regard to (RO)₂P(O)OH, formula (5) and (6) was existed:
logK_d = -5.06 + 0.59 SC (5)
logK_d = -7.93 + 0.70 SC (6)
With regard to R₂P(O)OH， formula (7) was existed:
logK_d = -5.51 + 0.61 SC (7)
When the carbon number in acidic organophosphorus ester molecular >16(SC of carboxylic ester >14), logK_d kept stable or decreased. It might be concerned with the cluster of long chain molecular^[8]. logK_d would be increased with the increase of the oxygen atoms number which neighboured to the phosohrous atom in the molecular^[9]:
R₂P(O)OH R(RO)P(O)OH (RO)₂P(O)OH
    It revealed that logK_d would be increased about 0.6 unit with a CH₂ unit increased in the molecular; and would be increased about 0.3 unit with an oxygen atom increased. The alkyl group changed from normal to isomer(a , b isomer), logK_d was decreased a little in Hansch^'s 1-octanol/water system; while logK_d was increased a little in n-octane/water system which we determined. Otherwise, during the course of determining the partition coefficient in oil-water system, the dimeric constant K₂ of acidic organophosphorus esters in n-octane was obtained, and K₂ was influenced by molecular weight (M) and the substituent group effect.
logK₂ = c + d^.logM + rSs_I + dSE^P_S (8)
And every regression coefficient was c 35.8 d -12.5  1.8  -1.0 r=0.97, s=0.26, n=10. T-test was conducted among every regression coefficient, and t was respectively t_c 7.26 t_d -4.92 t_r 1.25 t_d -1.32. The relative standard error of every regression coefficient was respectively S_c 4.9 S_d 2.5 S_r 1.4 S_d 0.8.
    From the results, we can see that the error between the calculated value and the determined value was comparatively small. It also revealed from Table 1 that K₂ was big in apolar (or low polar) solvents. It demonstrated that acidic organophosphorus esters was existed as dimeric type in apolar(or low polar) solvents^[10].
2.3 Determination deviation of pK_a of acidic organophosphorus esters

Table 2 pK_a values and substituent parameters of acidic organophosphorus esters (X)(Y)P(O)OH

Note: X and Y was two substituent group of acidic organophosphorus esters respectively. DpK_a = pK_a(Found）- pK_a(Calcd.）

    The determination results of pK_a of acidic organophosphorus esters are given in Table 2. The molar quantity of samples must be controlled to near equally during the course of determination. We observed first that the bigger the consumptive weight of the same sample, the smaller the pK_a was. In order to prove the truth of this phenomenon, we used 0.101mol/L NaOH solution titrated the standard sample benzoic acid in 40mL 75% ethanol. The determination results are given in Table 3.

Table 3 pK_a of benzoic acid

    This was consistant with the phenomenon we observed. The data in Table 2 were monobasic linear regressed and the better linear relation could be obtained between pK_a and the square root of sample concentration:
pK_a = 6.71- 2.95 (9)
And the correlation coefficient r was -0.9897( n=5). It might be explained by the ionic activity coefficient.
    Generally speaking, ionization equilibrium constant pK_a^a expressed by activity does not change with the concentration, and pK_a what we determined by 636 automaticall titration apparatus was pK_a^c that expressed by concentration. So the equation (10) was existed:
pK_a^a = log { a_HL / a_Ha_L } = log {r_HL．[HL] / r_Hr_L ．[H][L]}= log { r_HL / r_H r_L } + log {[HL]/ [H][L]} = pK_a^c - log (r_H r_L) (10)
In which, the acitivity coefficient of neutral molecule HL r_HL=1 .
    According to the Debye-Hückel theory of electrolyte dilute solution, we have the following formula:
(11)
In which, r_i was average activity coefficient of ion, I was ionic strength I= 1/2 Sm_iZ_i², which was controlled by the consumptive quantity of strong base. Although ionic strength I was changenable during the course of titration, its average value could be obtained. Ionic strength I was consistant with the initial concentration near the end point of the titration, so the relation of (11) was established. Taken formula (11) into formula (10) and formula (12) was obtained:
pK_a^c = pK_a^a - a (12)
That formula (12) we obtained was consistant with formula (9) obtained by experiment. It revealed that Debye-Hückel theory also applied to ethanol-water solution system in a certain extent.
pK_a changed with the consumptive weight of sample, the determination absolute error of pK_a was about ±0.05 if the consumptive weight of sample was controlled among 0.4-0.6mmol, and was about ±0.025 if that was in 0.1mmol.
2.4 Substituent effects on pK_a
Many scholars put forward the experience formulas of pK_a , the base of these formulas was the contribution of the substituent polar effect which led to the change of pK_a. Actually, the reverse reaction of acid disassociation was hydrated proton combined with acid group anion. Although the direct steric hindrance effect was not clear, because it was important for solvation to the stability of anion so that influenced ionization equilibrium, and thus influenced the values of pK_a . Therefore we put forward the solvated steric effect parameter E_s^p of organophosphorus compounds^[11]. In general condition, pK_a could be expressed as (13) that be influenced by substituent polar and solvated steric effects in the molecule.
pK_a = c +rSs_I + dSE^P_S (13)
Ss_I was induction polar effect parameter. The data in Table 2 were polybasic linear regressed according to formula (13), and the better related result was obtained: linear coefficient r=0.987 and relative standard deviation s=0.089(n=70). Every regression coefficient were c 3.78, r -3.10, d 0.80. Relative standard deviation of every regression coefficient were S_c 0.085, S_r 0.115, S_d 0.033. Put every regression coefficient to t-test and t values could be obtained t_c 44.73, t_r -26.97, t_d 23.81.
    Generally, the error between pK_a which calculated by formula (13) and which determined by experiment was < 0.1. It demonstrated that this formula reflected substituent effcet of pK_a sufficiently. That is to say , electron withdrawing group substitution led to decrease of pK_a and increase of acidity. Increase of substituent group steric effect led to increase of pK_a and decrease of acidity.

REFERENCES
[1] Hu W X. [Thesis for the Doctorate]. Shanghai: Shanghai Institute of Organic Chemistry, Chinese Academy of Science; Beijing Library ,1989:166-170.
[2] Fujita T. Progr. Phys. Org. Chem., 1983, 14: 75-86.
[3] Leo A, Hansch C, Elkins D. Chem. Rev., 1971, 71: 525-537.
[4] Mastryukova T A, Kabachnik M I. J. Org .Chem., 1971, 36 (9): 1201-1216.
[5] Li D Q, Xu F W, Pan C W. Wuhan University Journal, 1983, 2: 74-79.
[6] Li B G, Li J R, Cui L F. Chem. Reagent., 1984, 6 (3): 129-134.
[7] Yuauol B C, Clupugola P A. Paguoxuua, 1970, 12: 47-60.
[8] Jiang X K. Chem. Journal, 1984, 42: 1276-1284.
[9] Kolarik Z. Solvent Extraction Reviews. New York: Marcel Dekker Inc., 1971, 1: 36-38.
[10] Yuan C Y, Ye W Z, Ma H L et al. Chin. Sci (B)., 1982, 25: 7-13.
[11] Yuan C Y, Li S, Hu W X et al. Heteroatom Chem, 1993, 4 (1): 23-31.

[ Back ] [ Home ] [ Up ] [ Next ] Mirror Site in USA  Europe  China  CSTNet ChinaNet