Acidic Properties of Kaolinite in Water and Acetonitrile1

R. H. LOEPPERT, JR., L. W. ZELAZNY, AND B. G. VoLK2

ABSTRACT Additional Index Words: acidity, pH-dependent sites, nonaqueous

iitrstions surfflcc flciditv
Potentiometric titration of weakly acidic sites on soil clays is limited '
in aqueous media by the acidic property of water which prevents ———————————————————
sharp endpoints. Generally, it is not possible to obtain sharp , . , . . .,
endpoints in aqueous systems for acids with PKa > 8. In this stud,, C OIL ACIDITY and aCld-base reaCtIOnS ln SOlls .*« ln"
kaolinite with K, H, or Al as the saturating cation was successfully v3 fenced predominantly by the clay and organic frac-
titrated in water and acetonitrile using the glass-calomel electrode tions and their associated surface reactions. Previous work
system for potentiometric determination. Sharper endpoints and in which titration curves have been used to classify and
wider working potential ranges were obtained in acetonitrile than in quantitate the sources of acidity in pure clay, organic
the aqueous system, liter of base needed to reach the final endpoint matter, and natural soil systems were reviewed by Jackson
was greater in acetonitrile than in water; for example, K-saturated (1963), Volk and Jackson (1964), and Coleman and
clay required a negligible titer in water (0.4 meq/100 g clay) in Thomas (1967). Titrations reported for soil systems have
contrast to acetonitrile, in which a significant titer (7.4 meq/100 g generally been conducted in aqueous media, although
clay) was required. The larger titer in acetonitrile was attributed to (i) experiments have involved the use of nonaqueous
pH-dependent sites for which a quantitative endpoint was not . S.
solvents For
. ... . , ,c , . ,,ni-n\ i
obtained in water due to the acid-base properties of this solvent, and ' example, Wnght and Schmtzer (1960) used
(ii) an increased surface acidity in acetonitrile. A comparable dif- titrations in nonaqueous media to assess the acidic proper-
ference in total liters at the final endpoint was observed for the ties of organic matter, and Mitra and Kapoor (1969) and
titration of H- of Al-saturated kaolinite in water and acetonitrile. Kapoor (1972) titrated acidified nontronite in several
————— nonaqueous solvents.
GSllbe"t!Tsnj±Jhseef0noSC6i2I?ep- °f *" *"** ^ ^ *"" The
amphiprotic character, acid-base properties, and
"Graduate Assistant, presently Assistant Research Scientist, Univ. of dielectric constant of the solvent Can significantly influence
Florida, Gainesville FL 32611 Associate Professor of Agronomy solvent-Solute interactions (Kolthoff, 1974). Solvents Can
Virginia Polytechnic Institute and State Univ., Blacksburg, 24061, and , . .,, , . . . , ^, ... ., .
Associate Professor of Soil Sci., Univ. of Florida, Gainesville, 32611, be classified mtO two broad groups: (l) amphlpTOtlC SOl-
respectively. vents, and (ii) aprotic solvents (Fritz, 1973; Huber, 1967;
 1102 SOIL SCI. SOC. AM. J., VOL. 41, 1977

Kucharsky and Safarik, 1965). Amphiprotic solvents such the appropriate solvent were then titrated. Primary standard
as water, methanol, and acetic acid exhibit varying degrees benzoic acid was weighed directly into the appropriate solvent and
of autodissociation; however, aprotic solvents such as titrated.
pyridine, acetone, benzene acetonitrile, dimethylform-
amide (DMF), and nitromethane show very little or no Preparation of Clays
autodissociation. Kaolinite, ultrafine hydrite grade, was obtained from the
Solvent-solute acid-base interactions are summarized in Georgia Kaolin Company. The <2.0 ^im clay fraction was
the following equations: separated by Na2CO3 dispersion, centrifugation, and decantation.
Aliquots were A1-, H-, and K-saturated with the appropriate
chloride solution. Saturation was accomplished by four 20-min
AH + SH A- + SH,+ [1] mechanical shakings of the clay fraction with 20 symmetries of
appropriate solution. The suspension was centrifuged and the
B + SH BH [2] supernate discarded after each washing.
The clay samples were desalted by successive 10-min washings
where AH, B, and SH represent acidic-solute, basic-solute, with deionized water until the supernate gave a negative test for
and solvent species, respectively. If these reactions proceed Cl~". The clay was subsequently freeze-dried, stored at 0°C, and
equilibrated over P2O5 in a desiccator for 18 hours prior to
to the right to any extent as will occur in solvents with titration.
appreciable acidic character, a sharp titration endpoint will
not be obtained (Huber, 1967; Fritz, 1973). Amphiprotic Titration Procedures
solvents are subject to the protolysis reactions illustrated Titrant was added at a rate of 0.03125 ml/min using an
above, depending on the acidities and basicities of solutes autoburette (Radiometer ABU 13 with TTT50 titrator module).
in the system. Generally, it is not possible to obtain sharp Potentials were determined with a digital pH meter (Radiometer
endpoints in aqueous systems for acids with pKa > 8, such PHM 64) and recorded using a strip chart recorder (Radiometer
REC 61 with RE A 160 titrigraph module). Potentiometric de-
as some phenolic compounds or pH-dependent edge sites terminations were performed with one of the following electrode
on clays. Aprotic solvents, however, have been used to systems: (i) Orion rugged combination electrode, (ii) Orion glass
titrate acids having pKa values much > 8 (Huber, 1967). indicator electrode and Orion sleeve-type calomel reference
Inert aprotic solvents also have much wider working electrode, (iii) Beckman rugged combination electrode with fiber
potential ranges than most amphiprotic solvents (van der junction, and (iv) Radiometer combination electrode with porous
Heijde and Dahmen, 1957). Properly chosen solvents can plug liquid junction. Electrode response was standardized with a
pH 7.00 buffer.
be used to obtain sharp stepwise titration curves for Solvent (40 ml) previously dried over 3A molecular sieves was
mixtures of weak acids (Kucharsky and Safarik, 1965). added to the sample, or sample plus salt, and allowed to
The objectives of this study were to investigate the equilibrate for approximately 20 min prior to continuous titration.
titration of benzoic acid, HC1, and AlCls in selected All titrations were performed under dry N2, and suspensions were
solvents, and H-, A1-, and K-saturated kaolinite in water agitated with a mechanical stirrer during determinations. The
and acetonitrile. Special attention was given to experimen- glass electrode was soaked in water and wiped clean prior to
titration (Fritz, 1973). All titration curves were corrected for the
tal technique required for nonaqueous titrations. acidic liter of the solvent.
The excess salt used in this study was tetramethylammonium
MATERIALS AND METHODS chloride. Salt, previously dried over P2O5, was added to give a
0.1/V solution.
Solvents
RESULTS AND DISCUSSION
The solvents used in this study were deionized water, spectro-
quality methanol, 2-propanol, DMF, pyridine, acetonitrile, and Nonaqueous Titration Techniques
acetone. The presence of acidic impurities was checked by
titrating 20 ml of solvent with 0.0415Q/V tetrabutylammonium Of the four electrode systems originally tested, the Orion
hydroxide. The solvent was considered acceptable if < 0.04 ml of rugged combination electrode was found to be unacceptable
standard base was required to reach the potentiometric endpoint. for use in nonaqueous solvents. Potential readings were
All solvents were dried over 3A molecular sieves prior to use erratic, response was slow, and the capillary junction
(Forman and Hume, 1964). appeared to clog. Fritz (1973) observed a similar phenome-
non with a capillary liquid junction. The other electrode
Preparation of Basic Titrants systems were acceptable, although the fiber junction elec-
Potassium hydroxide titrant was prepared by dissolving 2.80 g trode apparently clogged and response was sometimes slow
of KOH in 100 ml of methanol and diluting to one liter with in the less polar solvents.
benzene. The resulting solution contained 0.041 SON KOH. The For titrations in water, aqueous KC1 was used as the
aqueous titrant was prepared by dissolving 2.80 g of KOH in filling solution for the calomel cell; however, for titrations
deionized H2O to give a solution containing exactly 0.041 SOW
KOH. Tetrabutylammonium hydroxide was prepared in ben- in organic solvents the calomel cell was filled with a
zene-methanol according to the procedure outlined by Cundiff saturated solution of KC1 in methanol (Cundiff and Mar-
and Markunas (1956) and diluted to exactly 0.04150^ with kunas, 1956). With aqueous KC1 the working potential
benzene. The basic titrants were standardized with primary ranges in the organic solvents were reduced and the
standard benzoic acid. potential endpoints were not as sharp as for titrations with
KC1 in methanol as the filling solution. Fritz (1973)
Preparation of Acid Solutions suggested that a saturated solution of tetramethylammo-
Solutions of HC1 and A1C13 were prepared in water to be nium chloride in 2-propanol would also make an excellent
approximately 5.ON. Additions of 10 /uliter of these solutions to filling solution.
 LOEPPERT ET AL.I ACIDIC PROPERTIES OF KAOLINITE 1103

+iooo r + 1000r-

+ 800 dimethylformamide + 800 -

• pyridine
• acetonltrlle
+ 600 2-propanol +600 - - — - pyridine

• water dimethylformamide
water
|+400 + 400 2-propanal
acetone
acetonitrile
+200 t-200
o
9>

I
-200 -200

-400 -400
••'I
0.5 1.0 1.5 2.0
0.04150 N Tetrabutylammonlum Hydroxide, ml
Fig. 1—Titration of benzoic acid in selected solvents with 0.0415Q/V -600
tetrabutylanunonium hydroxide, using a calomel reference elec- 0.5 1.0 1.5 2.0
trode with sleeve junction and glass indicator electrode. 0.04150 N Tetrabutylammonium Hydroxide, ml
Fig. 2—Titration of 10 juliter of 5.0V aqueous HC1 in 40 ml of
selected solvents with 0.04150/V tetrabutylammonium hydroxide,
Both the Orion sleeve junction calomel cell with glass using a combination electrode with porous plug liquid junction.
electrode and the Radiometer combination electrode were
found suitable for continuous potentiometric titrations in
nonaqueous solvents. The sleeve junction calomel cell quantity of standard base to reach the final potentiometric
required frequent filling, indicating a high rate of loss of endpoint in each of the solvents except DMF. When DMF
KC1 and methanol into the solvent system; therefore, in was used as the solvent medium, approximately 1.3 eq of
later titrations, the authors relied primarily on the Ra- standard base were required to reach the potentiometric
diometer electrode. endpoint of 1 eq of AlCb. A preliminary inflection was
Conditioning of the glass electrode affected the shapes of evident at approximately 80% of the titrant volume at the
the titration curves. Repeated titration in nonaqueous final endpoint. These phenomena may have resulted from
solvents apparently resulted in dehydration of the surface of formation of an Al-hydroxy complex with DMF or with
the glass electrode. The effect was especially evident in aluminate formation. Titration of AlCla yielded a single
DMF. Reproducible potentials were obtained when the endpoint in each of the other solvents. Titration steps
glass electrode was conditioned in an aqueous buffer for a corresponding to the separate stages of hydrolysis were not
few minutes between titrations in nonaqueous solvents. The evident.
electrode was removed from the aqueous buffer, rinsed
with deionized water, gently blotted dry, and placed Nonaqueous Titration of Kaolinite
immediately in the suspension to be titrated. Titrations of kaolinite samples in nonaqueous solvents
were affected by titrant flow rate, sample mixing, and
Nonaqueous Titration of Acids treatment of the glass electrode. For titrations of kaolinite
The glass electrode responded satisfactorily in water, samples, a stepped titration mode was employed with a
methanol, 2-propanol, DMF, pyridine, acetone, and ace- titrant addition rate of 0.03125 ml/min and a vertical sweep
tonitrile. Titration curves of benzoic acid (Fig. 1) and HC1 rate of 50mV/min. Solutions of KOH in benzene-methanol
(Fig. 2) showed sharp endpoints with each of the solvents. and tetrabutylammonium hydroxide in benzene-methanol
Wider potential ranges were obtained in the organic resulted in similarly shaped titration curves, however,
solvents than in the aqueous system. Working ranges of potential ranges were reduced when KOH was used as the
approximately 500, 1,100, and 1,200 mV were observed in titrant. Slightly lower potentials observed when KOH was
water, DMF, and acetonitrile, respectively, during the used as the titrant in the organic solvents may have been
titration of HC1. The wider potential ranges of organic due to an alkali error caused by K + .
solvents are advantageous in the resolution of mixtures of Adequate sample agitation was essential during non-
acids of different strengths (Fritz, 1973). aqueous titrations. Poor agitation resulted in low liters of
Titration of A1C13 (Fig. 3) required an equivalent standard base and nonreproducible inflections. Also, equil-
1104 SOIL SCI. SOC. AM. J., VOL. 41, 1977

+ 1000 +800

+600
+ 800 xx
X
. /
/
I +400
/ >
+ 600 dimethylformamide / E
pyridlne / o +200
i "^
water c
OJ
+400 methanol
I o - woter
• water with excess salt
———— acetonitrile
1+200 -200 ———— acetonitrile with excess salt

S
1 -400
I
I 2 3 4 5 6 7 8 9 10 12 13 14

Titratable Acidity, meq/100 g

-200 Fig. 4—Titration of 0.6 g of P2O5-dried K-kaolinite in 40 ml of water
with 0.0415<W KOH in water or in 40 ml acetonitrile with 0.04150/V
tetrabutylammonium hydroxide in benzene-methanol using a com-
bination electrode with porous plug liquid junction.
-400

(1972) observed a similar phenomenon in the nonaqueous

-600 tilralion of acidic nonlronile.
0.5 1.0 1.5 2.0
The surface acidily of kaolinile has been shown lo
0.04150 N Tetrabutylammonium Hydroxide, ml
increase wilh decreasing waler conlenl (Solomon and
ig. 3—Titration of 10 /iliter of 5.0N aqueous A1CI3 in 40 ml of Murray, 1972; Henmi and Wada, 1974). Maximum surface
selected solvents with 0.04150W tetrabutylammonium hydroxide
using a combination electrode with porous plug liquid junction. acidities equivalent to > 80% H2SO4 have been determined
with Hammett indicators. Acidities of protons associated
with structural water at the clay edges would be influenced
by the properties of the solvent. In a solvent such as water
ibration time of sample with solvent affected shape of the the charge of protons on the structural waters would be
titration curve. This particular aspect deserves further distributed due to H-bonding. In acetonitrile, the charge
study. would be distributed only by the weaker dipole-dipole
Working potential ranges for K-, H-, and Al-kaolinite interactions; therefore, the effective charge on the proton
(Fig. 4, 5, and 6, respectively) were significantly greater in would be greater .and repulsion of the positive charge
acetonitrile than in water due to the lower acidity and centers may be greater than in water. For this reason,
basicity of acetonitrile (King, 1973; Fritz, 1973; Kolthoff, acidity of edge sites of K-kaolinite may be greater in a very
1974). weakly basic aprotic solvent such as acetonitrile than in an
Titration of K-kaolinite in water resulted in an inflection amphiprotic solvent such as water. The larger liters of
at approximately 1.0 meq/100 g in contrast to titration in K-kaolinite in acetonitrile may be partially attributed to a
acetonitrile which produced a final inflection at approxi- stronger acidic character compared lo Ihe weaker acidic
mately 7.4 meq/100 g. In K-kaolinite the constant charge character in water.
sites would be occupied by K + ; therefore, acidic species Separate K-kaolinite samples were air-dried and dried
which were titrated were assumed to originate predomi- over PzOs. Titration curves in water were identical;
nantly from pH-dependent sites at the clay surface. The therefore, drying over P2Os did not result in irreversible
trailing nature of titration curves in water in the absence of production of protons.
excess neutral salt were indicative of neutralization of these Samples of P2O5-dried K-kaolinite were agitated in
weakly acidic pH-dependent sites. The larger liters in acetonitrile for 20 min and centrifuged. The supernate was
acetonitrile could partially be due to the weaker acidic lilraled and shown to contain a negligible acidic com-
character of acetonitrile compared to water (Kolthoff, ponent. Residual acetonitrile was removed from the clay
1974). Acidic properties of water prevent sharp endpoints under vacuum at room temperature. Titration curves of the
during titration of very weak acids (Fritz, 1973). Aprotic treated clay in acetonitrile and water were identical to
solvents with weakly acidic properties, such as acetonitrile, titration curves of the untreated P2O5-dried K-kaolinite.
are generally more suitable than water for the titration of Treatment of K-kaolinite with acetonitrile did not result in
weakly acidic species (Fritz, 1973; Kolthoff, 1974). For acidic products as determined by titration in water and
this reason larger total tilers of base were required to reach acetonitrile or in the irreversible production of protons. It
Ihe final polenliomelric endpoinl when acelonilrile was was therefore concluded that the acidic component mea-
used as Ihe lilration medium instead of water. Kapoor sured was due to surface properties of the kaolinite.
 LOEPPERT ET AL.: ACIDIC PROPERTIES OF KAOLINITE 1105
+ 800 _ +800

+600 +600

+400 +400

•g +200 _- +200
g
O
"c
<a s
£. o
-water £ - water
-water with excess salt >• water with excess salt
-200 ——.—acetonitrile -200 — acetonitrile
————acetanitrile with excess salt — acetonitrile with excess salt

-400 -400

I 2 3 4 5 6 7 8 9 10 11 12 13 14 I 2 3 4 5 6 7 8 9 10 II 12 13 14

Titratable Acidity, meq/100 g Titratable Acidity, meq/100 g

Fig. 5—Titration of 0.6 g of P2O5-dried H-kaolinite in water with Fig. 6—Titration of 0.6 of P2Ordried Al-kaolinite in 40 ml of water
0.0415<W KOH in water or in acetonitrile with 0.04150/V tetra- with 0.041501V KOH in water or in 40 ml of acetonitrile with
butylammonium hydroxide in benzene-methanol, using a com- 0.0415Q/V tetrabutylammonium hydroxide in benzene-methanol
bination electrode with porous plug liquid junction. using a combination electrode with porous plug liquid junction.

systems. Conductivity is lower in nonaqueous solvents and

Titration curves of H- and Al-kaolinite in water (Fig. 5 electrode response is slower. Therefore, it is necessary to
and 6, respectively) showed final inflections at approxi- perform titrations in a stepped mode or with an extremely
mately 3.9 meq/100 g. The final endpoints were not slow addition of titrant. Although potentials have been
substantially affected by the presence of excess neutral salt shown to be less reproducible in nonaqueous solvents than
although the shapes of the curves were affected. in water (Fritz, 1973), we found that with reasonable care,
The addition of salt would lower the activity of water in especially with regard to solvent purily and trealmenl of Ihe
the electrical double layer and would influence the in- electrodes, we could obtain reproducbile potentials. The
teraction of free water with protons associated with struc- presence of impurities in commercially available solvents
tural water at the clay edge. Therefore, the acidic properties may be a problem. Solvents must be checked routinely for
of the weak acid sites would be influenced by the presence acidic impurities, and in some cases further purification
of excess neutral salt. may be necessary (Riddick and Bunger, 1970). Also, the
The titration curve of H-kaolinite in water (Fig. 5) catalytic properties of dehydrated clays (Solomon and
contained two inflection points, which is more characteris- Murray, 1973) may lead to chemical changes in the solvent
tic of a H-Al-kaolinite. It is recognized that H-kaolinite will at the clay surface which could affect titration curves and
revert to a H-Al-kaolinite by autodecomposition. produce anomalous results. Such changes were not evident
Final inflections of H- and Al-kaolinite in acetonitrile in these experiments.
occurred at approximately 10 meq/100 g clay (Figs. 5,6). This study demonstrated the application of nonaqueous
It is interesting to note that the difference in total liter titrations in the determination of weakly acidic edge sites
between H-A1- or Al-kaolinite and K-kaolinite in ace- and/or constant charge sites on kaolinite. Many questions
tonitrile is approximately equal to the total liter of H-A1- or remain unanswered. For example, the effect of dehydration
Al-kaolinite in water at the final inflection (Table 1). These or solvation in solvents other than water on the acidic
differences may be attributed to the constant charge acidic nature of pH-dependent sites is not well understood. The
component of kaolinite. Several weak reproducible in- use of nonaqueous titrations should be further investigated
flections were evident in the titration curves of H-A1- and on other soil and pure clay systems.
Al-kaolinite in acetonitrile, especially in the presence of
excess neutral salt, but no attempt was made to give Table 1—Quantity of standard base required to reach the final
definite assignments to the individual inflections. However, potentiometric endpoint during titration of kaolinite in water or
it is interesting to note that the initial inflection at acetonitrile with 0.04150/V KOH with the presence and
absence of excess neutral salt (tetramethylanunonium
approximately 1 meq/100 g clay in acetonitrile in the chloride).
presence of excess salt corresponds closely to the initial
Saturating cation
inflection in water in the absence of excess neutral salt
which was atlributed to H+ (Fig. 5). Solvent Neutral salt K H-A1 Al
——— ml ———
CONCLUSIONS Water Absent 0.4 3.7 3.9
Present 0.3 3.8 3.7
Numerous problems are associated with nonaqueous Acetonitrile Absent 7.4 10.3 10.6
Present 7.3 10.7 10.7
tilrations which are generally not encountered in aqueous
1106 SOIL SCI. SOC. AM. J., VOL. 41, 1977