ARTICLE IN PRESS

0043-1354/$ - se

doi:10.1016/j.w

�CorrespondE-mail addr1Present addr

Water Research 39 (2005) 3457–3466

www.elsevier.com/locate/watres

Effect of citric acid on aluminum hydrolytic speciation

Wen Hui Kuana, Ming Kuang Wangb,�, Pan Ming Huangb,1, Chia Wen Wub,Chia Ming Changc, Shan Li Wangc

aDepartment of Environmental and Safety Engineering, Ming-Chi University of Technology, Taipei County 243, TaiwanbDepartment of Agricultural Chemistry, National Taiwan University, Taipei 106, Taiwan

cDepartment of Soil and Environmental Sciences, National Chung Hsing University, Taichung 402, Taiwan

Received 20 December 2004; received in revised form 8 April 2005; accepted 27 May 2005

Abstract

The mechanisms of the influence of organics on Al transformation were not fully understood. This study investigated

the effect of citric acid on Al speciation in partially neutralized aluminum solution. The partially neutralized solution

was prepared with 20mmol L�1 AlCl3 without citrate (citrate/Al molar ratio of 0, control) or with citrate (citrate to Al

molar ratios between 0.1 and 3.0) at pH between 3.0 and 7.0. The nature of aluminum hydrolytic products as influenced

by citrate complexation was investigated by turbidity measurement, ferron kinetic color development, peak line width

in 27Al nuclear magnetic resonance (NMR) deconvolution demodule quantitative methods, and the MinteqA2

chemical speciation program. Sulfate precipitates from Al solution as influenced by citrate were examined by X-ray

diffraction (XRD) analysis and atomic force microscopy (AFM). The turbidity of the Al solution increased with

increasing pH values. Increases in citrate/Al molar ratio from 0 to 0.1 decreased dramatically the turbidity due to citrate

complexation. The 27Al NMR peak at 6 ppm of the Al solution at a citrate/Al molar ratio of 0.5 shifted to 8 and

10 ppm in the solutions at citrate/Al molar ratio of 1.0 and 3.0, respectively. Comparison of 27Al NMR data and the

data obtained from the MinteqA2 chemical speciation program, indicate that the Al-citrate complexes as revealed by27Al NMR data are largely Al(citrate), AlH(citrate)+, and AlðcitrateÞ3-2 complexes (99–112%) in the pH range of

4.5–6.5. The non-detected Al fractions by the MinteqA2 program account for 82–99% of the non-detected Al fractions

by NMR quantitation in the same pH range. The AFM of sulfate precipitates from solutions with low citrate/Al molar

ratios (i.e., 0.01, pH 4.5, aged 40 days) shows that Al13 sulfate precipitates were ellipse-shaped. These ellipse-shaped

precipitates were aggregated when solution pH increased from 4.5 to 7.0 (aged 40 days), indicating the fast hydrolytic

rate of Al at high pH. The sulfate precipitates from solution with a high citrate/Al molar ratio (i.e., 0.05, pH 4.5, aged

40 days) also shows aggregate of particles, and XRD non-crystallized precipitates the hampering effect of citrate on Al

precipitates.

r 2005 Elsevier Ltd. All rights reserved.

Keywords: Al-citrate complexes; Al hydrolysis; 27Al NMR; Ferron kinetic method; MinteqA2 chemical speciation program; Sulfate

precipitates; X-ray diffraction (XRD)

e front matter r 2005 Elsevier Ltd. All rights reserved.

atres.2005.05.052

ing author. Tel.: (0118862) 2363-0231x2491 or 3066; fax: (0118862) 2366 0751.

ess: mkwang@ntu.edu.tw (M.K. Wang).

ess: Department of Soil Science, University of Saskachewan, 51 Campus Drive, Saskatoon, SK, Canada S7N 5A8.

ARTICLE IN PRESSW.H. Kuan et al. / Water Research 39 (2005) 3457–34663458

1. Introduction

Aluminum is released from aluminosilicate minerals

into soil solution due to acidity generated by natural

processes. Subsequently, Al ion undergoes hydrolysis or

forms strong complexes with naturally occurring organic

acids commonly present in soils. Organic acids are

involved in influencing the formation of most of the Al

oxides under pedogenic environments (Huang et al.,

2002). The strength of metal and anion retention by

aluminum hydroxides and oxyhydroxides plays a sig-

nificant role in governing the mobility of these anions in

soils (Huang et al., 2002; Parfitt, 1978). Organic ligands

of low molecular weight (Kwong et al., 1979; Krishna-

murti et al., 1999; Wu and Wang, 2000), tannic (Kwong

et al., 1981), fulvic (Kodama and Schnitzner, 1980), and

humic (Singer and Huang, 1990) perturb the hydrolytic

reactions of Al, resulting in the formation of X-ray non-

crystalline products. The relative distribution of the

hydrolyzed and complexed Al species is a determining

factor for Al mobility in soil solution and toxicity to

natural ecosystems (Buffle et al., 1985; Huang, 1988;

Bertsch and Parker, 1996; Chen et al., 2001).

Citric acid, one of the major low-molecular-weight

organic acids derived from root exudates, decomposing

organic matter and other sources, is an important

organic ligand in nature. The concentration of citric

acid ranged from 10 to 1000 mM (Robert and Berthelin,

1986). Citric acid can greatly influence the hydrolysis of

Al and modify the crystalline structure of hydrolytic

products (Huang, 1988). The interaction of polybasic

anions with aluminum is of interest because these salts

and acids are often used in the extraction of aluminum

and iron from poorly crystallized phases in soils (i.e.,

oxalic and citric acids). However, the mechanism of the

influence of citrate on Al transformation is still not fully

understood. In the present study, the nature of

complexes of hydrolytic products of Al with citrate

was investigated by turbidity measurement, ferron

kinetic color development, and 27Al nuclear magnetic

resonance (NMR) deconvolution demodule quantitative

methods of Al fractionations, and MinteqA2 program

to calculate the Al hydrolytic products and Al-citrate

speciation. Sulfate precipitates were investigated by X-

ray diffraction (XRD) analysis and atomic force

microscopy (AFM) to examine further the nature of

Al-citrate complexes.

2. Materials and methods

2.1. Synthesis of Al-polycations

2.1.1. Stock solution

(a) Stock solution of AlCl3: We added 120.7 g of

AlCl3.6H2O (Merck, GR) to 1L of deionized water and

stored it in a high-density polyethylene bottle (HDPE).

The Al concentration of stock solution determined by

atomic absorption spectrophotometer (AAS Hitachi

180-30) was 0.998M. (b) Standardized NaOH solution:

We added 500 g NaOH (Merck, GR) to boil deionized

water to prepare 0.1M NaOH solution which was

standardized by potassium hydrogen phthalate (Merck,

GR). The standardized NaOH solution was stored in

HDPE bottle.

2.1.2. Sample preparation

The control Al concentration was 20mmol L�1

(citrate to Al molar ratio of 0). Citrate (i.e., Na3-citrate,

Merck, GR) to Al molar ratio ranged from 0.01 to 3.0.

The Al-citrate and control Al solutions were titrated

with a Mettler DL 25 titroprocessor in the set mode to

pH 4.0, 4.5, 5.0, 5.5, 6.0, 6.5 and 7.0 with NaOH or HCl

solution.

2.1.3. Analytical methods

The Al-citrate and control Al solutions were examined

by measurement of their pH and turbidity, Al-ferron

kinetic color development, and 27Al solution NMR

analyses. XRD analysis and AFM were employed to

characterize the sulfate precipitates with hydrolytic

products of Al and Al-citrate complexes.

2.1.4. pH and turbidity determination

The pH and turbidity of the fresh or aged control Al

and Al-citrate (with various citrate/Al molar ratios)

solutions analyzed using Radiometer pHm83 pH meter

and Hack ratio/XR turbidity meter, respectively. For-

mazin solution was used to calibrate the turbidity.

2.1.5. Determination of Al solution with Al-ferron kinetic

color development method

The Al concentrations were below 0.30mmol L�1.

Two-milliliters of Al (control) or Al-citrate (different

citrate/Al molar ratios) solutions were added into a

10mm path quartz cell followed by the addition of 2mL

of ferron reagents. The mixture was shaken for 10 s and

Al was determined at 366 nm using UV–Vis spectro-

photometer (GBC 911 model) by the kinetics of Al-

ferron color development. The detailed procedures of

fractionations by ferron kinetic color development were

reported in Wang and Hsu (1994), Hsu (1997) and Wang

et al. (2003).

2.1.6. 27Al nuclear magnetic resonance (NMR) analyses27Al NMR spectra were collected at 104.206MHz

using a Varian 400 FT-NMR spectrometer. Five

milliliters of Al solution was loaded into a 10mm

sample tube (Wilmad 513-7-pp) and 2mL 0.05M

Al(OD)4� solution in a 5mm sample tube (Wilmad

507-pp) were co-inserted as the internal standard. The

Al(OD)4� solution was prepared by dissolving Al metal

ARTICLE IN PRESSW.H. Kuan et al. / Water Research 39 (2005) 3457–3466 3459

into 40% NaOD solution (Acros, 99.5% D). The

chemical shift of Al(OD)4� is 80 ppm. Data acquisition

conditions were: spectrometer frequency, 104.206MHz;

pulse length, 10ms; recycle delay, 3 s; spectra width,22857.1Hz; spinning speed, 20 rps; and number of

transient, 1600 times.

Spectral analyses were conducted using Bruker Xwin-

nmr (Version 1.3). The spectral features were curve-

fitted using Lorentz line shape. The concentration of

each Al species was then determined by the ratio of the

integrated intensity of the corresponding peak to that of

Al(OD)4� at 80 ppm. The amount of the NMR

undetected species was calculated by subtracting the

sum of the detected species from the total Al concentra-

tion in the samples. The non-detected Al fraction (Alnd)

from the NMR analysis was the total amount of Al

minus monomeric Al (0 ppm) and Al13 (62.5 ppm) and/

or Al-citrate complex (6–10 ppm) fractions.

Quantitative analysis of the NMR deconvolution

module of integration of peak area was made using

Bruker vwin-nmr 1.3 vesion software. The amplitude,

position, width and Gauss/Lorentz of NMR peak was

adjusted by the deconvolution module and setting

Gauss/Lorentz as zero. The quantitative data were

obtained based on the method for the regression of

Lorentzian curve (Wang et al., 2003).

2.1.7. Sulfate precipitates

For each sample, 100mL of 0.02M sodium sulfate

solution were added into the same volume of a sample

solution. The sulfate precipitates formed after aging for 1

day were collected using a Millipore filter of 0.2mm poresize. Detailed procedures for the Al fractionation and

sulfate precipitation methods can be found in Tsai and Hsu

(1984, 1985), Bertsch et al. (1986a), Hsu (1988), Hsu and

Cao (1991), Wang and Hsu (1994) and Wang et al. (2003).

The Al sulfate precipitates were examined with an X-

ray diffractometer (Rigaku Miniflex) with CuKa radia-tion with Ni filter. The XRD patterns were recorded in

the range of 2–5012ywith a scan rate of 112ymin�1. Theoperation voltage was 30 kV with 10mA current.

Precipitates were ultrasonication and one drop of

suspensions was dried on double-stick tape of stainless

disk. Atomic force microscopic analysis of sulfate

precipitates was employed using the NanoScope III

multimode scanning probe microscope (Si3N4 Cantile-

vers with integral tips; Digital Instruments, Inc., Santa

Barbara, CA) and NanoScope III software version 2.32rl

(1998 version) for morphology investigation. The oper-

ating conditions were: scanning mode, tapping mode;

scanner type, AS-2V; scanner size, 16mm and scanningrate, 0.2–1.5Hz. The investigation was run at 25 1C.

2.1.8. MinteqA2 chemical speciation method

The MinteqA2 Chemical Speciation Program (version

4.02) was employed to calculate the Al citrate speciation

in partially hydrolyzed Al solution (MinteqA2 Chemical

Speciation Program, 2000).

3. Results and discussion

In the control samples aged for 40 days, the

monomeric Al and polymeric Al fractions decreased

with increasing pH, but their turbidity, and colloidal and

precipitate fractions increased (Table 1a). However, with

the addition of citrate (citrate to Al molar ratio of 0.1,

aged 40 days), the solution became less turbid especially

at the pH of 5.0 and higher (Table 1b). In the ferron

reaction kinetics, the reaction rate of the control samples

was faster than that of the citrate-Al solution when the

citrate to aluminum molar ratio was below the critical

value (i.e., citrate/Al molar ratiop0.13) (Jardine andZelazny, 1987). Except for the systems at pH 6.0, the

monomeric Al content in the citrate/Al ¼ 0.1 solution

was lower than that of the control samples (Table 1).

The polymeric Al fractions decreased significantly, while

the colloidal or precipitate Al fractions increased with

increasing pH in the citrate/Al ¼ 0.1 system (Table 1b).

Similar results were also observed in the complexation of

succinate and tartrate with aluminum (Krishnamurti et

al., 1999; Earl et al., 1979; Hue et al., 1986). Jardine and

Zelazny (1987) reported that citrate ion interferes the

ferron color development in Al systems. In addition,

the citrate-Al complexes cannot be distinguished from

the OH–Al complexes. Therefore, the samples were

further investigated using 27Al NMR analysis as

discussed below.

3.1. 27Al NMR analysis of Al solution at citrate/Al

molar ratio of 0

A peak of chemical shift was present in the 27Al

NMR spectrum of the control solution at 0 ppm in pH

range between 3.0 and 5.0 (Fig. 1a–d). This peak of

0 ppm should include Al(H2O)63+, Al(H2O)5(OH)

2+ and

Al(H2O)4(OH)2+ species (Akitt et al., 1972). The

intensity of the 0 ppm peak decreases when pH increases

and the peak vanishes at pH45.0. On the contrary, apeak of chemical shift at 62.5 ppm, corresponding to

Al13 polycation was observed at pH44.0, and its

intensity reaches maximum at pH ¼ 5.0 (Bertsch et al.,

1986a; Akitt and Farthing, 1978; Teagarden et al., 1981;

Bertsch et al., 1986b). As seen in Table 2, the quantity of

the Al13 complex (62.5 ppm) increased with increasing

pH and reached the maximum at pH 5.0. Subsequently,

the Al13 polycation fractions decreased when pH

increases to 7.0 (Fig. 1e–h). The disappearance of Al13polycation is due to the formation of Al hydroxide

precipitates, as indicated by the increase in the turbidity

of the samples and the formation of colloidal and

precipitated Al (Table 1). The increase in turbidity with

ARTICLE IN PRESS

80 60 40 20 0chemical shift (ppm)

(a)(b)(c)(d)

(e)(f)

(g)

(h)

Fig. 1. The 27Al NMR spectra of partially neutralized

aluminum solutions (citrate/Al ¼ 0) at (a) pH 3.0, (b) pH 4.0,

(c) pH 4.5, (d) pH 5.0, (e) pH 5.5, (f) pH 6.0, (g) pH 6.5, and (h)

pH 7.0.

Table 1

Aluminum speciation of partially neutralized solutions determined by ferron kinetic color development

PH OH/Al molar ratio Turbidity (FTU)a Ferron

Monomeric Al Polymeric Al Colloidal or ppt. Al

mmolL�1 as Al

(a) Citrate/Al ¼ 0 of control solution

4.0 0.17 0.7 9.33 10.66 0

4.5 1.12 0.4 8.48 11.38 0.14

5.0 2.30 72.5 7.41 10.89 1.70

5.5 2.42 80.5 7.11 10.67 2.22

6.0 2.52 170.3 2.67 8.89 8.44

(b)Citrate/Al ¼ 0.1 solution at different pH

4.0 0.54 0.3 7.44 12.56 0

4.5 0.66 0.4 7.11 12.89 0

5.0 0.98 0.6 6.85 4.48 8.67

5.5 1.12 0.8 6.59 3.82 9.59

6.0 1.18 0.8 6.30 3.37 10.33

Control and citrate/Al molar ratio solution at 0.1 were aged for 40 days (Al concentration is 20mmolL�1).aFTU: Formazin turbidity unit.

W.H. Kuan et al. / Water Research 39 (2005) 3457–34663460

increasing solution pH may be due to the formation of

the clusters and sheet structure (Hsu, 1997; Bottero

et al., 1982). The non-detected Al fractions (Aln), which

are apparently the colloidal and precipitated Al, also

increased with increasing solution pH (Table 2).

Several studies on Al (i.e., 0.5M) hydrolysis under

high concentration of Al showed the occurrence of a

dimmer species (Al2(OH)24+), which was detected at

3 ppm in the 27Al NMR spectra (Akitt et al., 1972;

Akitt and Farthing, 1978; Bottero et al., 1982; Partha-

sarathy and Buffle, 1985; Greenaway, 1986). Such

dimmer species was not observed here because the Al

concentration (i.e., 20mmolL�1) of the samples is

relatively low in this study. Therefore, the amount of

Al dimmer is considered negligible.

3.2. 27Al NMR spectra of Al solutions at citrate to Al

molar ratios of (a) 0.1, (b) 0.2, (c) 0.5, (d) 1.0,

and (e) 3.0

In a series of different citrate/Al molar ratios

(pH ¼ 4.0), a broad peak at chemical shift range of

6–8 ppm as well as the 0 ppm peak are present in the27Al NMR spectra of the solutions (Fig. 2). The new

peak at 6–8 ppm corresponds to Al-citrate complexes.

On the other hand, Al13 does not occur as no 62.5 ppm

peak was observed. The formation of Al13 is inhibited in

the presence of citrate ions in the samples (Kwong et al.,

1979; Jardine and Zelazny, 1987; Thomas et al., 1993).

When the citrate/Al molar ratios increase from 0.5 to

1.0, the line width of the broad peaks at 6–8 ppm

increase from 500 to 714Hz (Wu and Wang, 2000). As

ARTICLE IN PRESS

Table 2

Aluminum speciation of partially neutralized solutions determined by 27Al NMR deconvolution demodule quantitative method and

calculated by the MinteqA2 chemical program

pH NMR deconvolution MinteqA2

Monomeric Al (0 ppm) Al13 (62.5 ppm) Alnda Monomeric

AlbAl13 AlðOHÞ�3

mmolL�1 as Al

(a) Citrate/Al ¼ 0 of control solution

3.3 20.00 0 0 20.00 0 0

4.0 17.33 0 2.67 17.07 0.44 2.49

4.5 7.77 0.52 11.71 6.81 1.08 12.11

5.0 0.59 1.33 18.08 0 1.30 18.70

5.5 0 0.96 19.04 0 0 20.00

6.0 0 0.82 19.18 0 0 20.00

6.5 0 0.37 19.63 0 0 20.00

pH NMR deconvolution MinteqA2

Monomeric Al (0 ppm) Al-citrate complex

(6–10ppm)

Alnda Mono-meric Alb Al(citrate) AlH(citrate)+ AlðcitrateÞ3-2 AlðOHÞ�3

mmolL�1 as Al

(b) Citrate/Al ¼ 0.5 solution at different pH

4.0 12.07 7.93 0 8.59 0.26 9.60 0.07 1.48

4.5 1.74 7.70 10.56 1.52 0.56 6.52 1.44 9.96

5.0 0 5.52 14.48 0.11 0.19 0.78 4.48 14.44

5.5 0 4.56 15.44 0 0.04 0.04 4.89 15.03

6.0 0 4.30 15.70 0 0 0 4.81 15.19

6.5 0 3.96 16.04 0 0 0 4.26 15.74

7.0 0 2.15 17.85 0 0 0 2.37 17.63

AlðOHÞ�3 : represents the non-settling gel of Al-hydroxide.Control and citrate/Al molar ratio solution at 0.5 were fresh in preparation

(Al concentration is 20mmolL�1).aAlnd represents the fraction of NMR not detected.bMonomeric Al considered in MinteqA2 modeling, including Al3+, AlOH2+, AlðOHÞþ2 , and AlðOHÞ

-4.

W.H. Kuan et al. / Water Research 39 (2005) 3457–3466 3461

the citrate/Al molar ratio further increased to 3.0, the

chemical shift of the broad peak shifted to 10 ppm and

the line width became very broad (i.e., 1251Hz). On the

contrary, the peak intensity of monomeric Al speciation

(0 ppm) decreased with increasing citrate/Al molar ratios

from 0.1 to 1.0, which reveals the transformation of

monomeric Al species to Al-citrate species with increas-

ing citrate/Al molar ratios. At the citrate/Al molar ratio

of 3.0, monomeric Al species was not observed. Citrate

anion gradually replaced the coordination of water and

formed Al-citrate complexes. In 27Al NMR spectra, the

chemical shift and line width of the peaks corresponding

to Al-citrate complexes were influenced by Al coordina-

tion geometry, and symmetry, and exchange rates

between species (Parthasarathy and Buffle, 1985; Green-

away, 1986). In general, the 27Al NMR analysis is a

sensitive and useful tool for determining the broad line

width of resonance peaks caused by the asymmetry of

complexes and the fast exchange rate. As pH increases,

the chemical shift and line width of the 6-ppm peak

increases. The increase in line width is mainly due to fast

exchange rate between citrate and OH. The chemical

shift of the peak was gradually moved to low magnetic

field due to octahedral Al coordination slowly changing

to tetrahedral Al coordination and water line width

(Karlik et al., 1982; Motekaitis and Martell, 1983).

Thomas et al. (1991) also reported that organic acid

possessed anti-shielding magnetic effect. Therefore, the

chemical shift of the peak moved to lower magnetic field

due to anti-shielding effect of Al-organic ligand.

3.3. Comparison between MinteqA2 chemical speciation

program calculation and 27Al NMR deconvolution

demodule quantitative method of control Al solution

(fresh sample preparation)

The equilibrium equations and constants used in the

MinteqA2 calculation for monomeric Al, including

Al3+, AlOH2+, AlðOHÞþ2 , and AlðOHÞ-4, are as follows

(Huang, 1988; Bertsch and Parker, 1996):

Al3þ þH2O ¼ AlðOHÞ2þ þHþ logK ¼ �4:79; (1)

ARTICLE IN PRESS

80 60 40 20 0chemical shift (ppm)

(a)

(b)

(c)(d)

(e)

Fig. 2. The 27Al NMR spectra of citrate to Al molar ratios of

(a) 0.1, (b) 0.2, (c) 0.5, (d) 1.0, and (e) 3.0 solutions (pH ¼ 3.0).

Fig. 3. Schematic structures of (a) Al(citrate), (b) AlH(citrate)+

and (c) AlðcitrateÞ3-2 ((a) and (b) adapted from Motekaitis and

Martell (1983)).

W.H. Kuan et al. / Water Research 39 (2005) 3457–34663462

Al3þ þ 2H2O ¼ AlðOHÞ2þ þ 2Hþ logK ¼ �10:10;

(2)

Al3þ þ 4H2O ¼ AlðOHÞ4� þ 4Hþ logK ¼ �23:00:

(3)

Several previous reports (Bottero et al., 1982; Rubin

et al., 1973; Bottero et al., 1980; Bottero et al., 1987)

displayed the sudden increase in turbidity, which is

attributable due to the formation in solution of the non-

settling gel, and this gel form of Al-hydroxide repre-

sented as AlðOHÞ�3 in order to better differentiate these

species from amorphous Al(OH)3(s).

Al3þ þ 3H2O ¼ AlðOHÞ3� þ 3Hþ logK ¼ �11:80:

(4)

This gel (or sol) could result from the primary

nucleation of the Al-hydroxide (Sohnel and Garside,

1992). The primary particulates are formed in freshly

prepared solution and the size of precipitates is very

tiny. The freshly prepared precipitates are more soluble

than that in aged precipitates, and this decreasing

solubility with time is probably a result of the

elimination of the smallest particles at the expense of

growing larger particles with prolonged aging (Sohnel

and Garside, 1992). In other words, the equilibrium

solubility of a substance increases with decreasing

crystal size and this phenomenon is apparent for smaller

than about 1mm (Sohnel and Garside, 1992). Accordingto the logK (�10.8) of amorphous Al(OH)3(s) in

MinteqA2 version 4.02, it was found that the equili-

brium solubility of Al3+ for AlðOHÞ�3 is 10 times larger

than for amorphous Al(OH)3(s) at corresponding condi-

tions.

The hydrolytic reaction of Al to form of Al13 is shown

in Eq. (5):

13Al3þ þ 40H2O ¼ Al13O4ðOHÞ24ðH2OÞ127þ

þ 32Hþ logK ¼ �105:00: (5)

We used log K ¼ �105 to calculate Al13 fractions

(Thomas et al., 1993). In control aluminum solution,

ARTICLE IN PRESSW.H. Kuan et al. / Water Research 39 (2005) 3457–3466 3463

100% of monomeric Al was present at pH 3.3; however,

both the monomeric Al by NMR quantitation and by

MinteqA2 calculation decreased with increasing solu-

tion pH (Table 2). The Al13 fractions determined by

NMR quantitation and MinteqA2 calculation appeared

at pH 4.0–4.5 and reached the maximum at pH 5.0, and

then substantially decreased at higher pH (Table 2a).

The amounts of non-detected Al fractions (Alnd) from

NMR quantitation were close to AlðOHÞ�3 fractions

calculated by MinteqA2 chemical speciation program

(Table 2a). Quantity of these two species increased with

increasing solution pH.

3.4. Comparison between MinteqA2 program of chemical

speciation calculation and 27Al NMR deconvolution

module quantitative analysis of solution with citrate/Al

molar ratio of 0.5 (fresh sample preparation)

The citrate bonds with different Al species with

varying bonding strengths. The dipole–dipole interac-

tion cannot distinguish between different organic-Al

species, resulting in the broader line width in the NMR

spectrum. Bottero et al. (1980), Karlik et al. (1982), and

Thomas et al. (1991, 1993); employed 27Al NMR,

potentiometric titration and MinteqA2 program to

quantify the Al-citrate spectrum in the complex form.

Here, we used the method of Kwong et al. (1977, 1979);

to interpret the influence of citrate on aluminum

hydrolytic reaction and the stability constants of Al-

Table 3

X-ray diffraction data of sulfate precipitates from Al solution at citrate

(pH ¼ 7.0, aged 40 days, and reference data

Citrate/Al ¼ 0.01 Citrate/Al ¼ 0.05

pH 4.5 d (nm) Peak intensity pH 7.0 d (nm )

1.283 (10) 1.254

1.028 (5) 1.013

0.921 (1) —

— 0.885

0.712 (1) —

— 0.611

— 0.537

0.464 (1) 0.454

0.443 (1) —

0.424 (1) —

0.410 (1) 0.404

0.389 (2) —

0.336 (1) 0.330

0.302 (3) 0.300

0.269 (1) —

0.245 (1) —

0.242 (1) —

0.235

citrate complexes reported by them and the database of

MinteqA2 program are as follows:

Al3þ þ citrate3� ¼ AlðcitrateÞ logK ¼ 7:37 (6)

Al3þ þHþ þ citrate3� ¼ AlHðcitrateÞþ logK ¼ 12:85

(7)

Al3þ þ 2 citrate3� ¼ AlðcitrateÞ23� logK ¼ 13:90 (8)

As shown in Table 2b (citrate/Al ¼ 0.5 solution), the

sum of Al(citrate), AlH(citrate)+, and AlðcitrateÞ3-2complexes calculated by MinteqA2 program accounts

for 99–112% Al-citrate complexes determined by NMR

quantitation in the pH range of 4.5–6.5. Therefore, Al-

citrate complexes determined by NMR quantitation are

largely attributable to Al(citrate), AlH(citrate)+ and

AlðcitrateÞ3-2 complexes. Both Al-citrate complexes (i.e.,

chemical shift at 6–10 ppm) determined by NMR

quantitation and the sum of Al(citrate), AlH(citrate)+

and AlðcitrateÞ3-2 complexes calculated by MinteqA2

method decreased with increasing pH. By contrast,

both the fractions (Alnd) not detected by NMR

quantitation and the AlðOHÞ�3 calculated by MinteqA2

speciation increased with increasing solution pH. The

AlðOHÞ�3 fractions determined by MinteqA2 method

account for 82–99% of those determined by NMR

quantitation in the pH range of 4.5–6.5. Motekaitis

and Martell (1983) studied the chemical structure

of Al-citrate complexes. The stability constant of

/Al ¼ 0.01 (pH ¼ 4.5, aged 40 days), citrate/Al ¼ 0.05 solutions

Kwong et al. (1977)

Peak intensity d (nm ) Peak intensity

(10)

(10)

(2)

0.708 (9)

(1) —

(2) —

(1) 0.462 (9)

—

0.432 (10)

(3) —

—

(3) —

(1) 0.302 (6)

—

—

—

(4)

ARTICLE IN PRESSW.H. Kuan et al. / Water Research 39 (2005) 3457–34663464

Al-citrate is high and forms a stable complex. The

COOH and OH groups in the citrate can act as a

coordination donor to the reaction site of Al. Thus,

Al(citrate), AlH(citrate)+ and AlðcitrateÞ3-2 can be

schemated as shown in Fig. 3. Kwong et al.

(1977,1979) reported that citrate possessed strong

affinity with Al and inhibited Al hydrolysis. Hue et al.

(1986) reported that the stability of Al-citrate complexes

are related to the position of OH and CO groups in the

citrate. Two para-sites of OH and CO groups (i.e.,

Fig. 4. Atomic force micrographs of sulfate precipitates from solution

and (b) pH 7.0; and citrate/Al molar ratio of 0.05 at (c) pH 4.5, and (

citrate and succinate) easily complex with Al and form

five- or six-member-ring stable structure, inhibiting Al

hydrolysis (Sohnel and Garside, 1992; Masion et al.,

1994a, b, c).

3.5. X-ray diffraction (XRD) and AFM of Al-sulfate

precipitates

XRD of sulfate precipitates from citrate/Al ¼ 0.01

solutions (pH 4.5, aged 40 days) yielded intense peaks at

with citrate/Al molar ratio of 0.01 aged 40 days at (a) pH 4.5,

d) pH 7.0 aged 40 days. The scale is the same from Figs. 4a–d.

ARTICLE IN PRESSW.H. Kuan et al. / Water Research 39 (2005) 3457–3466 3465

1.283 and 1.028 nm, and the other tiny XRD peaks

(Table 3). These XRD peaks are attributed to the basic

aluminum sulfate (13Al2O3.6SO3.xH2O) (Tsai and Hsu,

1984). The 0.712, 0.464 and 0.302 nm XRD peaks are

close to those reported by Kwong et al. (1977,1979). The

citrate/Al molar ratio of 0.05 solution (pH ¼ 7.0, aged

40 days) yielded intense 1.254 and 1.013 nm, and other

tiny XRD peaks (Table 3). However, precipitates from

citrate/Al molar ratio of 0.01 solutions (pH ¼ 7.0, aged

40 days) shows X-ray non-crystallized precipitates,

indicating fast hydrolytic rate of Al hydroxide precipi-

tates. Increasing citrate/Al molar ratio to 0.05

(pH ¼ 4.5, aged 40 days) yielded X-ray non-crystallized

precipitates. This is caused by the enhanced structural

perturbation of Al hydroxides due to the increasing

citrate/Al molar ratio.

The XRD data (Table 3) show that the sulfate

precipitates from the citrate/Al molar ratio of 0.01

solutions at pH 4.5 (aged 40 days) was the Al13 sulfate

complexes. The combination of the atomic force

micrographs and the X-ray data, thus, indicate that

these ellipse-shaped particles are Al13 sulfate complex

(Fig. 4a). The ellipse-shaped Al13 sulfate complexes were

transformed to aggregates of particles when solution pH

was increased to pH 7.0 (Fig. 4b). The atomic force

micrograph of sulfate precipitates from solution with

high citrate/Al molar ratio of 0.05 and pH 4.5 (aged 40

days) shows that the precipitates were smaller in size

compared with those formed at the citrate/Al molar

ratio of 0.01 and pH 4.5 (Figs. 4a and c), indicating the

hampering effect of citrate on Al precipitates. However,

the precipitates from solution with citrate/Al molar ratio

of 0.05 and pH 7.0 substantially increased in particle

sizes (Fig. 4d), resulting in formation of new Al

complexes with hydroxide, citrate and sulfate.

4. Conclusions

Comparison of the data obtained from 27Al NMR

spectroscopy and the MinteqA2 chemical speciation

program calculations indicate that these Al-citrate

complexes as revealed in 27Al NMR spectra are largely

Al(citrate), AlH(citrate)+ and AlðcitrateÞ3-2 complexes.

The sum of Al(citrate), AlH(citrate)+ and AlðcitrateÞ3-2complexes determined by MinteqA2 calculation ac-

counts for 99–112% of Al-citrate complexes determined

by 27Al NMR quantitation in the pH range of 4.5–6.5.

The present study shows, for the first time, that the non-

settling gel of Al-hydroxide (AlðOHÞ�3), determined by

the MinteqA2 calculation, accounts for 82–99% of the

non-detected Al fraction determined by NMR quantita-

tion. This Al fraction increases with increasing pH from

4.5 to 6.5 at the citrate/Al molar ratio of 0.5. Atomic

force micrographs and XRD analysis of sulfate pre-

cipitates from solutions with low citrate/Al molar ratios

(pH 4.5, aged 40 days) show ellipse-shaped Al13 sulfate

complexes. With citrate to Al molar ratio of 0.05

solution at pH 4.5 the precipitate particles were smaller

in size and XRD non-crystalline, indicating the hamper-

ing effect of citrate on Al precipitation. When solution

pH was increased to pH 7.0 at low citrate to Al molar

ratio (i.e., citrate/Al ¼ 0.01), sulfate precipitates were

aggregated as XRD non-crystalline. This is attributed to

the enhanced hydrolytic rate of Al in the formation of Al

hydroxide precipitates at high pH.

Acknowledgments

This work was financially supported by the National

Science Council, Taiwan, Republic of China, under

project #NSC88-2313-B002-279, 89-2621B002-006, 89-

2313-B002-279 and B90-2313-279.

References

Akitt, J.W., Farthing, A., 1978. New Al-27 NMR studies of the

hydrolysis of the aluminum (III) cation. J. Magn. Reson. 32,

345–352.

Akitt, J.W., Greenwood, N.N., Khandelwal, B.L., Lester,

G.D., 1972. Al-27 nuclear magnetic resonance studies of

the hydrolysis and polymerization of the hexa-aluminu-

m(III) cation. J. Chem. Soc. Dalton., 604–610.

Bertsch, P.M., Parker, D.R., 1996. Aqueous polynuclear

aluminum species. In: Sposito, G. (Ed.), The Environmental

Chemistry of Aluminum. Lewis Publications, New York,

pp. 117–168.

Bertsch, P.M., Layton, W.J., Barnhisel, R.I., 1986a. Speciation

of hydroxy-aluminum solutions by wet chemical and

aluminum-27 NMR method. Soil Sci. Soc. Am. J. 50,

1449–1454.

Bertsch, P.M., Thomas, G.W., Barnhisel, R.I., 1986b. Char-

acterization of hydroxy-aluminum solutions by aluminum-

27 nuclear magnetic resonance spectroscopy. Soil Sci. Soc.

Am. J. 50, 2583–2585.

Bottero, J.Y., Cases, J.M., Fiessinger, F., Poirier, J.E., 1980.

Studies of hydrolyzed aluminum chloride solutions. 1.

Nature of aluminum species and composition of aqueous

solutions. J. Phys. Chem. 84, 2933–2939.

Bottero, J.Y., Tchoubar, D., Caes, J.M., Fiessinger, F., 1982.

Investigation of the hydrolysis of aqueous solutions of Al

chloride. 2. Nature and structure by small-angle X-ray

scattering. J. Phys. Chem. 86, 3667–3673.

Bottero, J.Y., Axelos, M., Tchoubar, D., Cases, J.M., Fripiat,

J.J., Fiessinger, F., 1987. Mechanism of formation of

aluminum trihydroxide from Keggin Al polymers. J. Colloid

Interface Sci. 117, 47–57.

Buffle, J., Parthasarathy, N., Haerdi, W., 1985. Importance of

speciation methods in analytical control of water treatment

processes with application to fluoride removal from waste-

waters. Water Res. 19, 7–23.

Chen, M.-C., Wang, M.K., Chiu, C.Y., Huang, P.M.,

King, H.B., 2001. Determination of low-molecular-weight

ARTICLE IN PRESSW.H. Kuan et al. / Water Research 39 (2005) 3457–34663466

dicarboxlic acids and organic functional groups in rhizo-

sphere and bulk soils of Tsuga and Yushania in a temperate

rain forest. Plant Soil 231, 37–44.

Earl, K.D., Syers, J.K., McLaughlin, J.R., 1979. Origin of the

effect of citrate, tartrate and acetate on phosphate sorption

by soils and synthetic gels. Soil Sci. Soc. Am. J. 43, 674–678.

Greenaway, F.T., 1986. Aluminum-27 NMR study of alumi-

num (III) interactions with hydroxy carboxylic acids. Inorg.

Chem. Acta 116, L21–L23.

Hsu, P.H., 1988. Mechanisms of gibbsite crystallization from

partially neutralized aluminum chloride solutions. Clays

Clay Miner. 36, 25–30.

Hsu, P.H., 1997. Effect of temperature on the degradation of

Al13 complex. Clays Clay Miner. 45, 286–289.

Hsu, P.H., Cao, D., 1991. Effect of acidity and hydroxylamine

on the determination of aluminum with ferron. 1. Soil Sci.

152 (3), 210–219.

Huang, P.M., 1988. Ionic factors affecting aluminum transfor-

mations and the impact on soil and environmental sciences.

Adv. Soil Sci. 8, 1–78.

Huang, P.M., Wang, M.K., Kampf, N., Schulze, D.G., 2002.

Aluminum hydroxides. In: Dixon, J.B., Schulze, D.G.

(Eds.), Soil Mineralogy with Environmental Applications.

Soil Science Society of America, Inc. Book Series Number 7.

Soil Science Society of America, Madison, WI, pp. 261–290

(Chapter 8).

Hue, N.V., Craddock, G.R., Adams, F., 1986. Effect of organic

acids on aluminum toxicity in subsoils. Soil Sci. Soc. Am.

J. 50, 28–34.

Jardine, P.M., Zelazny, L.W., 1987. Influence of organic anions

on the speciation of mononuclear and polynuclear alumi-

num by ferron. Soil Sci. Soc. Am. J. 51, 885–889.

Karlik, S.J., Tarien, E., Elgavish, G.A., Eichhorn, G.L., 1982.

Aluminum nuclear magnetic resonance study of aluminu-

m(III) interactions with carboxylate ligands. Inorg. Chem.

22, 525–529.

Kodama, H., Schnitzner, M., 1980. Effect of fulvic acid on the

crystallization of aluminum hydroxides. Geoderma 24,

195–205.

Krishnamurti, G.S.R., Wang, M.K., Huang, P.M., 1999. Role

of tartaric acid in the inhibition of the formation of Al13tridecamer using sulfate precipitation of aluminum. Clays

Clay Miner. 47, 658–663.

Kwong, K.F., Kee, Ng., Huang, P.M., 1977. Influence of citric

acid on the hydrolytic reactions of aluminum. Soil Sci. Soc.

Am. J. 41, 692–697.

Kwong, K.F., Kee, Ng., Huang, P.M., 1979. The relative

influence of low-molecular-weight, complexing organic

acids on the hydrolysis and precipitation of aluminum. Soil

Sci. 128, 337–342.

Kwong, K.F., Kee, Ng., Huang, P.M., 1981. Comparison of the

influence of tannic acid and selected low molecular weight

organic acids on precipitation products of aluminum.

Geoderma 20, 179–193.

Masion, A., Bottero, J.Y., Thomas, F., Tchoubar, D., 1994a.

Chemistry and structure of Al(OH)/organic precipitates. A

small-angle X-ray scattering study. 2. Speciation and

structure of aggregates. Langmuir 10, 4349–4352.

Masion, A., Tchoubar, D., Bottero, J.Y., Thomas, F., Villieras,

F., 1994b. Chemistry and structure of Al(OH)/organic

precipitates. A small-angle X-ray scattering study. 1.

Numerical procedure for speciation from scattering curves.

Langmuir 10, 4344–4348.

Masion, A., Thomas, F., Tchoubar, D., Bottero, J.Y., Tekely,

P., 1994c. Chemistry and structure of Al(OH)/organic

precipitates. A small-angle X-ray scattering study. 3.

Depolymerization of the Al13 polycation by organic ligand.

Langmuir 10, 4353–4356.

MinteqA2 Chemical Speciation Program, 2000. Environmental

Protection Agency, Version 4.02, USA.

Motekaitis, R.J., Martell, A.E., 1983. Complexes of aluminu-

m(III) with hydroxy carboxylic acids. Inorg. Chem. 23, 18–23.

Parfitt, R.L., 1978. Anion adsorption by soils and soil

materials. Adv. Agron. 30, 1–50.

Parthasarathy, N., Buffle, J., 1985. Study of polymeric

aluminum(III) hydroxide solutions for speciation in waste

water treatment. Properties of the polymer and optimal

conditions of preparation. Water Res. 19, 25–36.

Robert, M., Berthelin, J., 1986. Role of biological and

biochemical factors in soil minerals weathering. In: Huang,

P.M., Schnitzer, M. (Eds.), Interaction of Soil Minerals with

Natural Organics and Microbes. Soil Science Society of

America, Special Publication, Number 17. Soil Science

Society of America, Madison, WI, pp. 453-496.

Rubin, A.L., Hayden, P.L., Columbus, O.H., 1973. Water

Resources Center. The Ohio State University, Project

Completion Report no. 364, pp. 66–91.

Singer, A., Huang, P.M., 1990. Effect of humic acid on the

crystallization of aluminum hydroxides. Clays Clay Miner.

38, 47–52.

Sohnel, O., Garside, J., 1992. Precipitation: Basic Principles and

Industrial Applications. Butterworth-Heinemann Ltd, Li-

nacre House, Jordan Hill, Oxford.

Teagarden, D.L., Kozolowski, J.F., White, J.L., Hem, S.L.,

1981. Aluminum chlorohydrate. 1. Structure studies.

J. Pharm. Sci. 70, 758–761.

Thomas, F., Masion, A., Bottero, J.Y., Rouiller, J., Genevrier,

F., Boudot, D., 1991. Aluminum(III) speciation with acetate

and oxalate. A potentiometric and 27Al NMR study.

Environ. Sci. Technol. 25, 1553–1559.

Thomas, F., Masion, A., Bottero, J.Y., Rouiller, J., Genevrier,

F., 1993. Aluminum(III) speciation with hydroxy carboxylic

acids. 27Al NMR study. Environ. Sci. Technol. 27,

2511–2516.

Tsai, P.P., Hsu, P.H., 1984. Studies of aged OH–Al solutions

using kinetics of Al-ferron reactions and sulfate precipita-

tion. Soil Sci. Soc. Am. J. 48, 59–65.

Tsai, P.P., Hsu, P.H., 1985. Aging of partially neutralized

aluminum solutions of sodium hydroxide/aluminum molar

ratio ¼ 2.2. Soil Sci. Soc. Am. J. 49, 1060–1065.

Wang, W.Z., Hsu, P.H., 1994. The nature of polynuclear

OH–Al complexes in laboratory-hydrolyzed and commer-

cial hydroxyaluminum solutions. Clays Clay Miner. 42,

356–368.

Wang, S.L., Wang, M.K., Tzou, Y.M., 2003. Effect of

temperature on formation and transformation of hydrolytic

aluminum in aqueous solutions. Colloid Surf. A: Physico-

chem. Eng. Aspects 231, 143–157.

Wu, C.W., Wang, M.K., 2000. Effect of acetate, oxalate, and

citrate anions on the formation of Al13 and aluminum

hydroxide in partially neutralized aqueous aluminum salt

system. Food Sci. Agri. Chem. 2, 15–19.