The Association of Water with Various Mono- and Bifunctional Phosphine Oxides in Carbon Tetrachloride Jerome W. O'Laughlin, John J. Richard, Jerry W. Ferguson,' and Charles V. Banks Institute for Atomic Research and Department of Chemistry, Iowa State University, Ames, Iowa 50010 The extraction of water into carbon tetrachloride solutions of various alkyl phosphine oxides was studied at various phosphine oxide concentrations. Infrared, nuclear magnetic resonance, and vapor pressure lowering data were used in addition to the partition data to elucidate the nature of the water species existing in the organic phase. Tri-n-octyl phosphine oxide, TOPO, appears to extract water to form a monohydrate at TOPO concentrations less than 0.1M. An overall equilibrium constant, KrHzO,of 0.68 f 0.04 was obtained ,from the present data. Methylenebisrdi-nhexylphosphine oxide], MHDPO, also appears to extract water as a monohydrate at very low MHDPO concentrations. At MHDPO concentrations greater than 0.02M, polymeric hydrated species are indicated and only a maximum value of 1.6 for KrEZofor the extraction of water by MHDPO as the monohydrate could be obtained.
IN THE PREVIOUS PUBLICATIONS from this laboratory (1-4) it was shown that several bifunctional phosphine oxides, particularly methylenebis[dialkylphosphineoxides], are powerful solvent extractants for many metal salts. Although these the solubility of extractants are quite insoluble in water (3, water in solutions of these extractants in inert solvents such as carbon tetrachloride is appreciable (6). The extraction of water can be viewed as a competing reaction in the extraction of metal salts or mineral acids from aqueous solutions with alkyl phosphine oxides and related extractants. Consequently, a somewhat more rigorous study of the extraction of water into solutions of methylenebis[di-n-hexylphosphine oxide], MHDPO, in carbon tetrachloride than previously reported (6) was undertaken. The extraction of water by tri-n-octyl phosphine oxide, TOPO, which had been previously investigated by Conocchioli, Tocher, and Diamond (7) was also studied so that the data obtained on these two extractants could be compared with the assurance that it was obtained under identical conditions. Infrared and nuclear magnetic reasonance spectrometric and vapor-pressure lowering data were obtained in an effort Present address, Dow Chemical Co., Rocky Flats Division, Golden, Colo. (1) J. J. Richard, K. E. Burke, J. W. O'Laughlin, and C. V. Banks, J. Am. Chem. Soc., 83,1722(1961). (2) J. E. Mrochek, J. W. O'Laughlin, H. Sakurai, and C. V. Banks, J. Inorg. Nucl. Chem., 25, 955 (1963). (3) J. W. O'Laughlin and C. V. Banks, ANAL. CHEM., 36, 1222 ( 1964). (4) J. E. Mrochek and C. V. Banks, J . Inorg. Nucl. Chem., 27, 589 (1965). ( 5 ) J. W. O'Laughlin, F. W. Sealock, and C. V. Banks, ANAL. CHEM.,36, 224 (1964). (6) J. E. Mrochek, J. J. Richard, and C. V. Banks, J. Inorg. Nucl. Chem., 27, 625 (1965). (7) T. J. Conocchioli, M. I. Tocher, and R. M. Diamond, J. Phys. Chem., 69, 1106 (1965).
146
ANALYTICAL CHEMISTRY
to elucidate the nature of the water species present in the organic phase. In addition to the data obtained for MHDPO and TOPO, some data on various other bifunctional alkyl phosphine oxides also are reported in the present paper. EXPERIMENTAL
The bifunctional phosphine oxides used were synthesized and purified by previously published procedures (1, 8-10>. TOPO, an Eastman organic chemical, was further purified by a procedure suggested by Brown et al. (11). The names, structural formulas, and abbreviations for the compounds studied in the present paper are given in Table I. The carbon tetrachloride was a Fisher spectroanalyzed reagent. Infrared data over 3000 cm-1 were obtained on a Beckman IR-7 double-beam infrared spectrophotometer. Matched, fixed-thickness (0.1-mm) cells with sodium chloride windows were used. The spectra were obtained with a cell containing the diluent alone in the reference beam. The cell compartment temperature was approximately 35 O C. Near infrared data were obtained using a Cary Model 14 spectrophotometer. Matched quartz cells (1-cm or 10-cm) were used. The cell compartment temperature was 25' C. Chemical analyses for water were performed by the Karl Fischer method using the dead-stop end point technique. Equal volumes of water and carbon tetrachloride solutions of the phosphine oxides were equilibrated for 10 hours on a mechanical shaker at ambient room temperature (23" ==I 0.5" C). The phases were allowed to separate for at least 12 hours. Aliquots of the organic phase were then taken for water analysis. The proton resonance spectra were obtained on a Varian Associates, Model HR-60, high resolution spectromete rat 40" C. Tetramethylsilane was used as the internal standard. A Thomas isothermal molecular weight apparatus, Model 12, was used to obtain the vapor pressure lowering data. The temperature of the compartment was thermostated at 30" C. Azobenzene was used as a standard. RESULTS AND DISCUSSION
Partition Data. The amount of water present in solutions of TOPO, MHDPO, and MEHDPO in carbon tetrachloride which have been equilibrated with water increases with the phosphine oxide concentration as shown in Table 11. The following general equation can be written which expresses the extraction of water as the hydrate nY mHzO: mHsOw
+ nYw = nY.mHzO(,)
(1)
(8) J. W. O'Laughlin, "Progress in Nuclear Energy," Series IX, Vol. 6, D. C. Stewart and H. A. Elion, Eds., Pergamon Press, New York, 1966, Chap. 2. (9) J. E. Mrochek and C. V. Banks, U. S. At. Energy Comm. Rept. IS-827 (1964). (10) J. W. O'Laughlin, J. W. Ferguson, J. J. Richard, and C. V. Banks, J . Chromatog., 24, 376 (1966). (11) K. B. Brown, C. F. Coleman, D. J. Crouse, and A. D. Ryon, U. S. At. Energy Comm. Rept. ORNL-2399 (1957).
Table I. Organophosphorus Compounds Studied Tri-n-octylphosphineoxide Methylenebis[di-n-hexylphosphine oxide] Methylenebis[di-(2-ethylhexyl)phosphine oxide] Methylenebis[di-(3,3-dimethylbutyl)phosphine oxide] Trirnethylenebis[di-n-hexylphosphineoxide] Tetramethylenebis[di-n-hexylphosphineoxide]
TOPO MHDPO MEHDPO MNHDPO PHDPO BHDPO
(CsHid3PO (CBH~~)ZP(O)CHZP(O)(CBH~~)Z (CsHi~)zP(O)CHzP(O)(CsH~~)2 (CBH~~)~P(O)CH~P(OXCIH~~)Z (CBHI~)ZP(O)(CHZ)~P(O)(C~H~ 312 (C~H~~)ZP(O)(CH~)~P(O)(CBH~~)~
where Y represents any of the phosphine oxides, and the subscripts indicate the aqueous or organic phases, respectively. If it is further assumed that the activity of water in the aqueous phase, (H20), is unity and that the equilibrium concentrations of the other species present can be substituted for their activities, the overall equilibrium constant for the formation of the species nY mH20can be expressed as
The total concentration of water in the organic phase, [H201(o.tota~), is then given by [HzOI(o.tata~) = [HDIco)
+ .WmHzO .nYl(o)
(3)
where [H20](,) is the solubility of free water in the organic phase. The total concentration of phosphine oxide, [Y](o,tota~) is likewise given by [Yl(o,tobl)=
rYI(0)
+ Zn[mHsO*nYI(o)
(4)
Substituting Equation 2 in Equations 3 and 4 gives
+ hKm,n’[Yl(o)n rYI(0) + ZnKrn,n’Wl(o)n
[H201(o.tota~) = [HzOI(,) and WI(o.total)
=
(5) I
(6)
[H~Ol(o,tata~) - [HDI(o) = ZmKm,n’krYI(o,total) (7) A log-log plot of the left side of Equation 7 us. [Y](o.tabl) should then be linear with unit slope. Plots of the data in Table I1 for TOPO, MHDPO, and MEHDPO are shown in Figure 1. It was assumed in this section that [H20](,), the solubility of free water in the organic phase, was the same as in pure carbon tetrachloride, Data for 0.1 and 0.2M PHDPO which were reported previously (6) are also shown. The data for MEHDPO fall on a straight line of unit slope. At total TOPO concentrations less than 0.1M, the TOPO data, except for the point for 0.01M TOPO, fall on a straight line of unit slope within the expected experimental error. The difference in the amount of water in pure carbon tetrachloride and in a 0.01M solution of TOPO in carbon tetrachloride is small compared with the total amounts in each solution, and this point is subject to considerable error. Conocchioli, Tocher, and Diamond (7) have studied the extraction of water into solutions of TOPO in carbon tet- [H2O](,)) us. rachloride and reported log ( [H20](o.totn~) log [Y]co.totnl) for concentrations of TOPO less than 0.1M gave a straight line of unit slope. They assumed a monohydrate and reported that the unit slope indicated a monosolvate-i.e., one molecule of TOPO in the water-TOP0 complex. They calculated a value for K(A20,TOPO)’ of 0.56. Based on the same assumptions and excluding the data for 0.01M TOPO and for concentrations of TOPO greater than 0.1M, the average value for KH2~,TOPO’ based on the data in
10-1
10-2
It is evident from Equation 6 that if n is 1, then [Y](o,tobl) is directly proportional to [Y](o)and it follows from Equation 5
[Yl(o,total)
Figure 1. Extraction of water by various alkyl phosphine oxides 0 is TOPO 0
A
is MHDPO
is MEHDPO
X is PHDPO
Table 11. Solubility of Water (Molar) in Solutions of TOPO, MEHDPO, and MHDPO in Carbon Tetrachloride {[HzO](o,totai)- [H20](o)IX 10’ Extractant, M TOPOa MEHDPOb MHDPO? 0.01 0.02 0.05 0.10 0.20 0.50
0.51 f 0.03 0.81 f 0.03 1.96f 0.06 4.16f 0.16 9.07 f 0.07 (37P
0.52 f 0.04 0.91 i 0 06 2.43 f 0.11 4.63 =k 0.03 9.78& 0.33
0.61 rt 0.02 1.18i 0.15 3.39rt 0.22 8.36i 0.35 19.3 i 0.3 65.5 i 0.2
... Average values for two separate equilibrations. * Average values for four separate equilibrations. From Reference 7. a
Table I is 0.68 i 0.04. The slightly higher value for KIr,o’in the present paper over that reported by Conocchioli, Tocher, and Diamond results from the different corrections used for the solubility of water in carbon tetrachloride. The latter authors found the solubility of water in carbon tetrachloride was 0.010M, while a value of 0.008M was found in the present investigation. Healy (12) gives the solubility of water in (12) T.V. Healy, J . Znorg. Nucl. Chem., 19, 328 (1961). VOL. 40, NO. 1, JANUARY 1968
147
~~~~~
z 0.01 0.02 0.03 0.04 0.05 0.06 0.08 0.1 0.2 0.25 0.30 0.35 0.4
~
Table 111. Aggregation Numbers, ii, for Several Alkyl Phosphine Oxides in Carbon Tetrachloride MHDPO PHDPO TOPO~ MEHDPOa MNHDPOa Dry Wet Dry Wet 1.00 0.98 0.98 0.97 0.98 0.98 0.96 0.93
1.09 0.94
0.90 0.86 0.82
1.00 0.99 1.00 1.01 1.05 1.06 1.08
0.96
1.00 1.01 1.04
1.00 1.01 1.04
1.06
1.09
1.01
1.01
.oo
1.01
1.06 1.05 1.06 1.07 1.13 1.16
1.13 1.19 1.21 1.28 1.34 1.43
1
1.10 1.11 1.23
1.11 1.24 1.37
1.17 1.16
1.31 1.28
a The values obtained for & with TOPO, MEHDPO, and MNHDPO were the same when the organic phase was dry or equilibrated with water.
carbon tetrachloride (presumably at 22” C) as 0.0083M, and Greinacher, Luttke, and Mecke (13) report the solubility is 0.0075 at 20” C. The use of the 0.OlM value for the solubility of water with the data obtained in the present investigation would lead to a value of 0.54 for K E 1 ~ . ~ ~ p ~ ’ . At concentrations of TOPO greater than 0.1M and for concentrations of MHDPO greater than 0.02M, the curves rise more steeply than for a slope of one (Figure 1). The data for MHDPO fall on a straight line of slope 1.31 from 0.5M to MHDPO which indicates the formation of species in which n is greater than one, or a breakdown of the assumption that [H20](,, remains constant. The two points for PHDPO also suggest a species with more than one molecule of extractant. Even if [H20](,, remains constant, the values of m in the various possible hydrates are not known and it is not possible to calculate [Y](o, from the [H20](,,t,t,l,data. Log [Y](o)must be plotted on the abscissa in order to determine n, because when n is greater than one, [Y](o) is not necessarily directly proportional to [Y](o.tota~). The observed slope will then not necessarily give the correct value for the average value of n. The limiting slope at low concentrations of MHDPO approaches one and if the only species are Y and Y . H 2 0 , a maximum ualue for KH20,M ~ = 1.6~can be~calculated. ~ ‘ Vapor-Pressure Lowering Measurements, It might be expected that the extent of aggregation of phosphine oxidewater adducts in carbon tetrachloride could be determined by the vapor-pressure lowering technique. Scibona et al. (14) have recently studied the aggregation of quaternary ammonium salts in carbon tetrachloride by this method. They defined an average aggregation number, 7i, by L
a = -
A
where 2 is the stoichiometric concentration of the solute and A is the apparent concentration. The latter was obtained from the vapor-pressure lowering data. Data for fi, Table 111, for several phosphine oxides in carbon tetrachloride solutions both dry and saturated with water were obtained using essentially the same technique as that used by
(13) E. Greinacher, W. Luttke, and R. Mecke, Z . Elecktrochem., 59, 23 (1955). (14) G . Scibona, S. Basol, P. R., Danesi, and F. Orlandi, J . Inorg. Nircl. Clrem., 28, 1441 (1966).
148
ANALYTICAL CHEMISTRY
Scibona et al. (14). The stoichiometric concentrations were all calculated on the basis of the dry phosphine oxide. It is apparent from the data in Table I11 that equilibration with water has little effect on fi in the case of TOPO, MEHDPO, and MNHDPO, which suggests only species monomeric with respect to these extractants. In the cases of MHDPO and PHDPO the data suggest some aggregation. The average aggregation numbers for the latter two extractants for the water-saturated solutions for concentrations of extractant over 0.1Mare not too different from the slopes of the log-log plots of { [H*O](,,t,~l,-[HzO](,,} us. total concentration of extractant for MHDPO and PHDPO. However, the fact that fi values for water-saturated solutions of TOPO did not increase with concentration of TOPO but actually decreased must be considered. The slope of the log-log curve for the water-solubility data increased substantially over one for concentrations of TOPO greater than 0.1M. This complicates any simple correlation of data obtained by these two techniques. Actually, the assumption that solutions of these extractants behave in an ideal manner (upon which this method of calculating ii is based) is questionable and nonideal behavior might account for the observed changes in A with increasing extractant concentration. The increase in fi with extractant concentration for MHDPO and PHDPO to values considerably larger than 1 does seem significant, however, and together with the water solubility data does suggest polymeric species form with these two extractants. Presumably, the substituted side chains greatly reduce the stability of polymeric species in the case of MEHDPO and MNHDPO. Infrared Spectra. It is well known that the infrared spectrum (particularly the 0-H stretching frequency) of species containing the hydroxyl group changes significantly when hydrogen bonds are formed (15). Consequently, infrared spectral studies might be expected to give much useful information on the nature and extent of hydrogen bonding between water and the basic phosphoryl oxygen in compounds such as the phosphine oxides. The various bands associated with water which are observed in the infrared spectrum of water-saturated carbon tetrachloride have been assigned (13). Mohr, Wilk, and Barrow (16) have investigated the changes that occur in the (15) A. L. McClellan, “The Hydrogen Bond,” G. C. Pimentel, Ed., W. H. Freeman and Co., San Francisco, 1960, Chap. 3. (16) S. C. Mohr, W. D. Wilk, and G. M. Barrow, J . Am. Cliem. SOC., 87, 3048 (1965).
0.8
4
I
0.7
I
0.6
II
!Ii i II
0.5
0.4-
I{1I
m 9 1 a 0.3-
ti
2
II II
0 v)
-
;i
I II PI
-
II II
I
3700
3500
3100
3300
WAVENUMBER, cm-1
Figure 2. Infrared spectra of 0.15M solutions of various alkyl phosphine oxides in carbon tetrachloride
-
0.0-
2.2 2.1
2.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
WAVE L ENGT H
Figure 3. Near infrared spectra of dry and water-saturated solutions of MHDPO in carbon tetrachloride
~-
MHDPO, O.2M
Dry _ _ _ _ - -Water-saturated
1.0-cm light path infrared spectrum of water in carbon tetrachloride on the addition of various organic bases. With relatively strong bases, bands near 3680-3690 cm-1 and 3350-3500 cm-1 are observed at low concentrations of base which were attributed to the free and hydrogen-bonded OH stretch, respectively. The bands for free water may also be observed. At high concentrations of base or water, the band near 3680-3690 cm-1 disappears and a broad band appears in the 3400- to 3500-cm-1 region. For high concentrations of base and low concentrations of water, this presumably is caused by the formation of species such as B . . . H-0-H. . B, where B is the added base, When the water concentration is high, hydrogen bonding involving other water molecules seems likely. Roland and Duyckaerts (17) and Conocchioli, Tocher, and Diamond ( 7 ) have interpreted infrared data in the 3100- to 3700-cm-' region in terms of the hydrate species present. The infrared spectra in the 3100- to 3700-cm-l region for 0.15Msolutions of MHDPO, PHDPO, and BHDPO in carbon tetrachloride saturated with water are shown in Figure 2. These spectra are similar to those reported for tributylphosphine oxide (17) and TOPO (7). The intensity of the broad band centered near 3400 cm-1 and the solubility of water both increase considerably in the order TOPO < MHDPO < PHDPO < BHDPO. The ratios of the concentration of water to the concentration of phosphine oxide for 0.2M solutions of the above phosphine oxides were previously reported as 0.52, 0.96, 2.56, and 2.90, respectively (6). The
frequencies of these bands did not change on dilution but their relative intensities changed appreciably as shown in Table IV (18). These intensity data relate to the peak heights of the bands and the ratios would be larger and increase more rapidly with concentration if the areas under the bands were compared. The near infrared spectra for 0.20M solutions of MHDPO in carbon tetrachloride saturated with water and in carbon tetrachloride dried over molecular sieves are shown in Figure 3. The three fairly intense bands in the solution saturated with water at 1.88, 1.95 and 2.00 microns are assumed to be combination bands (U 6) of the various OH stretch bands in the region from 3100 to 3700 cm-1 and the H-0-H bending mode near 1600 crn-'. The spectrum of a solution prepared by mixing equal volumes of water-saturated and dry carbon tetrachloride and read against a dry carbon tetrachloride blank is shown in Figure 4. Two solutions of MHDPO in carbon tetrachloride, 0.005
(17) G. Roland and G. Duyckaerts, Spectrochim. Acta, 22, 793 (1966).
(18) J. E. Mrochek, Iowa State University, private communication, 1963.
9
-
e
Table IV. Relative Intensities of Water Bands TOPO MHDPO PHDPO BHDPO Dilution I w ~ I ~ M s I 3 3 9 4 / 1 3 6 5 2 Is400lI3ss4 I3r02113ssn Undiluted 1.35 1.44 2.51 4.13 1:l 1 :3 1 :7
1.06
1.27 1.13 1.17
...
...
2.46 1.87 2.12
3.05 2.38 2.11
+
VOL 40, NO. I , JANUARY 1968
149
2.0 1.9 1.8 WAVELENGTH , p
Figure 4. Effect of MHDPO on near infrared spectrum of water in carbon tetrachloride 10-cm light path Bottom spectrum, carbon tetrachloride (0.00411.1 in water) read cs. dry carbon tetrachloride Top spectrum, carbon tetrachloride solution (0.004M in water and 0.46211.1 in MHDPO) read cs. a dry solution 0.462Min MHDPO
.I
.2
.3
.4
.5
[YJ(o,totot)
Figure 5. Absorbances and Karl Fischer data for water-saturated solutions of MHDPO and TOPO in carbon tetrachloride
- - - - - -Karl Fischer data (upper curve, MHDPO, and lower curve, TOPO) Open symbols, MHDPO Solid symbols, TOPO 0 or 0; 1.88-micron band
and 0.05M, respectively, which contained the same amount of water (approximately 0.004M) were read us. dry blanks 0.005 or 0.05M in MHDPO. The spectra for both solutions were essentially the same (not shown) in this region as for the solution which was 0.004M in water alone. When the concentration of MHDPO was increased to 0.462M, however, the water bands in this region changed considerably as seen in Figure 4. The concentration of water was the same for both spectra. The solution 0.004M in water and 0.462M in MHDPO was read against a 0.462M solution of MHDPO dried over molecular sieves. The sharp band at 1.89 microns attributed to free water in the solutions 0.005 or 0.05M in MHDPO essentially disappears and a much more intense band at a slightly lower wavelength appears in the solution 150
ANALYTICAL CHEMISTRY
.,
A or A, 2.00-micron band ci
or
1.95-micron band
0.462M in MHDPO. This presumably is caused by the free OH stretch for water hydrogen bonded via one hydrogen atom to the phosphine oxide. The band at 2.00 microns would then represent the hydrogen-bonded OH stretch. The band at 1.95 microns presumably is caused by water species where both hydrogen atoms are involved in hydrogen bonds. Absorbance data for the 1.88-, 1-95, and 2.00-micron bands for water-saturated solutions of MHDPO and TOPO in carbon tetrachloride at concentrations of extractant from 0.01 to 0.5M are presented in Figure 5. The amount of water in the organic phase is also shown. It should be noted that the cell length was 1 cm in this case and 10 cm for the spectra in Figure 4. The spectra for solutions of TOPO in this region were identical to those for solutions of MHDPO except for
I
I
I
I
I
I
3.I
2,61B
i i
2.5
2. I
SOLVENT - C C14 TEMF! - 4 0 C o
L
1.5
6
t
1 0.1
I 0.2
, 03
REF.-TMS
, ,
0.4
EXTRACTANT
0.5
CONC.,M
, d 0.6
Figure 6. NMR data for water-saturated solutions of MHDPO in carbon tetrachloride
the ratios of the absorbances of the water bands. The absorbance data given in Figure 5 represent the difference in the absorbance of a water-saturated solution of the extractant of a given concentration and the absorbance of a solution of the extractant of the same concentration which was dried over molecular sieves. The absorbance of either the 1.88- or 2.00-micron band should correlate with the concentration of the species Y . . . .H -0-H for either extractant. Plots of the absorbance of either band against concentration of extractant up to 0.2M are nearly linear. At higher concentrations of extractant the absorbance-concentration plots for the band at 1.88 microns for both TOPO and MHDPO curve toward the abscissa (Figure 5 ) . This is consistent with the notion of increased hydrogen bonding in these more concentrated solutions. The absorbance of the band at 1.95 microns increases in an exponential manner with concentration of extractant. This band is presumably caused by water molecules which are doubly hydrogen bonded. These could be bridging water molecules in discrete polymeric hydrates. Alternately, the increase in the absorbance of this band might be attributed to an increase in the solubility of “free” water with concentration of extractant. Although the concentration of water in the organic phase for any given concentration of extractant is approximately twice as great with MHDPO (Figure 5 ) as with TOPO, it is obvious the absorbance of either the 1.88- or 2.00-micron band is not nearly twice as large. On the other hand, the absorbance of the 1.95-micron band increases much more rapidly with extractant concentration in the case of MHDPO. Attempts to correlate quantitatively the absorbance of the bands at 1.88, 1.95, and 2.00 microns in the case of MHDPO in terms of monomeric and polymeric hydrates were not successful. Furthermore, the 1.95-micron band also increases in an exponential manner in the case of TOPO where the vapor-pressure lowering data indicated no aggregation.
Proton magnetic resonance studies (see subsequent section) indicate a rapid chemical exchange of water between ail species present; and attempts to identify positively any species other than perhaps the monohydrate on the basis of the data in the present paper would, at best, be of questionable significance. It seems entirely possible that water loosely held in polymeric hydrates and “free” water in moderately concentrated solutions of these polar extractants would absorb in the same region. NMR Data. The NMR spectrum of a water-saturated solution of MHDPO in carbon tetrachloride has been previously published (6). It was found that the chemical shift for the water resonance band and for the triplet for the unique protons on the methylene bridge shifted with concentration. The data obtained are shown in Figure 6. The data shown for the unique protons on the methylene bridging group were obtained on water-saturated solutions. It was found, however, the data on solutions of carbon tetrachloride in MHDPO dried over molecular sieves fell, within experimental error, on exactly the same curve. The fact that only one resonance band is observed for water indicates rapid exchange of water (on an NMR time scale) between any water species existing in solution. The shift to lower fields with increasing concentration is consistent with the view of an increase in the extent of hydrogen bonding with concentration (19, 20). Similar shifts have been observed with increasing concentration of water in tri-n-butyl phosphate and a rather elaborate model for the structure of hydrates present has been proposed (21). In view of the complex reasons advanced for the usual shift to lower fields on hydrogen bonding (20) (which from a naive point of view are in the wrong direction anyway) only the following qualitative conclusions have been drawn from the data in Figure 6. At low concentrations of MHDPO both free water and water hydrogen-bonded to MHDPO presumably exist in solution. With increasing concentration of MHDPO the ratio of bound to free water would increase rapidly. This is consistent with the initial rapid shift of the water resonance band to lower fields. The continued shift to lower fields at concentrations of MHDPO greater than 0.1 to 0.2M is also consistent with the infrared picture of an increase in the extent of hydrogen bonding in this region. The fact that the chemical shift for the unique methylene protons depended only on the concentration of extractant and was the same in dry or water-saturated solutions is of interest. It indicates this shift is not due to any rearrangement of the MHDPO molecule which might occur in the formation of some species such as a cyclic hydrate, These protons have been shown to be acidic (22) and presumably could hydrogen-bond with the basic phosphoryl oxygen. This might explain the aggregation of MHDPO in dry carbon tetrachloride solutions. RECEIVED for review August 10, 1967. Accepted October 20, 1967. Work performed in the Ames Laboratory of the U. S. Atomic Energy Commission.
(19) W. G. Schneider, “Hydrogen Bonding,” D. Hadzi, Ed., Pergamon Press, New York, 1959, p. 55. (20) J. A. Pople, “Hydrogen Bonding,” D. Hadzi, Ed., Pergamon Press, New York, 1959, p. 71. (21) E. Bullock and D. G. Tuck, Trans. Faraday Soc., 59, 1293 (1963). (22) J. J. Richard and C. V. Banks, J. Org. Chem., 28,123 (1963). VOL. 40, NO. 1, JANUARY 1968
151
