NOTE TO USERS

This reproduction is the best copy available

UMI

INVESTIGATIONS ON THE CHEMISTRY OF PHOSPHORUS

AND ORGANIC MATTER REMOVAL DURING

WASTEWATER TREATMENT

Anselrn ldernudia Omoike

A thesis submitted to the Department of Chemistry

in conformity with the requirements for

the degree of Doctor of Philosophy

Queen's University

Kingston, Ontario, Canada

July, 1999,

Copyright Q Anselrn Idemudia Omoike

National Library I*I of Canada Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services services bibliographiques

395 Wellington Street 395. rue Wellington Ottaw ON KiAON4 Ottawa ON K1A ON4 Canada Canada

The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distrï'bute or sell copies of this thesis in microform, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it m2y be p ~ t e d or otheNvise reproduced without the author's permission.

L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, districbuer ou vendre des copies de cette thèse sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent ê e imprimés ou autrement reproduits sans son autorisation-

ABSTRACT

Phosphorus is a growth-limiting plant nutrient that can cause eutrophication of

water bodies- Due to growing concern over the problem of eutrophication, especiaily in

ecologically fragile areas, stnngent effluent Iimits for phosphorus have been established

in different jurisdictions. As one means used to meet these Iimits. chemical coagulants

are used during wastewater treatment and the coagulant can be added at various points

in the treatment plant Depending on the design, the dissolved organic matter (DOM)

concentration is variable in the wastewater at these different points- The effect of

variation in DOM concentration on phosphorus removal and on the nature of the solid

product is still not cleariy understood,

In order to contribute to a better understanding of the interactions between

aluminurn, DOM and phosphonis, coprecipitation and postprecipitation expenments

were conducted to simulate various wastewater treatment processes. The solids that

were precipitated from solution were characterized using ferron reagent, Fourier

transfomi infrared s pectroscop y, atomic force microscop y (AFM) , interfacial force

microscopy (IFM) and solid-state aluminum nuclear magnetic resonance (27 Al NMR)

spectroscopy. In addition, residual phosphonis and tannic acid (used as a surrogate for

organic matter) concentrations in the filtrates were deterrnined.

Alurninurn 27 NMR spectroscopic data showed that the solids contain

predominantly skcoordinated aluminum. In the absence of tannic acid , phosphonis

was removed to a greater extent when coprecipitated rather than postprecipitated ont0

prehydrolyzed solid. In contrast, postprecipitation pmcesses removed a greater fraction

of tannic acid, Where both phosphonis and tmnic acid were present, there was a

cornpetition between them, interîering with the rernoval of both cornponents, Tannic acid

appeared to have a larger inhibitory effect on phosphorus than did phosphonis on

tannic acid,

The surface reacüvity of solids obtained from postprecipitation experiments

reacted much more slowiy with ferron reagents compared to solids obtained from

coprecipitation experiments, A mode1 in which the postprecipitation particles are coated

with an organic layer of tannic acid was proposed and verified using AFM and IFM

techniques. The observations from AFM and IFM data are consistent with this mode[-

Results suggest that it is advantageous to add at least a portion of the alum ai

the exit of the aeratoc This enhances phosphonis removal by wprecipitation under

conditions where the concentration of organic matter is relatively low and it enhances

removal of organic rnatter by postprecipitation onto the recycled sludge in the aerator.

iii

In loving memory

of

rny late rnother,

Alice Omoike

Acknowledgement

My sincere appreciaüon to Dr. G. W. vanloon for his friendship, support and

guidance during the penod of my program at Queen's University, Kingston-

I wish to thank al1 those in the department of Chemistry and other departrnents

who have helped me in vatious ways during rny program. Special thanks to Dr. H.

Horton for his assistanœ with the acquisition of Me IFM data and for his scanning probe

expertise. I also thank Dr. G. Chen, Ms. Sue Blake and Mr. Dave Kempson (Geological

Sciences department) for their assistance in the acquisition of the AFM images;

aluminum MAS NMR spectra and the microprobe analysis data respectively. l would

also like to thank Mr. P. Mulligan. Mr. S. Meskis and Mr. D. Axelson for their help during

my program.

To my colleagues and friends in Dr vanLoonls research faboratory and the

Analytical Services Unit I Say a big 'thank you' for the fnendship, fun and good-spirit we

al1 shared.

1 wish to express my special appreciation to al1 the parishioners of St Thomas

More Parish and members of Servants of the Light Prayer Group for the spiritual support

and friendship. Special thanks to Dr. vanloon and family, Dr. S. Hesp, Dr. Stephen

Duffy, Ohi Izirein and family. Bill Reason and Fatima Dias for making me feel at home in

Kingston,

I gratekrlly acknowledge the financiai assistance the University and department

provided during my program.

Table of Content

Title page.-. ... ------ --- ---------.-- -----. --. -.---- ----.-.---.. .-- .-- .-.--. .--- Abstract,,, . .- ,.. ,,, . , , .- - --- - .- --. --- - .- --- --. --. . . . . -. --- -. - - -- - - - -- - - -. -- - . - - - Dedication,. . . , , . . - - . - . - - -. - , - - . . . - -. . - - . . . - - . . - - - - - - - - - - - . - - - . - - -- - - - -. - -. . . . . . Acknowledgernents- -- -. . . - -.. C C - -- - -- - -- - -- - -- - -- - - - - - - - --- --- - - - -- - - - - - Table of Contents--. - - - --. -.- -- - --- --- --- --- - - - --. --- .-- --- --. -.- --- --- - - - ---- List of Tables .-. -., -.- .-- .-- ..- . .- ---. -. ... ..- -.. --- .-- .-. -.- .. - .. . .-- -.. ... - List of Figures.-- -., ,.- --- -.- --- . .. -.- ... -.- ..- --- ..- --. -. - - - - --- .-- --- -.- -- - .- Abbreviations,.. .,. . - .. - .- -. . . .. . .- . .- .., -.. -. . .-- -.- --. - -. --. -. . . -. - - - - -. . . - .. . Glossary of terms--- - -. - -. . -- . . . .. . - . . .-. . .. . .. --. .. - .. . - .- -.. ..- .. . . . - -- - - .- Statement of original contribution .,. --- -.. ,.. .-. ... .-. ..- ... --- ... ..

1 - Introduction, -. - -- . - . - - - - - -. - -. - - . - -- . . . . . -. . -. . . . -. . - . - - -. - . - -. . - - - . . - . . .

Wastewater. - - - - . - - . . -. - - - - . . - -. - . . . . - . - . - - - - . . - - - - . - . . . . - . - . . - - . - - . - Municipal wastewater quality parameters. .. -.. .. . . .. . .. ... . The composition of municipal wastewater..- ... ... .. . ... .-. . Wastewater treatment processes.-, . .. -.. ..- . .- -. - . . . . -- .-. .-- 7.4.1 Preliminary treatment -.. -.. ... -.. .-- -.- -.- - - a a.- -.. --a .. 1 -4.2 Primary treatment. -. .. , , ., , . , . . , .. . ._ - - -. .. . .-- . . . . . . -. - . 1.4.3 Secondary or biotogical treatmenL - - - --. .,- --- .-. - 1.4.4 Tertiary / Advanced treatment . . . . . - - - . . . . . - . . - . . . . - . Aluminum coagulation chemistry,,. ... _-- -.- .-. .-. --- .-. --. .-. Phosphorus and organic matter in municipal wastewater-

1.6.4. Phosphorus forms ... ,-. -.- . .,..- -.. ... ,-. -.. ... ... --. .-. 1.6.2. Reasons for phosphows control in wastewater-.

1.6.3. Phosphonis control in wastewater ... ..-...-.. ... --.. 1 -6.4. Mechanism of phosphorus removal--- .-- .- - *-. --. .-- - 1.6.5. Organic matter in wastewater ... ..- .. - ..- -.- ... .. . ..- ... 1-6.6. Mechanism of organic removal ... .-. ... .-. .-- ... ... -..

i

ii

iv

v

vi

ix

xi

m- xxiii

xxiv

1

1

3

7

9

9

9

11

14

15

20

20

22

22

27

31

32

1.6.7. Tannic acid as surrogate for organic matter in wastewater.. 33

Materials. ,- . . - . . . . -. . - - . -. . -. . - - -- - -- - - . - . - - -- - . - - - . - -- - - - - - - - -- - --- -- Simulated wastewater experiment--- . .. --- --- --- --- --- --. .-. -. Municipal wastewater and recycled sludge sample ... -. . . pH -. - -. - . - . - - - - - - . . . . . . -. - . . . . . - - . - - - - - - - - -. . - . . - - - - - - - - - - . . - - - - . . . . Dissolved organic carbon (DOC) determination of untreated

and treated wastewater,,. --. .-- .-. ... -- - .. . ..- -.- .-. -.. -- - ..- ..- - Kinetic experiments,. . -. , . .. ... .-. .., .-. ._- -. . . .- . -. -.- --- --. -.- .-. Residual phosphoms determination,.. --. --. --. -.. -.- --. -.- ... Residual aluminum determination--- -.- .-- -.- ... .-- ... --- --. ...

Method development for residual TA determination.. . . . . Fourier transfomi infrared spectroscopy (FTIR) analysis.

Atomic force microscopy (AFM) anaiysis-.. --. ..- . .. . .. ... .- lnterfacial force microscopy (IFM) ,,. -,- ... .-. .-- .-. .-- - - - .-- - Solid-state *?AI Magic angle spinning nuclear magnetic

resonance (MAS NMR) ..-... ... ... ... -.. .-. .-. ..-.-. .-. ... ... ... Electron microprobe analysis - - -. . . . - - . - . . . - - - -. - - -. . . . . - - - . . -

3.0. Results and Discussions ... ... --. -.. --. ..- --. .-- .-. .-. --. .-- -- - ..

31. The Ferron test and Fourier transfom infrared (FTIR)

spectroscopy.. . . . - ,. . . . . .-, , -- - .- - - . - -. . . . - -- . - - . . . . -. . . . . . . . . . . . - . . . 3-1.1 , Coprecipitation studies --* -. - --. --- -- - .. . . .- .. . -. . - -. . . . 3.1 -2. Postprecipitation studies . .- .-. . . . -.. . .. . -. -.- -. . -. . ..- .- 3.1 -3. Residual phosphorus and tannic acid.-. ... ... ... ... 3.1 -4. Summa y... ... ... ... ... -.. ... .-- ... .-- ..- ....- --. .-. ... ... .

vii

3.2. Atomic force microscopy (AFM) and interfacial force

microscop y (1 FM) studies: Evidence for organic coatings on

postprecipitation productç .., .,, -,- --. .-- --- .. - --- -.. .-. . .- ..- - 104

3.2.1 .. AFM studies on hydrous aluminum oxides precipitated

under vanous wastewater conditions--- --- --- -.+ ._- - 104

3-2-1 -1 - Surnmary.-. . .. --- .-- .-- ... --- .-- --- ..- . 1 36

3.2.2- IFM studies on hydrous aluminum oxides precipitated

under various wastewater conditions --- .-- --- -.- --. 3-2.2.1. Summary.. .. . . .- - .- .. . -.- .-. ... . .- .. - . .-

3.3. 2 7 ~ ~ MAS solid-state NMR spectroscopy --. .-. --- .-- .-- -.. ... 3.3.1. Summary.-- .., -.. .., --. -,- ... -.. -,, --. .-- --... ..- ..- -.. ... ..

3.4. Studies involving municipal wastewater and recycled sludge

samples.-. ... ..- ... ... .-- -.. ..- ... .-- .---..-. .-- ..- .-- --. ..- --. --- .-.

3.5. General conclusion.. . . ,. . .. - , . . -. , . . - . . -- - - -. . - - . -. . . - . .. - .. -. . .

4.0. Referenœs ..- .. . . .. -.- -. - ... .-- ... .-- ... .. . .. . -- --. --.. --. ..- ... ... -. .

viii

List of Tables

Table

Chernical contaminants of conœm in wastewater treatrnent-.. ..

Typical composition of untreafed municipal wastewater- ..........

A cornparison of phosphorus removal performance of three

primary treatment plants in Ontario, Canada with and

............... .................... without coagulant addition (FeC t3) ,.,

Cornparison of effluent quality obtained from the treatrnent of

influent with and without alum addition in an adivated sludge

............................................................................. process

Page

2

8

I O

13

.. The horganic phosphate species found in municipal wastewater. 21

Foms of organic phosphates found in municipal wastewater. ...

Coagulants commonly used for phosphorus removal in

...................................................... municipal wastewater

Analyücal features of the spectrophotometric method

.................................................. for the determination of TA

Operating conditions for electron microprobe data on

.............................. select hydrous aluminum oxide species

Percent fast reacthg aluminum species (FR) and values

obtained from ferron tests on coprecipitation systems

after different aging periods ................................................

Percent fast reactnrg alurninum species (FR) and values

obtained from ferron test on postprecipitation systems

.......................... ................... after different aging perioâs ....

Young's modulus and Poisson's data.,- --- --- .-- --. -,- ,,, ,-, --, ,,- ---,

2 7 ~ ~ MAS NMR chemical shifts and peak analysis results

for Al alone system--. .....-..................................................-.

2 7 ~ ~ MAS NMR chemical shifts and peak analysis

results for AIP systern--- ..................................................

''AI MAS NMR chemical shifts and peak analysis results for

AITA sysfern,,. .................................................................

2 7 ~ ~ MAS NMR chemical shifts and peak analysis results

for AISTA system ....................... -... ...................................

Z 7 ~ ~ MAS NMR chemical shifts and peakanalysis results

for AlPTA system.. ............ ,. .............................................

2 7 ~ ~ MAS NMR chemical shifts and peak analysis results

for AISPTA system,.. ..........................................................

Dissolved organic carbon and phosphorus concentrations of

wastewater samples obtained frorn an acu'vated sludge plant ...

Absorption spedra of sludge sampfes., ....-............................

Percent fast reacting aluminum species (FR) and values

obtained from ferron test on coprecipitation system (AIUS)

using municipal wastewater after different aging periods.. ........

Percent fast teacting aluminum species (FR) and fso values

O btained from ferron test on postprecipitaüon system (A15US)

using municipal wastewater after different aging periods,-. ........

List of Figures

Figures

Flow diagram showing the basic units in a prirnary treatment

Plant- .- --. - . -. - -. . . - --. . . -. -.-. - -. - -- -- --- -- - - - - - -- - - -,- - ,-- - -,. - - - -- - -. - - -. . . -- - -- Schematic diagram showing the basic units in a secondary or

(Biological) treatment plant--- .--- ---. --------.,,--.,,,.,- ,,-, ,,- -- - - - - - -_--

Solubility of aluminum as a krnction of pH, in the absence of

complexing agents other than hydroxide ions.-. -.-. _..- ..-- -.-. .-cc.

Alternative points (1, 2 and 3) of aluni addition in an

activateci sludge plant ... ---. ...- .-.. .--. ...- .... .... .,.. -., ,.. -,.- .--. -.-. ---.

The basic units in tannic acid .-- ..-- ..-- -- - - ..,. .,,. ..., .,.. --.. --. --. .-. ..

The structural formula of tannic acid,,, ,..,..-. ..,,..,-,-.- ,.,-,-,, -... ,-.

Schematics of Atomic Force Microscopy ....,., ,-.. ,... ,,., ,--- ---. .-.. .

Absorbance versus time plot for the differential reaction

of aluminum species with ferron ..,..,....-........ .....--..-...... ... .-., ,.

Calibration plot for phosphorus determination-., ,.-. .,,, .,.. --,, . --- -

Absorption spectra of TA (17.0 mg L-') in (a) distilled

de-ionized water (pH 4.5) and (b) aikaline sodium bicarbonate

solution (pH 8.4) ..............................................................

Absorption spectra for TA (5.0 ppm): Influence of pH of the

solution -. - -. - - . . - - -- -. - - . - - - - - - - - - -. - - . . -. - - - . . - . - - - -. . - -. - - . . - - - - -. , -. - . . . - -. -. --

Absorption spectra for filtrate obtained using 0.45 um

membrane filters from 30 minutes aged sarnples of AlTA

and AIPTA systems..- --,- .. -- .--- . .-- .- .- . --- . .-. ., .. ,.. ,. . , ,. . -. .-. --. . . .- .--

Absorption spectra of TA at hm different pH values ... ... ... ... ..-.

Page

11

14

18

26

34

34

42

50

51

53

54

56

57

Absorption spectra of a mucture of TA and aluminum ions at

two different pH values ........................................................ 57

Calibration curve obtained from standard TA solutions

............................................. acidified to a pH value of 2.0.

F-D curve of a surface. On the cuwe, 1 is the elastic

loading region (increasing contact area), 2 is both elastic

and plastic deformation region, 3 is the elastic unloading

region (constant contact area) and 4 is the region of

.......................... probe withdrawal (decreasing contact area)

Fourier transfom infrared (FTIR) spectnirn of the 'simple'

hydrous aluminum oxide solid (A15) in the region

...................................... . 4000 400 cm4 [Al] = 430 mg L*'

Fourier transfomi infrared (FTIR) spectrum of solid sample

obtained from AIP system (AIP5) in the region

- ................. 4000 400 cm? [Al] = 430 mg 1-'; [Pl = 500 mg L-'

Fourier transform infrared (FTIR) spectrum of waveflite in

- ............................................ the region 4000 400 cm- '.......

Fourier transfom infrared (FTIR) spectrum of tannic acid

- ............................................. In the region 4000 400 cm-'

Fourier transfom infrared (FTIR) spectnim of solid sample

obtained fmm AlTA system (AITAS) in the region

- ............. 4000 400 cm-' [Al] = 430 mg L-'; = 1700 mg L-'

Absorbance versus time for the ferron test on AIPTA solids

(120 min aged) before and after filtration using membrane

......................... filters.. ,_,. ...............................................

xii

Influence of increasing of TA concentration on surface reacüvity

...... of solids obtained from AlPTA systems with ferron reagent

Fourier transfomi infrared (FTIR) spectrum of solid sample

obtained from AlPTA system (AIPTAS) in the region

4000 - 400 cm-'. [Al] = 430 mg L-'; [Pl = 500 mg L?;

F A ] = 1700 mg L?.. ....................................................

Fourier tfansform infrared (FTIR) spectrum of solid sample

obtained from A15P system (Al5P5) in the reg ion

................ . 4000 - 400 cm-' [Al] = 430 mg L*'; [Pl = 500 mg L-'

Influence of long-tem penod agîng on surface reactivity of

AISTA solids, .....................................................................

Fourier transform infrared (FTIR) spectrum of solid sample

obtained from AISTA system (A15TA5) in the region

.......... . 4000 400 cm-', [Al] = 430 mg L-'; [TA] = 1700 mg L-'

Absorbante versus time for the ferron test on AISPTA solids

(1 5 h aged) before and after filtration using membrane

filters. ~.......................................-...........-.~~.....................

Wavelength scans for filtrates obtained from AISPTA

.............................. and AIPTA systems aged for 90 minutes

Companson of wavelength scan for a mixture of TA and

phosphate with that for a solution of TA alone: (a) TA (5.0 mg L")

...... alone and (b) Phosphate (5.0 mg c') and TA (1 7.0 mg L-l)

Compatison of surface reactivity of hydrous aluminum oxide

precipitated in the presence of phosphate or TA. Times shown

are the time of aging of the system after precipitation,., ,., ..,,.,.,

Cornparison of surface reactivity of hydrous aluminum oxide

precipitated in the presence of phosphate and tannic acid- Times

show are the trmes of aging of the system after precipitation-..

Phosphorus rernovai ( O h ) in coprecipitation systems- ln the

case of AIPSTA, phosphorus is removed by coprecipitation.

In the case of AIPSTA, TA was added after phosphonis had

been removed by coprecipitation of phosphate .....................

Phosphorus removal (%) in postprecipitation systems- .--

Tannic acid removai ( % ) in coprecipitation systems. In the

case of AlTASP,phosphate was added after TA had been

removed by postprecipitati0n.-.. .............................. ... ... ... .

Tannic acid removal (%) in postprecipitation systerns-.. ..........

Influence of increasing TA concentration on phosphonis

removal: AlPTA system, aged for 30 min ..............................

AFM images of TA adsorbed on a mica substrate.

fieight (left) (2 range = 25 nm) and phase imaging (right).

The image is 2.0 pm square and was acquired

at A, = 106 nm, and r, = 0.90. Aggregate particles of

various dimensions can be seen in the height image-.. ..........

AFM images of TA adsorbed on a mica substrate. Height (left)

(Z range = 10 nm) and phase imaging (right). The image

is 500 nm square and was acquired at

(a) A. = 106 nm, and r., = 0.90. and (b) A, = 106 nm,

.................................................................. and rs, = 0.30

(c) A. = 52 nm, and r, = 0.90. and (d) A. = 26 nm,

and r,, = 0.90 ..............-.................... .... ............................

Cross sectional profile analysis data for TA ........................

xiv

37. AFM images of the control particles dispersed on a mica

substrate. Height (left) (2 range = 200 nm) and phase

imaging (right) The images are 2 pm square and were

.......................... ....... acquired at A, = 58 nm and rsp 0-50 .... 112

38 a. Cross sedional profile for particles from Al system used as

control, showing that the height of the particle highlighted

............. on the image is 24.0 nm, ... .............................. .,. 113

38 b. Cross sectional profile for particles from Al system used as

control. showing that the height of the particle highlighted

.... ............................................ on the image is 124.3 nm ., 114

39. AFM images of coprecipitated particles adsorbed on a

mica substrate. Height (left) (Z range = 200 nm) and phase

imaging (right)- The images are 2 pm square and were

acquired at :(a) & = 116 nm, and r, = 0.65; and

.............................................. (b)Ao=58nm,andrs,=0.65 116

............................................... (c) A,, = 29 nm, and r, = 0.65 117

40 a. Cross sedional profile for CO precipitated Al PTA particles

showing that the height of the particle highlighted on the

image is 54.6 nm-.. ......................................................... 118

40 b. Cross section profile for coprecipitated AlPTA parücles

showing that the height of the particle highlighted on the

image is 81 -5 nm. .......................................................... 1 19

41. AFM images of the postprecipitated hydrous aluminum

oxide pa;ticles (A15PTA5) dispersed on a mica substrate.

Height (left) (2 range = 20 nm) and phase imaging (right).

Images are 2 prn square and were acquired ai A, of 104 nm

and at the following rs, (a), 0.90 and (b) 0.6L. ..................... 122

42. AFM images of the postprecipitated hydrous aluminum oxide

particles (AISPTAS) dispersed on a mica substrate. Heig ht (left)

(2 range = 20 nm) and phase imaging (right) Images are 2 pm

square and were acquired ai A, of 46 nrn and ai the following

.......... ........................................ r,,: (a)O.Wand (b)0-75 ,. 123

43. AFM images of the postprecipitated A15PTA5 dispersed.

on a mica substrate Height (left) (2 range = 20 nm) and

phase imaging (right). lmages are 2 pn square and were

acquired at A, of 22 nm and at the following r, (a). 0.90

................................................................... and (b)0.70 124

44 a. Cross sectional profile of postprecipitated AISPTA particles

showing that the height of the highlighted particle on the image

44 b. Cross sectional profile of postprecipitated AISPTA particles.

showing that the height of the highlighted particle on the image

is 2-1 nm- ..........~.............-....~......~...~.........~................... 128

45. AFM images of the postprecipitated AISPTA particles

(1 hour aged) dispersed on a mica substrate. Height (left)

(Z range = 70 nm) and phase imaging (right). lmages are

5 pm square and were acquired at A. of 1.30 nm and at r,,

of0.88 ..................-......................................................... 131

46. AFM images of the postprecipitated AISPTA particles

(1 hou? aged) dispersed on a mica substrate- Height (left)

and phase imaging (right). The 2 pm square images are

images of different ragions of the sample and were acquired

.... under identical conditions. (Z range = 50 nm, high & and rsp)

AFM images of the postprecipitated AISPTA particles

(24 hour aged) dispersed on a mica substrate, Height

(left) (2 range = 40 nm) and phase imaging (nght)- lmages

are 5 pm square and were aquired at A, of 1-12 nm and

at r,, of 0-92 ................. --.. ...............................-................

AFM images of the postprecipitated AISPTA particles

(24 hour aged) dispersed on a mica substrate, Height

(left) (2 range = 40 nm) and phase imaging (rïght). lmages

are 5 pm square and were acquired at & of 1-12 nm and

at r,, of 0.92 ............................................................ .. .......

AFM images of the postprecipitated AISPTA partictes

(24 hour aged) dispersed on a mica substrate. Height (left)

(2 range = 30 nm) and phase imaging (nght)- Images are

2 prn square and were acquired at A, of 4-12 nm and at

r of0-92 ........................................................................ SP

Contact mode IFM image of (a) freshly cleaved mica

surface (6 prn x 6pm), (b) control particles (6 pm x 6pm)-

(b) coprecipitation particles (1 2 prn x 12 pm ) and

..................... (d) postprecipitation particles (12 pm x 12 pm )

F-D curve obtained from indenting freshiy cleaved mica surface..

F-D curve obtained from ïndenting control partides on a mica

substrate.,, ....................................................................

F-D curve obtained from indenting coprecipitated AlPTA

............................................. particles on a mica substrate.

F-D cuwe obtained from indenting postprecipitated

AISPTA particles on a mica substrate ................................... 145

F-D cuwe obtained from indenting postprecipitated AISPTA

partides on a mica substrate, In this enlarged figure two distinct

slopes indicating that the particles consist of a bilayer is shown.. 146

F-D curve obtained from indenting postprecipitaed AISPTA

particles on a mica substrate showing the Hertzian ffi

for the organic layer ............................................,.....,....... 147

F-0 cuwe showing the Hertzian fit for the inner core of

the postprecipitated AISPTA particles .................................. 148

'?AI MAS NMR spectnim of hydrous aluminum om'de precipitated

in the presence of alkatinity and aged for 5 minutes (A15):

[Ag = 430 mg L-l ............................................................ 152

2 7 ~ 1 MAS NMR spectfum of hydrous aluminum oxide precipitated

in the presence of alkalinity and aged for 30 days (A143290):

................................................................ [AI] = 430 mg L-'

2 7 ~ ~ MAS NMR spectnirn of A143200 showing the resolved

............... aluminum coordination signals: [AI] = 430 mg L?..

'?AI MAS NMR spectnim of solids obtained from the

coprecipitation of aluminum and phosphate aged for 5 minutes

(AIP5). [Al] = 430 mg L-l: [Pl = 500 mg L-' .................. .... .......

'?AI MAS NMR spectnrm of solids obtained fmm the

coprecipitation of aluminum and phosphate aged for 43200

......... minutes (AIP43200). [AI] = 430 mg L-l: [PI = 500 mg L-'

2 7 ~ ~ MAS NMR spectrum of solids obtained from

postprecipitation of prehydrolyzed aluminum and phosphate

aged for 5 minutes (A15P5). [Al] = 430 mg L-': [Pl = 500 mg L?..

2 7 ~ ~ MAS NMR spednim of solids obtained from

coprecipitation of aluminum and TA aged for 5 minutes

(AITA5). [AI] = 430 mg L-': [TA] = 1700 mg L-' ......................

2 7 ~ ~ MAS NMR spectnim of solids obtained from

coprecipitation of aluminum and TA aged for 3 days

(AlTA43200). [Al] = 430 mg L-l: F A ] = 1700 mg 1- '... .............

MAS NMR spectrum of solids obtained fmrn

postprecipitation of prehydrolyzed aluminum and TA

aged for 5 minutes (A15TA5). [AI] = 430 mg 1-'

FA] = 1700 mg L-' ...........................................................

2 7 ~ ~ MAS NMR spectnim of solids obtained from

postprecipitation of prehydrolyzed aluminurn and

TA aged for 30 days (A15TA43200). [Al] = 430 mg L"

FA] = 1700 mg L-' ............................................................

2 7 ~ 1 MAS NMR spectnim of solids obtained from the

coprecipitation of aluminurn, phosphate and TA aged for

minutes (AIPTAS). [Ai] = 430 mg L-', [Pl = 500 mg L-l

[TA] = 1700 mg CI. ...........................................................

2 7 ~ ~ MAS NMR spectrum of solids obtained from

coprecipitation of aluminum, phosphate and TA aged for

30 days (AIPTA43200). [Al] = 430 mg [Pl = 500 mg L-'

[TA] = 1700 mg L-' ............................................................

2 7 ~ 1 MAS NMR spectnim of solids obtained from aging

prehydrolyzed aluminum in the presenœ of a mixture of

xix

phosphate and TA for 5 minutes (A15PTA5). [Al] = 430 mg L-':

[ P ] = 5 0 0 m g ~ ~ ' ~ A ] = 1 7 0 O m g ~ " ...............-.................... 170

59 b. 2 7 ~ 1 MAS NMR spectmm of solids obtained from aging

prehydrolyzed aluminum in the presence of a mixture of

phosphate and TA for 43200 minutes (A15PTA43200).

[ Al] = 430 mg [Pl = 500 mg L-' UA] = 1700 mg L-' ...... ... . 1 70

60. FTlR spectnim of dried recycled sludge obtained from Kingston

West Sewage Treatrnent Plant, Ontario, Canada-.. -.. ..- ... .. -.. .. 175

61 2 7 ~ ~ MAS NMR spectmm of recycled sfudge obtained from

Kingston Township Water pollution Control Plant,

Ontario, Canada.,- ,... ...- ,,,--,,. ,,.-.,,- ...,-..--....-.-.... .-. -.- ..- ..-,,. ,, 179

62 a. Backscattered electron image of AIPTA system aged for 5 min.. 199

62 b. Elemental distribution of alurninum and phosphorus in AIPTA

particles aged for 5 min-.. ,.. ,.. .., .. . . ,. - -- . .- -.- --. --. .-. ... -.. .-. -.. .,. . 199

64a. Backscattered electron imageofAI5PTA partides agedfor5 min ... 200

64 b. Elemental distribution of aluminum and phosphorus in AISPTA

particles aged for 5 min.. . .. . . . . . .. . . . - .. --. . . . .. . . . . . .. . . . ... . . . .. . ..,. -. .. 200

AFM

A0

As,

AIPTA

AISPTA

d

DDW

DOC

DOM

FTlR

GPa

h

IFM

MAS NMR

nrn

PAHs

PCBs

PPm

~ S P

S

TA

Abbreviations

atomic force microscopy

free oscillation amplitude of AFM probe

setpoint amplitude of AFM probe

hydrous aluminurn oxide species formed by rnixing alum

(Al), phosphate (P) and tannic acid (TA) in the presence of

alkalinity

hydrous aluminum oxide species fomed by aging a

mixture containing alkalinity and alum (Al) for 5 min

followed by addition of a mixture of phosphate (P) and

tannic acid (TA)

hydrous aluminum oxide fomed by mùa'ng alum and

untreated wastewater

hydrous aluminum oxide fomed by mixing prehydrolyzed

alum aged for 5 min before mixing with untreated

wastewater

day distilled deionized water

dissolve organic carbon

dissolved organic matter

foufier transform infrared

giga pascal

hour

interfacial force microscopy

mag ic angle spinning nuclear magnetic resonance

nanometer

polyaromatic hydrocarbon

polychlorinated biphenyl

parts per million (mg L -')

setpoint ratio (setpoint)

second

tannic acid

xxi

Pm

PN UV1 vis

time taken to recover 50% of spiked aluminum dunng the

analysis of alumhurn species in aged solutions using

fenon reagent

micrometer

micro Newton

ultra-violet / visible

Glossary of tenns

Agglomerates- a group of relatively large floc partides formed from srnaller floc

particles

Al system- hydrous aluminum oxide precipitated in the presence of alkalinity alone

Amorphous- terni used forfreshly precipitated hydrous aluminum oxide specles

Amplitude- maximum displacement during the oscillatory motion of AFM probe

Coprecipitation- refers to the precipitation of a normaily soluble camponent in the

absence of an initial solid phase,

Downfield- terni used when a resonance signal in an NMR spectnrm occurs at higher

frequency or lower applied field than an arbitrarily selected reference

Effluent- wastewater from a treatment plant discharged from into surface waters

Eutrophication- a phenornenon causing bioiogical, chernical and physical changes in a

water body due to its enrichment with nutn'ents such as phosphorus and nitrogen.

Influent- wastewater flowing into a treatment plant

Postprecipitaion-refers to a precipitation reaction in the presence of an initial solid

phase.

Protonation-an acid-base reaction dunng which proton(s) are accepted-

Receiving water- a water body such as lake, strearn or river into which wastewater

effluent is discharged

Setpoint ratio = seuoint amplitude of AFM robe (A ,a free oscillation amplitude of AFM probe (A&

Upfield- terni used when a resonance signal in an NMR spectrum occurs at lower

frequency or higher applied field than an arbitrarily selected reference

Statement of original contributions

The contributions of this work are as follow:

1. The influence of presence of organic matter at the points at which alum are

introduced into wastewater during the treatment process was investigated using

simulated wastewater and the solid phases were characterized using different

analyücal rnethods-

2. Tannic acid, as a surrogate for natural organic matter, was shown to inhibit

phosphorus removal from wastewater

3. A method was developed for the detemination of tannic acid in water samples

containing residual alurninum-

4. A mode1 based on the formation of an organic layer of tannic acid on solids

generated when prehydrolyzed aluminum is used to treat wastewater was verified

using atornic and interfacial force microscopy.

5- The elastic modulus of an organic layer of tannic acid and solids precipitated under

wastewater conditions were deterrnined using interfacial force microscopy.

xxiv

1 .O. Introduction

1 1. Wastewater

Wastewater is a cornplex aqueous mixture of soluble inorganic and organic

substances and suspended solids, induding a large population of various

microorganisms. Depending on source. the@ are many types of wastewater. Some

important categones include:

Municipal wastewater. This type of wastewater originates from households and

from activities related to the support of commercial and institutional facilities.

Industrial wastewater. This terni is used for water-borne wastes from a variety of

industries. These wastes may inciude agricultural by-pmducts, biomass residues,

metals, and organic and inorganic chemicals.

Infiltration wastewater. This is water which percolates downward thmugh the soi1

during or after precipitation and enters the sewer system-

lnflow wastewater. lnflow wastewater is mnoff from residential areas and

agricultural sites, resulting from precipitation or snowrnelt It is discharged into

sewers from stom drain connections, manhole avers, roof drains and outdoor

paved areas. The composition of this type of wastewater includes silt and sediments

from land erosion, sali and other deicing compounds, road dust and petroleum

products, fertilizers and pesticides.

The main focus of the work descnbed in the present thesis is on municipal

wastewater, It is important to note however that in some locations, municipal wastewater

may be mixed with inflow and industrial wastewater '.

Over the past decades, there has been growing concem about water quality,

especially streams and nvers receiving wastewater effluents, The concem has

generated both fundamental and applied research that is directed towards improving the

design and operation of wastewater treafrnent plants- In modem communities, the

discharge of municipal wastewater is regulated by govemment agencies and the effluent

requirements are based on water quality standards set to iimit the concentration of

contaminants entering receiving water bodies. Table I shows some contaminants found

in municipal wastewater and reasons why they are of concem.

Table 1 . Chernical contaminants of conœm in wastewater treatment-

Contaminants Reason for concern

N utdents (Nitrogen and phosphorus in particular)

Dissolved inorganics ( e-g. fluotide, arsenic, iron, cyanide, selenium, and de- icing salts) and trace metals (e-g. lead, cadmium, chromium, manganese, and mercury)

Biodegradable organics

Specific organic compounds (Man y corn pounds, including hydrocarbons, chlorinateci solvents, pesticides, PAHs and PCBs)

Nutrients support excessive growth of undesirable aquatic life leading to eutrophication-

Some of these are essential nutn'ents, but in excessive concentrations may be toxic to living organisrns. Trace metais biomagnify in the food chah and can be toxic and / or carcinogenic.

Causes oxygen depletion in the reœiving water leading to anaerobic conditions, causing stress to many types of desirable aquatic fauna,

Depending on the wmpound, they may contribute to carcinogenicity, mutagenicity or other forms of toxicity. Some of these compounds may produœ objectionable taste and odor-

1 2 Municipal wastewater quality parametem

Municipal wastewater has varying charactefistics but there are a number of

quality parameters that are measured in order to idenüfy the physical, chemical and

biological characteristics of the water *. The results from the measurements are

em p Io yed in the design of wastewater treatment facilities-

Measured phy sical characteristics of wastewater include co lor, odor, suspended

solids, and temperature. For aesthetic reasons as well as because of potential heaith

concerns, it is necessary to remove compounds that produœ color, odor or taste- It is

atso important to rninimize the discharge of suspended solids since suspended material

may reduce the level of Iight penetration adequate for photosynthetic activities in

receiving water bodies. Finally it is necessary to discharge the effluent at a temperature

that may not advenely affect the receiving water body. An abnomally high temperature

resuIts in depletion of dissolved oxygen (by lowering the saturation value) and may

enhance microbial activity.

The chemical characteristics of wastewater are dosely related to the above-

mentioned physical properties and a complete chernical description of the nature of

wastewater would require evaluation of the nature and quantity of soiid particulates, as

well as al1 dissoived constituents including gases.

Solid partiwlates in municipal wastewater exist as either settleable or suspended

solids. Settleable solids accumulate in settling tanks forming deposits made up of

inorganic matter such as clay, sand, and grave1 as well as some high-density organic

substances. Suspended materials include organic forms, finely divided organic matter of

animal and vegetable origin and inorganic colIoidal solids.

Dissolved solids, liquids and gases are ail present in municipal wastewater. The

dissolved solids and Iiquids are made up of organic and inorganic matter that originate

from a variety of sources- The organic component includes substantiai concentrations of

humic matter, carbohydrates, proteins and arnino acids as well as smaller arnounts of

synthetic organic chemicals. The inorganic matter consists primanly of simple salts,

alkalinity-causing compounds. and nutrients such as compounds containing nitrogen and

phosphonis-

Besides gases originating from the atmosphere, gaseous products - most

importantly carbon dioxide, rnethane and in some cases ammonia and hydrogen sulfide

- anse from the demmposition of waterbome organic matter-

The biological component of wastewater is comprised of a wide vanety of

rnicroorganisms- These organisms take part in diverse chernical reacüons and, to some

degree, the product of these reacüons defines the organic chemistry of the wastewater.

Furthemore, some biological organisms are capable of causing waterbome diseases.

Several chemical and biolagical parameters are used to characterite wastewater.

Some measurements are used to determine the bulk chemical characteristics while

others are made to identify and quantify specific compounds. The principal bulk

charaderistics are g iven below ** '. The hydrogen ion concentration (measured as pH) of wastewater is frequently

deterrnined because proper pH adjustrnent is essential during treatment in order to

optimize both chemical coagulation and biological acüvity.

The proton-accepting capability of water is known as alkalinity. Chemical substances

that contribute to alkalinity in municipal wastewater include hydrogen carbonate,

carbonate and hydroxide as well as other species that can a d as proton acceptors.

In most cases, hydrogen carbonate is the major contributor because it occurs

naturally from the reaction of carbon dioxide with water, and is the principal

carbonate species at the near neutral pH values that are typical of wastewater.

Alkalinity is a very important parameter since the proton-accepting components react

with the acid generated by some coagulants. Alkalinity also serves to minirnize the

pH changes that wuld result from biological proœsses in the wastewater- In less

common situations, the acidity instead of alkalinity of wastewater rnay be

detemined, This is done where the water has no proton-accepting capacity- Acidity

is defined as the capacity of the water to neutralize base- Acidity onginates from

wastewater constituents such as fatty acids, carbon dioxide, hydrogen sulfide and

iron (Ill) salts,

= BiochemicaI olcygen demand (BOD) is a measure of the amount of oxygen used by

bacteria and other organisms during aerobic biological decornposition of organic

matter- Degradation of organic matter is an oxygen-cansuming pmcess and can be

represented in simple fom as follows:

0 2 (aq.) + organic matter -+ CO2 (aq.) + transformed biological products

The BOD is detemined by placing an aliquot of wastewater in a 30GrnL BOD boffle

containing the nutrients required for biological growth. The sample is diluted with

water saturated in oxygen, and then incubated for 5 days at 20' C. The BOD is a

measure of the amount of dissolved oxygen that is used up by the microorganisms

dufing this 5day period.

The total amount of oxygen required for chemical oxidation of the organic matter to

its end products is designated chemical oxygen demand (COD). It involves the use

of a potassium dichromate oxidation procedure to measure both readily degradable

and more refractory (Le. substantially non-biodegradable) organic and inorganic

oxidizable compounds. The refractory organics comrnonly include fluorinated

h ydrocarbons, chlorinated pesticides and detergents with aryl and alkyl sulfonated

groups. Because COD measurements account for additional classes of compounds,

the values are nonnally higher than 80D values. The BOD and COD tests do not

account for organic cornpounds that are partially or totally resistant to either

biochemical or dichromate oxidation,

The total organic carbon (TOC) value is a measure of the overall organic matter

content of wastewater- The assessrnent of TOC involves the complete oxidation of

organic matter in a high temperature furnace to produce carbon dioxide, which is

detected and quantified by infrared spectroswpy. In addition to TOC, the dissolved

organic carbon (DOC) can also be similarly measured on the filtrate from a filtered

sam ple of wastewater.

Nitmgen in wastewater exists in both organic and inorganic (ammonium and nitrate)

foms with ammonium ion being predominant. Ammonium ion is toxic to fish at

relatively low concentrations and can exert a significant oxygen demand- Nitrates on

the other hand enhance eutrophication of water bodies and can cause severe illness

in infants. The primary sources of nitrogen in municipal wastewater are feœs, urine

and food-proœssing discharges4. Seveal technologies have been developed for the

removal of nitrogen; among these, biological nitrification-denitrification processes are

established, well understood and widely used 2v '. Nitrification-denitrification relies on

designing a system with an aerobic (with dissolved oxygen) zone for nitrification and

anaerobic (dissolved oxygen-depleted) zone for denitfification- ln the aerobic zone,

ammonium-nitrogen is converted into nitntenitrogen and subsequently into nitrate-

nitrogen.

Bactefial decomposition Organic-nitrogen b NH4' t energy

Nitrifying bacteria 2NH4'+ 302 b 2N02-+ 4H' + 2H20 + energy

(Nitrosornonas)

Nitn'fying bacteria 2NOg (nitrile) + Oz (Nitrobader) b 2NO3' + energy

Nitrate nitrogen is then converteci into nitrogen gas in the anaerobic zone- These

reactions are as follows.

Bacterial denitrification 2N02' + CH30H ) N2 + CO2 + Hz0 + 20H-

6N03-+ 5CH30H Bacterial denitfification

b Nz + 5C02 + 7H20 + 60H'

It has been shom6 that the denitrification step is more diftiwlt to achieve than

nitrification and depends on variables such as carbon substrate type and

concentration, dissolved oxygen concentration, temperature, alkalinity and pH of the

wastewater,

Phosphonis is a key nutrient supporthg growth of algae and aquatic plants; it

therefore contributes to eutrophication of natural water bodies. In municipal

wastewater, phosphonis may be present as a vanety of organk species and also in

the fom of inorganic orthophosphate and polyphosphates- Much more will be said

about phosphonis removal in later parts of the thesis-

1.3. The composition of municipal wastewater

Due to variations in the source of the water, the composition of municipal

wastewater is highly variable. Table 2 presents the ranges of values for some of the

aforernentioned physical and chernical parameters for untreated municipal wastewater.

Table 2. Typical composition of untreated municipal wastewater

Constituents Concentration, mg C'

Range Average

Solids, total (TS)

Suspended solids (SS)

Fixed

Volatile

Settieable solids (mUL)

Dissolved, total (TDS)

Fixed

Volatile

Al kalinity (as CaC03)

Biochemical oxygen demand, (BOD5,ZO OC) Chernical oxygen demand (COD),

Total organic carbon (TOC)

Nitrogen (total as N)

Organic

Free ammonia

P hosphorus

Organic

l norganic

Chlorides

Sulfate

--

Source: Corbit, A. R. Standid Handbook of Envimnrnentaf Engineering McGraw-Hill

Inc., New York. 1990.

1 4 Wastewater treatment processes

Four levels of treatment processes are used to remove the undesirable

components in wastewater: prelirninary treatment, primary treatrnent, secondary or

biological treatrnent, and advanced treatment 3. The choice and configuration of these

processes and the detailed design depend on factors including characteristics of the

wastewater, guidelines for qualit. of the discharged effluent, the capital and operating

cost for the treatment system and site availability-

The goal of preliminary treatrnent is to remove large, coarse materials that may

create mechanical problerns or interfere with the optimum functioning of the plant

Screens (parallel bars with openings of 3.8 to 6.4 cm) are used to remove coarse and

settleable solids by surface straining. The large particles with specific gravity higher than

about 1.6 are ailowed to settle out in the grit chambers. To enhance the setüing, the flow

velocity of the wastewater into the grit chambers is greatiy reduced.

This treatrnent process removes suspended matter (inorganic and organic solids

that are not removed during preliminary treatment) as sludge in sedirnentation tanks

while, at the same time, skirnrning off the floaüng scum. In some plants, primary

treatment and prelirninary treatment are part of a single unit. If there is no further

treatment, the effluent is disinfeded and discharged. Typically, about two-thirds of the

suspended rnatter, one-third of the BOD and most of the floating rnatter are removed

during the primary treatment step A flow diagram for a plant using only preliminary

and primary treatrnent is illustrated in Figure 1.

In older wastewater treatment plants that cannot easily expand, chernical

coagulants (often aluminum- or iron-based) are used to en hance primary treatment b y

desta bilizing suspended particles and precipitating dissolved inorganic and organic

cornponents The benefits derived from the use of coagulants to enhance primary

treatment have been reported ' to niclude the following:

Suspended solids. BO0 and phosphonis are reduced substantially-

Settling rates of solids generated as sludge are increased.

Significantiy smalIer seffling tanks are needed than would be required when there is

no chernical treatment

Table 3 shows the removal efficiency achieved with and dttiout the use of iron (III)

chloride coagulant in three wastewater treatment plants in Ontario.

Table 3. A cornparison of phosphorus removal performance of three pnmary

treatrnent plants in Ontario, Canada, with and without coagulant addition

(F~cI~)'.

% Vernoval Parameters Sarnia Windsor Burlington

Wthout With Without Witti Wthout Wth FeClj FeCb FeC13 FeCla FeCla FeCh

Suspended solids, 64 84 40 66 55 54 (mg 0 BOD, 34 64 24 62 40 62 (mg L") Total phosphorus 10 86 6 68 33 78 as P (mg c')

* For ovemow rate less than 24 m3/ m2d-

In most plants, primary treatment is not the final process, but serves as a

preparation for further treatment of the municipal wastewater,

Wasfe siudge

Sanitary landfiIl sife

Chlonnating point

Figure 1. Schematic diagram showing the basic units in a primary treatment plant,

1.4.3. Secondary or bÏologÏcal tteatmnt

Secondary treatment is now used in many wastewater plants as it has been

found that combined biological and chernical processes are essential for achieving

discharge water of acceptable quality.

Several biological processes have been developed for the secondary treatment

of municipal wastewater 21 The activated sludge process is a widely employed

secondary treatment system in which sludge containing mixed populations of

degradative, aerobic, heterotrophic microorganisms is utilized to break down organic

matter (Figure 2). In this process, the pnmary effluent is mixed with return activated

sludge and is passed continuously into an aeration tank fMed with bubble diffusers or

surface aerators This ensures an adequate supply of dissolved oxygen required for

maintaining a high ievel of rnicrobial activity. It also keeps the sludge in suspension to

facilitate contact beîween the microorganisms and the biodegradable organic matter in

the wastewater- Dunng the aeration process, microorganisrns in the mixture oxidize a

portion of the suspended and dissolved organic matter into carbon dioxide while the

rernainder is incorporated into new microbial cells. The miXed muor from the aeration

tank then flows into a final sedimentation tank or clarifier. Under quiescent conditions,

the sludge containhg living microorganisms setfies to the bottom of the clarifîer. The

settled sludge is removed and some of it is recycfed into the head end of the aeration

chamber where the microbial population is able to grow and stimulate the rnetabolic

degradation of additional waste organic matter. In order to maintain a proper balance

between food supply and mass of inicroorganisms in the aeration tank, excess sludge is

removed (wasted). This balance, known as food-to-micrwrganism ratio (FM), is a vital

component of the operation of an activated sludge system. Optirnization of influent BOD

removal can be achieveâ by regulating factors such as the F/M ratio, detention time and

the quantity of air used for the aeration of the wastewater 1

In order to achieve a superïor effluent quality, a large number of plants utilize

chernical coagulants during biological treatment The advantages arising from the use of

chernical coagulants have been highlighted to include high removal efficiencies of

phosphorus. BOD and suspended solids of al1 types Furthemore. the adivated

sludge generated is more compact, settleable and has lower water content than sludge

that has not been enriched by an inorganic component

A companson of the activated siudge process performance with and without

coagulant (alum) addition is illustrated on Table 4.

Table 4. Cornparison of effluent quality obtained from the treatrnent of influent

with and without alurn addition in an activated sludge proœss.

Parameter Influent

( ms L-'1 ( ms L-'1

Nomal Alum assisted

ss 110

BOD 71

COD 1 72

Soluble P 6.7

Total P 10.0

Source: Handbook of Wafer Resoums and Pollution Contmi, Gehrn, H .W. and Bregman, J. I., Eds. Van Nostrand Reinhold Company, New York, 1979.

Bar s c m n Gnf mmoval

Aemtion tank

final effluent

Chlonnating point

Figure 2. Schernatic diagram showing the basic units in a secondary or biological

treatment plant.

Tertiary or advanced treatment processes are used for the removal of residual

suspended solids and dissotved materials soch as nutrients and trace organics following

secondary treatment A wide vanety of physical, chernical and biological processes have

been applied to enhance the rernoval of specific contaminants in order to meet or

exceed specified quality levels. Many of the processes require high capital and running

costs and so they are used only where needed in specific situations.

1.5. Aluminum coagulant chemistry

It was indicated prevîousiy that both pnmary and secondary treatrnent processes

may be improved by the simuîtaneous use of coagulants to enhance the removal of

undesirable components in the water. in many situations, aluminum salts are the

coagulants of choiœ because they are able to promote the destabilization of suspended

inorganic colloids and at the same tirne effect more efficient phosphorus and organic

matter removal. When alum, (A12(S04)s.n&O), is added to water, it dissociates to give

trivalent AI", which exists in hydrated fomi as the hexaaquoalurnïrtum ion AI(H~O)~*. In

water containing carbonate or hydrogen carbonate species or other sources of alkalinity,

the aquo aiumïnum ion, AI(H~O)~* undergoes a series of rapid hydrolytic reactions to

fom soluble monomenc and polymeric species as well as solid hydrous aluminum oxide,

whose formula is usualfy given in simple form as AI(OH)3. The equilibriurn reactions (ai

25OC) for the hydrolysis of aquo aluminurn ions (AI(H~O)~? can be represented as

shown below :

Controversy still exits over the nature of hydrolysis intemediates prior to the

precipitation of aluminum hydroxide, and several different polymeric foms of aluminum

have been proposed. For example, one researcher 'O has suggested that [AI&H)~~" is

the predominant aluminum species in the acidic pH range rather than the products

obtained from Reactions 1 and 2 above. The existence of other polyrneric aluminum

species in partially neutralized aluminum systems in which base was slowly added to

aluminum salts solutions has been reported in other studies 12- 13. The

[AI~~O~(H~O)~~(OH)~J" species (known as Ab3) has been show to be the predominant

aluminurn species in some partially neutralized aluminum systems '3. Most of the

species forrned during aluminum hydrolysis are of a transitory character, which makes it

difficult to characterize them using available analyücal methods* Another cause of the

controversy is related to differences in the experimental conditions used in order to fom

the pol ynuclear species. investigators who prepared their hydrolyzed aluminurn solutions

by mking a solution of aluminum sa1 with base concurrently l4 '" 'T la or by the

dissolution of a solid phase l9* 20 reported the existence of monomenc aluminum. On the

other hand investigators who prepared their hydrolyzed aluminum solutions by adding

base very slowiy to a solution of hexaaquo aluminum ion reported the existence of

polymeric aluminum species 'Oe "* 24.

Several techniques have been used to study parüally hydrolyzed aluminum

specie~'~. Among these techniques, potentïometric methods feature prominently for the

rneasurernent of the hydrolysis equilibria. Potentiometric data were used to provide

evidence for the existence of aluminum species such as [AI~(OH)J'+ and [AI~,(oH)~]*

while 2 7 ~ 1 nuclear rnagneüc resonance (NMR) 12* 25 was used for the identification and

characterization of the Ali3 polycation. Some other polyrneric species that have baen

shown to exist in aqueous solutions of aluminum l3 are Iisted below.

~AI(H~o)? + 2H20 - [A12(~20)8(0~)d4+ + 2H30' (Log K = - 7-7)- (5)

3~1(~20)6* + 4Hz0 - [AI~(H~O)~~(OH)~]~ + 4H30* (Log K = - 13.9)-(6) 13AI(H 2 0 ) ~ ~ + 28H20 - [ . & 4 (H~O)I~(QH)~~]~+ + 32H&* (Log K = - 98.7)-(7)

Other techniques that have been used for the identification of polynuclear

species include light scattering and small-angle x-ray scattering.

Despite the controversy over the nature of the hydrolysis reacüons and products,

the following deductions have been made:

Aluminum hydmlyzes to form soluble monomenc and polymeric species and

solid precipitate. The overall hydrolysis reaction as indicated in the reacüons

involves H30' ions (on the produd side of the equaions as -tten) and hence

the concentrations of the vanous species are pH dependent Other factors that

influence the speciation of aluminum in aqueous systems include rate of base

addition. temperature. mixing conditions, presence and cuncentration of other

species

Hydroxo aluminum complexes readily adsorb on surfaces of suspended colloids,

and the charges they possess tend to reduce the negative charge of the

surface.

The thermodynamic data obtained by Baes and Mesmer are considered

accurate l3 and the aqueous ~lubility diagram for aluminum in ternis of monometic

species (Figure 3) is based on these data,

2

-2

Log CA!

-6

-1 0

Figure 3. Solubility of aluminum as a function of pH, in the absence of

complexing agents other than hydroxide Ions.

This diagram indicates that aluminum solubility, in the absence of additional

complexing agents, is at a minimum between pH 6.0 and 8.5.

In summary. the simple aqueous chemistry of aluminum can be described by five

rnonomeric species (AI(H~O)~~', AI(H~~)PH~+, AI(H2O).@H);, AI(H20)3(OH)I, and

AI(HzO)2(OH);). polymefic species (such as [AI~(H~O)~(OH)J"*, AI~(H~O)~~(OH])$+ ,

[AIB(oH)~~" and [ A I ~ I O ~ ( H ~ O ) I ~ ( ~ H ] ) ~ ~ ] ~ ~ and a solid precipitate (AI(OH)3).

The destabilization of suspended solids and wlloids using hydrolytic products of

metal ions such as iron (III) and aluminurn (III) is referred to as chemical coagulation.

During coagulation processes, rapid mixing is used to facilitate production of a

homogeneous mixture of the coagulant and the colIoidal suspension. The mechanism of

the destabilization proœss depends, in part. on the aluminum species ttiat interact with

the dissolved or solid mnstituents of the wastewatet-

When aluminurn salts are used for coagulation, the principal removal

mechanisrns are referred to as charge neutralizaüon and sweep coagulation:

Charge neutralizaüon involves the adsorption of positively charged aluminum

hydrolysis species ont0 the negatively charged colloiclal partfcies in the wastewater,

thus reducing the intnnsic negative charge on the particles and consequently

allowing for particle aggregation and precipitation- Because the reactions that

produce monomeric species such as AIOH" are extremely fasf occuning within

microseconds 26, the alum should be dispersed as rapidly as possible (< 0.1 s) so

that the most reactive of the hydrolysis products will be available to effect charge

neutralization. The optimum aluminum concentration produœs sufficient positive

charge to exadly balance the negative charge of the colloids. Overdosing the

wastewater with excess coagulant will cause charge reversal of the wlloids followed

by restabilization and this condition worsens treatment performance.

The addition of a sufficient amount of alum to wastewater containing alkalinity

ultimately results in the precipitation of aluminum hydroxide- As the voluminous and

flocculent inorganic precipitate settles, it is able to physically entrap colloidal particles

and cany them down with the precipitate- At the same time, some dissolved species

are removed by adsorption onto the precipitate- This former mechanism of particle

destabilization is referred to as sweep-coagulation or enmeshment The precipitation

of aluminum hydroxide occun within 1-7 s "* In sweep coagulation, extremely

short dispersion times and high intensities of mwng are not as crucial as in charge

neutralization-

1.6. Phosphoms and oigrnic matter in municipal wastewater

1.6.7. Phosphorus fonns

In municipal wastewater, phosphorus exists as simple inorganic phosphate

(orthophosphate), complex inorganic phosphates (tripolyphosphate and pytophosphate)

and organic phosphates. These various species of phosphorus are present in soluble as

well as diverse partiwlate forms. The amount of phosphorus is highly variable but in a

typical municipal wastewater, the approximate concenmons of the various foms have

been estirnated a as orthophosphate (5 mg P L*')'), tripolyphosphate (3 mg P L"),

pyrophosphate (1 mg P LI), and organic phosphates (1 mg P L-'). The distribution of

ortho-, pyro- and tripolyphosphate species depends on the nature of the source and is

also govemed by pH (Table 5 a). During the treatment proœss, complex inorganic

phosphates are converted to stable orthophosphate, which is the fom that is most

readily available for biological adiviües ? The organic phosphates undergo bacterÏal

decomposition and are also converted into orthophosphate. mus, orthophosphate is the

principal form of phosphorus that is removed in wastewater treatment plants. The

representative organic phosphates likely to be found in municipal wastewater are shown

in fable 5 b.

Table 5 a. The inorganic phosphate species found in municipal wastewatec

Foms of inorganic Species depending on pH Predominant species in

phosphates wastewatet

Orthophosphate H3PO4 (pKa = 2.1)

H2POi (pKa = 7.2)

HPO,~- (~K~ = 12.3)

Tripolyphosphate ~ ~ ~ 3 0 4 0 ~ - (pKa = 2.3)

H ~ P ~ o ~ ~ ~ (pKa = 6.5)

H P ~ O ~ O ~ (pKa = 9.2)

Pyrophosphate

Table 5 b. Foms of organic phosphates found in municipal wastewater. - -

Foms of organic phosphates Representative compounds

Sugar phosphate Glucose-1-phosphate

Phospholipids Glycerolphosphate

Organic condensed phosphates Adenosine-5'-mphosphate

1.6.2. Reasons forphosphoms contml ri, wastewater

Nutnents such as phosphorus (as orthophosphate) and nitrogen (as ammonium,

nitrite or nitrate) cause eutrophication of water bodies. Municipal wastewater is rich in

both phosphorus and nitrogen but phosphorus release to a receiving water body is of

particular environmental significance because it is often the lirniting nutrient for plants

and growth of microorganisrns. Furthemore, the cantrol of eutrophication of water

bodies through phosphorus removal is usually preferred because the sources of nitrogen

in the aquatic environment are Iess controllable and the removaI processes more

cornplex 31. 32. Y

1.6-3. Phosphonrs contd

Control of phosphorus has received considerable attention due to problerns

associated with eutrophication. To curb this growing problem, standards have been

established for the manufacture of phosphate-containing detergents and stringent

restrictions imposed on the levels of phosphonis in wastewater effluent For example,

an effluent total phospho~s concentration of 1.0 mg is required for wastewater

treatment facilities in Ontario with operating capacity more than 450 m3 d". More

Mngent restrictions requiring less than 1.0 mg L-' phosphonis have been established

for effluents in some emlogically fragile areas designated as 'Areas of Concem'. In

Ontario, these areas include the Bay of Quinte, Collingwood and Hamilton H a h n .

Efforts to meet the established standard and Iimit the phosphorus loading into water

bodies have led to the widespread use of coagulant-assisted biological treatment

plants '.

Phosphorus is at least partially removed by biological processes during

secondary treatment In order to enhance uptake by microorganisms, modifications of

the standard activated sludge proœss are required j5. This is achieved by taking

advantage of a phenomenon known as luxury uptake. To encourage luxury uptake,

microorganisms are subjected to sequential anaerobic and aerobic conditions- Under

anaerobic conditions, the organisrns expenence a stress condition that enhances

release of phosphorus and uptake of soluble BOD due to lack of oxygen. In the

subsequent aerobic condition, the microorganisms prepare themselves for future stress

conditions by storing more than normal amounts of phosphorus in the form of

polyphosphates as an energy source '. As high as 97% total phospho~s removal can

be achieved ' but operational difficulties exist and make the method highly variable ".

The success of biological phosphonis control ' depends on substrate type, substrate

availability, and temperature and is Iimited when inactivation or death of the

microorganisms responsible for nutrient uptake occurs due to presence of toxic materials

in the wastewater.

While biological methods can be used to remove phosphonis, chernical methods

involving coagulants are more widely used because they oner more efficient phosphofus

removal, hproved seffleability of micmbial fioc, as well as redudion in the wnœntraon

of suspendeci substances and organic matter "-'. The tedinology is well proven and

has been practiced widely fhroughout the worfd for decades ". The chernical coagulants

thaï have been used include aluminum, imn and calcium salts (Table 6)-

Table 6. Coagulants comrnonly used for phosphorus removal in municipal

wastewater. - ppp . - - - -

Coagulant Typical example Charaderistics -- -

Aluminum Alum [A12(S04)3.nH20] Most effective between pH = 6.8 - 7.5 salts

lmproves disinfection systems such as UV

Dry alum is not corrosive but liquid alum is rnoderately corrosive

Produces less sludge than lime

lron salts Fenic chlonde

[Fe&]

Calcium salts

Calcium oxide

mol

Effective over a wide pH range (4 - 9)

More effective for odor removal

Corrosive

lron is a plant nutnent and can enhance afgae bloorns

Can cause staining of concrete and plant facilities

inhibits the use of UV light for disinfection

Requires pH adjustrnent into effective pH range

(pH=9-10)

Produces large quantity of sIudge

Can cause scaling in tanks and piping

lncreases water hardness

Of these chernical coagulants, alum, [Ah(SO&.nH20], is the most extensively

used. Through the coagulation mechanisms diswssed in Section 1.5, phosphorus is

removed from wastewater as precipitates of aluminum phosphate, complex aluminurn

hydroxyphosphate or adsorbed on the freshly precipitated hydrous aluminum oxide

floc At the same time, other components of wastewater, such as organic matter and

suspended and coltoidal matter, oompete for aluminum ion added to the wastewater by

foming complexes with aluminum species. For this reason, an arnount of alum in excess

of that requïred to rernove phosphorus in the absence of these cornpethg substances is

usually employed to achieve optimum removal of this element '. It has been reportad

that a 95% phosphorus redudion in municipal wastewater is generally achieved when an

aluminum to total phosphorus (TP) molar ratio range between 2.i:l and 2-33 is

utilized 2.

In the acüvated sludge wastewater treatment process, addition of alum can occur

at various points (Figure 4). including immediately upstream of the primary clarifier

(Point l), in the aeration chamber during aeratian (Point 2). immediately downstream

from the aeration chamber but prior to final clarification (Point 3) and at more than one

point simultaneously 4'.

Waste sludge

Ra w inasteMa&

(Influent)

Aeration tank

Chlorinating point

Figure 4. Alternative points (1, 2 and 3) of alum addition in an activated sludge

plant-

In the Orillia treatment plant in Ontario, both single-point and dual-point alum

addition schemes have been tested to achieve effluent with low levels of

phosphorus ''. For single-point alum addition, 64 mg L" of alum was adckd to the outlet

of the aerator (Point 3 in Figure 4) whereas in dual-point alum addition, an alum dosage

of approxirnately 16 mg L" at oie aeated grit chamber (before Point 1 in Figure 4) and

32 mg L-' at the tail end of the aerator tank (Point 2 in Figure 4) was used. Plant

performance data showed that average effluent phosphorus of 0.36 mg was achieved

using duabpoint addition compared with average effluent phosphonis of 0.65 mg L-' for

the single-point alum addition scheme. Similar improvements in phosphonis removal by

two point coagulant addition have not, however, been achieved in al1 orner situations.

In the Kingston West Sewage Treatment Plant (KWSTP). there are four parallel

activated sludge plants designated A, 6, C and D- ln these units, coagulant is added ai

the aeration tank (Point 2 in Plant A) behnreen the aerator and the secondary clarifier

(Point 3 in Plant 8) and at the head of the aeratar (Point in Plants C and D).

Depending on the parüwlar plant, each of these locations has advantages and

disadvantages- Furthemore, the choice of point of alum addiüon is often based on

engineering considerations. Because the complexing cornponents of wastewater

(including organic compounds) compete with hydroxyl ions and phosphate species for

aluminum species, the possible effect of organic matter on phosphorus removal is one

factor that should be considered in detemining point of addition of the coagulant.

1.6.4. Mechanr'sms of phosphorus mmoval

The precipitation of phosphonis dunng wastewater treatment using alum is often

illustrated as follows:

This ovenirnplified representation is misleading for several reasons. The free,

unhydrolyzed metal ion, AI* is highly hydrated in solution (AI(H~O)~Y) and as the pH of

the solution increases, it undergoes hydrolysis to fonn soluble and insoluble aluminum

hydrolytic products as previously show (Section 1.5). Under wastewater conditions

therefare, we expect the formation of monomeric and polymeric hydroxy aluminum

complexes rather than the free AI* species. The phosphorus in the equation is also not

correct because the phosphate species. ~0~~ exkts in aqueous systems only at a pH

much higher (pH > 11) than that of typical wastewater. The orthophosphate ions that

exist under most wastewater conditions are dihydrogen phosphate ion (H2Po43 and

hydrogen phosphate ion (HPO~~-) with the predominant form k i n g HPO,--

Furthemiore, the formula for the product in the equation is not correct Mare than one

reaction has been suggested for the formation of an insoluble material by the

coprecipitation of aqueous aluminum and phosphate. Same suggested reactions are

illustrated in Equations 9 - 1 1

Proton release (pH < 7)

Charge neutralization (pH < 7)

Partial charge neutralization (pH > 7)

Reiease of OH - (pH - 7)

Another reaction has been proposeci 42 to describe how solid prehydrolyzed aluminum

species interad with phosphate via a postprecipitation process. The reaction involves

the rapid precipitation of hydrous aluminum oxide floc followed by the adsorption of

phosphate onto or enmeshment of phosphate species by the floc. Adsorption ont0 solid

aluminum hydroxide is represented as follows:

Freshly precipitated amorphous hydmus aluminum oxide has a high density of

amphoteric surface hydroxyl groups that play a signifiant role in phosphonis removal- In

the absence of other complexing substances, aging of hydrous aluminurn oxide in the

solution from which it was formed results in its transformation to a more crystalline state-

Gibbsite develops in an acidic solution and pseudoboehmite or bayerite in a basic

solution, with consequent reduction in specific surface area

The tendency for adsorption of phosphate ont0 hydrous aluminurn oxide

therefore decreases with the age of the precipitate ". On this account, for effective

removal of phosphorus, postprecipitation reaction conditions that favor formation of

freshly precipitated amorphous hydrous aluminurn oxides are desirable. In the activated

sludge process, the Iifetime of the hydrous aluminurn oxide in recycled sludge in most

plants is 2-10 days, a period much too short for the complete transformation of the

amorphous precipitate into more crystalline aluminum species.

The soi1 and geological science Iiterature also has shown mat arnorphous foms

of hydrous aluminum oxides in soils are capable of binding phosphates ". One of the

proposed mechanisms involves the specific sorption of phosphates by replacing the

coordinated -OHz or -OH groups in either aluminum or iron hydrous oxides (Equations

14 and 15) a.

ii HO-P- OH

I

The reaction represented in Equation 15 is consistent with that describeci in

simple fom by Equation 13 above.

It has also been cleariy shown by wastewater plant performance data and

laboratov studies that removal of phosphonis does not occur through the formation of

aluminum phosphate (AlPo,) ". The equilibriurn phosphate concentration at vanous pH

levels for AlPo, has been detemined, and the data used to show that AlPo* is not

fomed at neutral pH Another study has demonstrated that phosphonis removal

occurs through adsorption ont0 amorphous aluminum hydroxide rather than direct

precipitation of aluminum phosphate ''. It has k e n suggested that the precipitation of

phosphate using aluminum salts is governed by the integration of both AI-OH-AI and

AI-PO4-AI Iinkages into an aluminum hydroxyphosphate solid M e r than by precipitation

of discrete phases such as AI(OHh or A l P Q ". For example. chernical formulae

proposed as representing the average composition of aluminum hydroxyphosphate

species preàpitateâ at pH = 5.0 and pH = 6.0 are Al13(OH)(H2P04)18H2 a and

AI, -4P04(0H)r z " respectively-

Apart from the organic phosphoms species Fable 5 b). other organic

wnstituents in wastewater, both soluble and in solid fom span a wide range of chernical

structures and rnolar masses- The soluble or dissolved fom of organic matter is defined

as that fraction of the total organic matter, which passes a 0.45 pm pore filter. A

considerable amount of general information is available, related to the classification of

this dissolved organic rnatter (DOM) ? In the latter paper it was shown that the

soluble fraction of secondary effluents contained ether extractables (IO%, of which two

thirds was organic acids), carbohydrates and polysaccharide (5%). proteins (10%),

tannins and lignins (5%), and anionic surfactants. for example. alkylbenzene sulfate (10

Oh). Studies conducted by Rebhun and Manka indicated that 40 to 50 Oh of the total

organic content of secondary effluents frorn trickling fifters used for aeration consisteci of

humic substances. Other organic components identifieci in the same effluent induded:

ether extractables (fatty auds. 8.3%). carbohydrates (1 1.5%), proteins (22.4%), tannins

(1 -7%) and anionic detergents (1 3.9%).

The organic material is therefore made up of substantially hydrophobic

compounds (humic acid, hydrocarbons, ethers) as well as hydrophilic ones (sugars,

hydroxyacids, polysaccharides, arnino acids). The functional groups in some of these

cornpounds undergo ionization and the degree of ionization is pH dependent.

R-COOH (carboxylic) + HzO - RCOO- + H30' (pK 4 - 5)-(16)

R-OH (phenolic) + Hz0 - RO- + H307 (pK-9.5) (1 7)

Where negatively charged sites are present on organic molecules this may enhanœ

their availability for interaction with hydrous aluminurn oxide species during the

coagulation,

7.66- Mechanism of otganiic matter mmoval during coagulation pmcess

The mechanism of organic removal from water and wastewater is a chernically

complex process and has attracted considerable attention *'* "* ".

Dissolved organic matter removal mechanisms are pH dependent and involve

reactions such as aluminurn hydrolysis, complexation benNeen aluminurn species and

organic matter, precipitation of soluble species and adsorption of DOM ont0 hydrous

aluminum oxide surfaces. At pH values less than 5, positively charged aluminum

hydrolysis species can interad with negatively charged suspended organic matter and

negativel y charged dissolved organic species with both reactions leading su bsequentf y

to organic removal by precipitation. This mechanism known as charge neutralization-

precipitation (CNP) " was discussed earlier in Section 1.5. At pH values greater than

7 the mechanism of organic matter removal occurs by adsorption of humic materiafs

ont0 hydrous aluminum oxide or by a coprecipitation process that results in the formation

of aluminum fulvate (or hurnate). It has been suggested that aluminum fulvate

complexes should exist at tygical water treatrnent conditions and the species are

probably sorbed ont0 hydrous aluminum oxide floc6'. Between the pH range of 5 - 7,

both removal rnechanisms can occur depending on factors such as concentration of

DOM and alum dosage,

The presenœ of suspended partides of mineral species during coagulation of

organic rnatter has been shown to influence the mechanism of organic matter

coagulation. Effective coagulation of kaolinite particles in the presence of fulvic acid was

found to be dependent on fuivic acid concentration and it is pmposed that coagulation

involves the precipitation of alurninum hilvate wmplex followed by physical enmeshment

of the kaolinite partÏcles ". The nature of the organic matter can also infiuenœ its

coagulation and removal. Coagulation studies on river water using alum "* have

shown that hydrophobie compounds were more cornpietely removed than hydrophilic

compounds- Furthemore, high molewlat weight compounds were more effecüvely

removed than low rnolecular weight compounds.

1.6.7. Tannk acîd as a surmgaie for organic maffer in wastewater,

In this work, tannic acid U A ) was chosen as a surrogate for soluble organic

matter in the expenments involving the use of simulated wastewater. Tannic acid has

been used as a model compound for soluble organic matter in studies on drinking water

treatment In the field of soi1 science, it has also been used as a model high

molecular weight organic compound in several studies 61. 68, 69, 70. 71, 72, 73

Tannic acid is water-soluble and has an average rnolar mass of 1701 g mol-'. It

occurs naturally in the bark and fruits of many plants and is extracted in quantity frorn

Turkish or Chinese nutgall. Tannic acid has k e n reported to have a structural formula

as shown in Figure 5 and to consist of digalloyl moieties linked to glucose. The structural

formula shown in Figure 5 suggests that TA contains ten ester linkages but has no free

carboxylic acid group 74. However, TA has been refened to as a polyhydroxy-carbowlic

acid in oie Iiterature There are also studies that indicate that TA has a pK. value that

is indicative of the presence of carboxylic acid groups in the rnolecule As wilf be

shown later in this thesis, TA shows acidic properties in aqueous solution and

measurements of the pKa and the FTlR data also suggest the presence of carboxylic

acid groups in the molecule-

Figure 5 a. The digalloyl basic unit in TA

Figure 5 b. The structural composition of TA (where R = the digalloyl shown in Figure

5 a above)

The sacdiaride, aromatic as well as 400- (or COOH) and phenolic -OH groups

in TA makes it simlar in some ways to soluble fulvic or humic acid. In contrast to TA

however, humic substances have a wide range of molar mass and Vary in composition

depending on their source. and method of exiracüon ".

Commercial humic substances have also been used as model organic

compounds in water and roi1 research, but they too have limitations. Each commercial

product has characteristics related to its source and method of extraction, and these

rnay not be tNly representative of the DOM in any other situation.

Although successes have been achieved in the removal of phosphorus and

organic matter during wastewater treatment, the influenœ of organic mater on

phosphorus removaf is still a controversial subject The sorption of organic matter ont0

mineral surfaces has been shown to change the physical and chemical properties of the

material and consequently influence its interactions with other solutes ". In soils, it has

been shown that phosphate fucation is reduced due to the specific sorption of organic

matter and subsequent blocking of adsorption sites on hydrous @des of aluminum and

iron 79* ? There are, however. other reports that show an opposite effect - increased

phosphate adsorption which is atûibuted to inhibition of iron oxide crystallization "* a ".

The confiicting results may be due, at least in part to expen'mental conditions used in

studying the reactions, including the tirne durhg which the organics and oxides were in

contact

The dissoived organic carbon (DOC) concentration in dïfferent units in the

activated sludge wastewater treatment plant in Kingston has been estimated '? The

differences in organic matter concentration in the various parts of the plant rnay

influence dephosphorization efficiency, residual aluminum concentration in the effluent

and the physical and chemical characteristics of the sludge.

The foregoing discussions therefore jusüfy working towards a better

understanding of the chemical processes associated with municipal wastewater

treatment in order to optirnite conditions for both phosphorus and organic matter

removai 86.

1.7. Chernical methods used for the chrtracterization of wastewater

constiiuents

1-7- 1 _ The tbm method

The ferron reagent is a mixture of o-phenanthroline, hyroxylamine hydrochloride,

and ferron (8-hydroxy -7-iodo-5-quinoiine -sulphonic acid) aged for at least 5 days. The

ferron rnethod is based on the reaction between aluminum species and ferron reagent to

fom colored aluminum-fenon complexes- By measunng the absorbanœ of the solution

containing the complexes at 370 nm, the concentration of aluminum species can be

detemined-

The ferron method was originally used to classify aluminum species in aqueous

solutions into ttiree fractions denoted by Al,, Alb and Al,. These species were classified

based on their dïfferential rate of reaction with ferton. Al,, represents monomenc

aluminum species which react instantaneously with ferron; Al are poiymeric aluminum

species which react slowly at a rate that can be show to follow first order kinetics, and

Al. represents the extremely slow-teacting or inert solid aluminurn species ".

Recentîy it has been shown that the ferron method can be extended to the

investigation of solid hydmus aluminum oxides The Cndings of that study are

summarized as follows:

Freshly precipitated solid hydrous aluminum oxides reacted with femn reagent over

a relatively short time (at a rate intermediate to that of the Alb and Al, species in the

original procedure) and as the hydrous aluminum oxide is allowed to age, the solids

become Iess reactive.

The source of aluminurn ion undergoing hydrolysis did not influence the reacüvity of

hydrous aluminum oxides formed with ferron, or the aging trends of the solids.

H ydrous aluminu m oxides precipitated under wel1-mixed mnditions were more

reactive than those precipitated under quiet or pooriy stirred conditions.

The presenœ of phosphate in the hydrous aluminum oxide resulted in decreased

reactivity of the solids with ferron at younger ages and increased the reactivity a i

older ages compared to solids confaining aluminurn abne-

1.7.2. Fourier transform rilfraried spectmscopy.

lnfrared spectroscopy (if?) has been widely used for the qualitative identificatÏon

of the absorption bands associated with specïfic functional groups in oxides and

hydroxides, and also to study solid metal hydroxide phase transformations- The IR

spectrum of hydrous aluminurn oxide shows absorption bands due to hydroxyl gmups in

the region between 2800 - 3800 cm ' (stretching modes) and 8W -1200 cm "

(deformation modes). An intense H-O-H bending vibration is usually observed neat

1640 cm-', which indicates the presence of water in the solid. The absorption bands that

are observed between 500 - 800 cm '' and 650 - 900 cm -' have been assigned to AC0

vibrations of Al in octahedral and tetrahedral coordination environments respectively. An

overlapping region exists between 650 and 800 cm -' where assignment of peak(s) as

octahedraf or tetrahedral is uncertain and this makes the application of this technique as

a diagnostic tool for the determination of aluminum coordination environment unreliable

in this region.

lnfrared spectroscopy has also been used to provide useful information on

functional groups involved in sorption studies as well as on the possible surface

structures of inorganic and organic anions or ligands adsorbed on hydrous aluminum

oxides surfaces. The presence of strong broad absorption band@) between 1100 and

1040 cm-' is indicative of P-O stretching vibrations in phosphatacontaining hydmus

aluminum material and the position of the band(s) depend on the metals or other species

associated with phosphate ". It has been usad to explain the role of carboxyl and

phenolic functional groups in the sorpüon of organic materials onto the surface of the

hydrous oxide. Spectral changes such as a shift in the absorption band in the region

1605 - 1725 cm-' to lower frequency and the appearance of new bands near 1600 cm"

and 1375 cm-' was associated with involvement of carboxyl groups in this

interaction n*m* a." lnfrared spectroscopy was usecf to show that surface complexaüon of

salicylic acid on goethite (iron (III) oxide) involves both a carboxylic and a phenolic

oxygen 95 whereas in the aluminum oxide-salicylate system, only carboxylic oxygen was

bound to the aluminum oxide

In the present study, we have used this technique to aid the identification of the

hydrolytic products, monitor the structural changes occumng dunng aging and to explain

the nature of the chemical interaction between aluminum and the wmponents

(phosphorus and organic matter).

Magic angle spinning aluminum-27 nuclear magnetic resonance (27~1 MAS NMR)

is now an established non-destructive technique for the chemical and structural

characterization of amorphous solid matenals that contain alurninum. Magic angle

spinning NMR studies have been used to elucidate the molecular structure of aluminurn-

containing glasses %. aluminophosphate molecular sieves or zeolites g9, aluminum

(iii)-glycolate complex lW ceramics 'O1, aluminum-oxygen compounds '5 and aluminum

hydroxophosphates Rie high resolution associated with MAS NMR makes it possible

to distinguish between the aluminum coordination environments in a solid sample. In the

MAS NMR specüa of solids, the chernical shifts depend not only on the coordination

nurnber of the aluminum atorns but also on the nature of its neighboring atoms and

hence provide structural information of interest. Octahedral Al sites with aluminum as

second neighbors (AlOs) have peaks with chemical shiRs in the range O to + 22 ppm ".

However if the second neighbors are phosphorus atoms instead of aluminum atoms, the

chemical shiis of the peaks are located upfield (O to -20 pprn). Tetrahedral aluminum

sites (A104), where the second neighbors are alurninum or phosphorus, yield chernical

shifts between 50 to 80 ppm, and 33 and 45 ppm respectively ". Several studies have

reported peaks occumng between 20 and 40 pprn and these investigators have

identified the coordination environment as fivecoordinate or distorîed odahedral

aluminum lm* lm- Five-coordinated AIQ units in phosphatesfq yield chemical shifts in

the range of 14 to 16 whereas in oxides lm the chemical shift is between 30 to 40 ppm.

Other peak parameters such as the relative intensity of the peak and peak width are

good indicators of the amount of aluminum in a particular environment and degree of

rnolecular order respedively. It has been shown that highly ordered materials identified

as crystalline by X-ray diffraction (XRD) generally have peaks with width less than 10

ppm Mile the peak width for arnorphous materials are greater than 10 pprn In the

present study, this technique was used to investigate the structural organization of

aluminum in precipitates obtained under various wastewater treatment conditions in

order to provide information conœming aluminurn-phosphate-organic matter interactions

during wastewater treatment-

7.7.4. Atomic force microscopy

Atomic force microscopy (AFM) is a scanning probe technique that has bmad

application in profiling surface structures of materials from the micron scale to the atomic

scale. Several researchen have demonstrated the unique potentials of AFM in the

characterization of the surface features of wide varieties of environmental materials such

as biological molecules, polymers, metals and catalyst surfaces 107. 1 ~ . 109.- ln ,=nt

years, AFM has developed frorn a simple, high-resolution profilometer to an advanœd

microscopic method for studying different properties sucti as composition, adhesive

forces, electrostatic forces and elasticity 110. Hl. 112

The basic components of an atomic force microscopy instrument are a cantilever

fitted with a tip (scanning probe) to serve as sensor, a detector to measure the cantilever

displacement and a computer to display the image of the sampIe (Figure 6)-

Two AFM imaging modes are commonly utilized: contact-mode AFM and

tapping-mode AFM. In contact-mode AFM, the scanning probe maintains contact with

the sarnple as it traverses the surface. The surface features of the partides deflect the

cantilever and the laser beam focussed on the cantilever is deflected to a photodiode,

which converts the deflection into ' topographie information ' such as the height of the

components in the sample. However, the use of contact-mode for imaging soft materials

results in undue sample deformation and hence the tapping-mode was developed to

overcome this problem '13.

Tapping-mode AFM has become an indispensable method for imaging soft

materials sucb as organic substances Il4* "'. It can be used to monitor changes in

surface properties of particles such as shape, size and stiffness in homogeneous and

heterogeneous surfaces '? In the tappïng mode AFM, a fast oscillating probe makes an

intermittent contact with the sampk during scanning. When the tip touches the surface

of the sample, the interadion forces shift the resonance frequency of the free oscillating

cantilever. The difference between the phase of the free oscillating cantilever and the

modified phase is monitored and used to generate compositional mapping. Thus the

phase changes and the optical defleaon of the oscillating probe are used to

simultaneously record both phase shift and height images. In the height and phase

images, dark regions correspond to lower values of height and phase shift, while brighter

regions correspond to higher values. .

Phase shift images show contrast for morphological features when the material

under investigation is homogeneous whereas, in heterogeneous rnaterials, îhey ïndicate

difïerenœs in the mechanical properties of individual components "'. Assigning the

cause of phase contrast to sarnple composition can be difficult since the magnitude of

the force applied to the surfaœ of a sample can signifscantiy change the data, especially

the phase data '? In tapping-mode AFM. the amplitude & of the free oscilfathg

mntilever and the set-point amplitude ratio (r,) are very important experimental

parameters. These parameters are used tu control the magnitude of the force applied to

the sample surface. The seGpoint amplitude ratio is a ratio of the set-point amplitude

(Arp ) to A, (rSp = Arp / & )- From tapping-mode AFM studies of polymer blends, it has

been suggested that by systematically varying A, and r,, image contrast caused mainly

by the stiffhess of a sampk surfaœ can be identifieci. Images that closely reflect the

'tme' topography of a sample were acquired at sufficiently high & and f, values and

image contrast indicative of stifhess of a sample surface at high & and moderate

r, values l lT .

l Laser

Ily coated cantilever

le and holder

Dellect ion Sensor

driving cantilever

Figure 6. Schematics of AFM

No information has been published conceming the use of this new technique ta

study the chernical identification of surface features of colloidal particles- In the present

work, the AFM technique is used to monitor changes in surface topography of hydrous

aluminum oxides precipitated frorn simulated wastewater in order to gain an insight into

aluminum-phosphate-organic rnatter interactions during wastewater treatment. Although

the contrast observed in phase images in tapping mode AFM are related to the

mechanical properties such as stiffness, these properties are diffiwlt to quantiry. There

are several disadvantages of using cantilever-based AFM to quantify nanomechanical

properties- AFM cantilevers are unstable in the attractive force regirne- There is also

difficulty in calibrating the cantilever spnng constant including cantiiever displacement

response over the entire range of displacements. Furthemiore, there can be prablems

associated with applying load verücally to the AFM cantilever tip "'.

Interfacial force microscopy is a scanning probe technique reœntiy developed to

overcome the problems inherent in tapping mode AFM such as the d-ficulty in obtaining

accurate measurements of forces. This technique uses a difFerential-capcitance

displacement sensor capable of imaging the surfaces of materials with sub-micrometer

spatial resolution throughout the attractive and repulsive force range 'la . The

interfacial force microscope senses force by displaœment of the sensor and uses an

automatic force-feedback scheme to rebalance the displacement and hence eliminate

the instability that plagues the cantilevers used in AFM techniques.

Its potential for the quantitative measurement of nanometer-scale adhesive and

mechanical properties of surfaces and intefiaces, with nN to pN force resolution has

been demonstrated I l 8 . The data generated during IFM measurements are displayed in

foms such as images and forces curves. The images are similar to AFM images but are

lower in resolution. Force cuwes are obtained by monitoring interfacial forces as a

function of distance prior to tip-sample contact, affer contact and upon wittidrawal of the

sensor. By analysing the force curves, useful sample properties such as adhesive forces

and elasticity can be obtained- Interfacial force wwes have been used to provide

information on adhesive monolayer film-tip interaction I2O* j2' and elastic modulus of

different types of samples "'- la* '"- In one of these studies 12', the interfacial forces

between a tungsten tip interacüng with a self-assembled monolayer of hexadecylthiol on

thin gold film was investigated. The IFM images revealed a number of large-scale

defects while the force profiles of the surface showed chemical differences in the vicinity

of the defects- The investigation demonstrated that by combining the imaging and force

profiling capabilities of IFM it is possible to differentiate topographical and chemical

features of heterogeneous samples. A typical force curve of a sample and how

information can be extrapolated frorn the curve are descn'bed in Secüon 2.12.

In order to quantify the qualitative data given by AFM studies, IFM

measurements of the sutface properties of some of the precipitates as well as the

surface of the mica substrate were made,

1.7.6. Scanning eIectmn probe ana\ysis

Scanning electton probe analysis is a standard technique used to provide

electron and elernental distribution images of solid samples- Electrons emitted from a

tungsten ' hairpin' filament heated to 2500 O C bombard a solid sample in order to obtain

'micrograms' showing not only the topography of the sample but also the distribution of

particular elements at the surface la.

Scanning electron probe analysis was used in the present study to ascertain the

homogeneity of the precipitates obtained from coprecipitation and postprecipitation

systems.

1.8. Summary of the research objective

The purpose of the present study, therafore, was to contribute to a better

understanding of the intefaciion between alumïnum. organic matter and phosphorus in

the wastewater treatment process. Because the organic matter content is variable

throughout the plant, it is hypothesized that the point of alum addition would influenœ

the extent of phosphate removal and the nature of the solid phase products. Knowiedge

of the interaction between alurninum, organic rnatter and phosphorus in the treatment

process would therefore aid in detemining conditions for wastewater plant design that

would optimize the removal of both phosphorus and organic matter.

2.1. Matenals

All chernicals used for analysis were obtained fmm BDH and were reagent

grade. Distilled de-ionized water (DDW) was used for sarnple preparations, The alum,

which was commercial grade, was obtained from the Kingston Watet Purification Plant-

Afum stock solutions were used either wWIthout dilution or in diluted fom (IMO). Aqueous

solutions of 0433 mol L~~ sodium bicarbonate (NaHC03), 0.323 mol L? potassium

dihydrogen phosphate (KHtPO,) and 1.176 x 10'~ mol L-' TA (C~enaOa) were prepared

and used as stock solutions for phosphorus and dissolved organic matter respectively.

The concentration of TA used in the simulated wastewater, 17 ppm (9.1 mg C L-'), was

dose to the concentration of soluble organic mater (9.8 mg C L*') found in the aerator of

the Kingston West Sewage Treatment Plant (KWSfP).

2.2. Simulated wastewater expriment

This study describes two types of experirnents - Copmcipitation Experimenfs,

which are synonymous with the addition of alum before or within the aerator, and

Posfpmcr;Oitation Expeninents, which are synonyrnous with addition of coagulant at the

aerator outlet, pnor to the recycling of sludge. The synthetic wastewater was made up of

NaHC03 (100 mg L-' as CaC03. an alkalinity typical of that found in wastewater). and

also contained other components - KH2P04 (5.0 mg P L") and TA (17 mg L'') which

were added as describeci below. Alum (4.3 mg Al L-l) was used as a coagulant in al1 the

experiments. The solution volume was 1 L and a constant mixhg speed of 380 rpm was

maintained using a mixer equipped with dwl blades. For al1 the experiments, aging

penods - the time taken between formation of the precipitate and obtaining a sample of

solid, initiating the ferron test or analyzing for residuals - ranged from 5 - 120 min- The

pH of the aged solutions ranged from 6.6 to 7.8 which is M i n the range of 6.0 to 9-0

specified for secondary emuent from wastewater facilities '24.

In the coprecipitation studies, alum was added to the synthetic wastewater

containing (i) orthophosphate ions (AIPI), ( Ï i ) TA (AITA) and (iii) both orthophosphate and

TA (AIPTA). The postprecipitation experiments were canied out as foilows: (i) adding

phosphate to prehydrolyzed aluminum thaï had b e n allowed to age for 5 min in the

absence of TA (A15P) or to a prehydrolyzed coprecipitate of aluminum and TA that had

aged for 5 min (AITASP), (ii) introducing TA to sirnilady aged prehydrolyzed aluminum

(A15TA) or coprecipitate of aIurninum and phosphate (AIPSTA), and finally (iii)

intrcducing TA and orthophosphate as a mixture to prehydrolyzed alum (AISPTA

systems). For cornparison, a controt experiment (Al), involving synthetic wastewater to

which only alum (4.3 mg aluminurn K') had been added, was also carned out In al1

cases, the molar ratio of afuminum: phosphorus was approximately 1 :1. Under actual

pfant treatment conditions, a several-fold excess of aluminum is usually added to ensure

complete phosphonis removal. We opted for the near stoichiometnc ratio in order to be

able to observe differences in phosphorus removal efficiency between the various

conditions.

Samples obbined under the conditions described above gave only small

1 The symbol AIP refers to a situation in which alum is added to a solution containing phosphate

so that the aluminum (AI) and phosphorus (P) are coprecipitated- The syrnbol AlSP means that

aluminum is precipitated alone by hydrolysis, the precipitate is allowed to age for five minutes and

then phosphate is added to the mixture- In this case, phosphoms is rernoved from solution by

postprecipitation processes. The other syrnbols can be interpreted in a sirnilar way.

amounts of solids and filtration of the 1-L mixture generally took several hours- Thus, for

infrared, 2 7 ~ ~ solid-state magic angle spinning nudear magnetic resonanœ, atomic force

microsco py, interfacial force rnicroswpy and electron micro probe analyses, 5 min to

30 d aged solids were obtained using ten times the amounts of al1 chernicals as

specified above, but only 100 mL volumes of solution.

2.3. Municipal wastewater and recycled sludge sample

Untreated municipal wastewater was mllected at the raw sewage feed of the

Kingston West Sewage Treatment Plant, Ontario, Canada- Recycled sludge was also

collected from the plant, Coprecipitation and postprecipitation experiments were

performed using aliquots of the untreated wastewater and the ferton method was used

to characterize the solid products- The recycled sludge obtained from the plant was

characteflzed using 2 7 ~ 1 MAS NMR and FTlR techniques.

The pH of the samples was deterrnined using an Accumet pH meter, (Mode1825

MP). Prior to pH determination of the sample, the pH meter was calibrated with standard

buffer solutions (pH values of 4-04 and 7-47),

2.5. Dissolved organic carbon (DOC) detemination of untmated and treabed

wastewater.

Samples for DOC deteminations wera filtered through a ~ i l l i p o r e ~ membrane

filter (0-45 prn pore site), The filtrates and standards were acidified and sparged with

nitmgen gas before DOC detemination.

DOC analysis was camed out using an Astro 2001 (System 2) Total Carbon

Analyzer. Standards in the range 0-10 mg L-' were prepared from 1000 mg 1-' stock

solution of potassium hydrogen phthalate, Carbon diaxidefree water was used for the

preparation of the stock solution and standards. Triplicates of 5.0 mL of the sarnple and

standards were used for the DOC measurements-

2.6. Kinetic experiments

For the kinetic proœdure, 0.25 mL of the aged suspension was added to 1.50 ml,

of ferron reagent and 1.00 mL of DDW in a 3-cm glass cell. The contents were well

mixed, &ter which the absorbance at 370 nm was monitored continuously (against a

blank) untii al1 the aluminum had reacted to fom the ferron-aluminum complexe The

blank was made up of 1.50 mL of ferron reagent and 1.25 mL of DDW. To quant@ in a

simple manner the rate of reaction of aluminum in the mixtures with ferron, the üme

required to recover 50% of the solid phase aluminum, designated as ~SQ, was

determined- lncreasing values are indicative of decreasing reactivity of aluminum in

the solid phases. As an additional factor, the percentage of aluminum that reacted

within the first 30 s was measured and this (presumably composed of soluble species) is

referred to as Yast reacüng aluminum'. Figure 7 shows the fast-reacting aluminum

species and solid phase in a typical absorbance versus time plot for the difFerential

reaction of aluminum species with ferron. Details of the procedure for the preparation of

ferron reagent are described alsewhere

8 0.1 - m L % Solid phase aluninun species 2 0.05 -

Fast-reacting alun inurn species l I L '

Figure 7. Absorbance versus time plot for the differential reaction of aluminum

species with ferron.

2.7. Residual phosphorus detemination

Direct measurernent of residuat soluble orthophosphate was made at room

temperature after filtration of samples through ~ i l l i p o r e ~ ~ membrane filten (0.45 prn

pore size. Ths residual phosphorus concentrations of the solutions were detennined by

the phosphomolybdate blue method- This method is based on the readion of

orthophosphate with molybdate in an acidic medium to produce a phosphomolybdate

complex which is reduced by ascorbic acid to an intense 'molybdenum blue' color. The

absorbance associated with the color was rneasured at 690 nm using an Ultrospec 3000

(Phamacia Biotech Ltd-) spectmphotornetef and a 1-cm path length cell. The Hach

phosver 3 Reagent Powder Pillows which contain al1 the reagents

(potassium pyrosuifate, L-ascorbic acid and sodium molybdate) for the procedure were

used to produce the colored species- The contents of a pillow were added to a 5-mL

aliquot of the sample, which had b e n diluted to 50 mL using distilleci de-ionizeà water

(DDW). The absorbante was determined after a 10-min reaction pend and cornpared

to an extemal calibration wrve (R' = 0.999). The calibration standards used to obtain

the linear calibration airve were in the range 0-1.0 mg (Figure 8).

O 0.1 0.2 0.3 0.4 0.5 0.6 0-7 0.8

Concentration of phosphonis, mg L-'

Figure 8. Calibration plot for phosphorus determination.

2.8. Residual aluminum determination

Residual aluminum determinations in the filtrates were carried out by two

techniques: the ferron rnethod and flame atornic absorption spectroscopy using a

Perkin-Elmer Atomic Absorption Spectrometer Model 1 1008 with a nitrous oxide-

acetylene fiame. Both gave linear calibrations within the range of O to 5.0 mg L*' Al

and O to 10.0 mg L-l AI respectively. The regression equation at 370 nm for ferron

was y = 0.034~- 0.01 (R' = 0.998) and for the flarne atomic absorption spectroscopy it

was y = 0 . 0 1 ~ (R' = 0-997).

2.9. Method development for residual TA determination

Because of the presence of unsaturated fundional gmups in natural organic

matter (NOM), these heterogeneous species abmb Iight in the UV region. The UV

spectra of humic substances are feâtureless and give a fising absorbante as the

wavelength decreases below about 400 nm. Absorption at 254 nrn is a standard method

for estimating the concentration of dissolved NOM in water, and a Iinear relationship has

been shown to exist between absorbance at that wavelength and the dissolved organic

matter (DOM) content of many types of water 12'* 127.

In Section 1.6-7, it was stated that TA was used as a sumgate for DOM. Unlike

humic substances, TA has specific peaks mat are useful for monitoring its concentraüon

in water. When dissolved in water, TA produces an acidic solution (pH = 4.5 when its

concentration is 17 mg c') and the UV / vis spectrum of the solution exhibits two strong

absorption bands - one with A,,,, = 214 nm and another with &- = 278 nm. In alkaline

media, however, signifiant reduction in these two peaks is observed and a new broad

peak appears with k- = 322 nrn. Figure 9 shows the absorption spectra of

TA (17.0 mg L-') in (a) distilleci de-ionized water (pH 4.5) and (b) aqueous sodium

bicarbonate solution (pH 8.45).

Depending on pH, one of these wavelengths may be chosen for spectmphotometn'c

analysis of TA in solutions containing no other species. However, the presence of

aluminum in solution affects the nature of the TA spectnim and it was necessary to

develop an analyücal method that could be used in mis situation.

Figure 9, Absorption spectra of TA (17.0 mg L-') in (a) distilled de-ionized water

(pH 4.5) and (b) alkaline sodium bicarbonate solution (pH 8.4).

Beginning with an aqueous solution containing TA and sodium hydrogen

carbonate, the pH was adjusted to ditferent values from 3.0 to 9.0 and the spectra

recorded. Figure 10 depicts the absorption spectra of these solutions and their

respective pH values. The absorption spectra for TA in the pH range 3.0 - 5.6 ware

significantfy different from the spectra a TA obtained in the pH range between 6.0 - 9.0.

It is clear from Figure 10 that as the pH increases, an increase in absorbance near 322

nm and a decrease near 278 nm are obsewed. An isosbestic point is observed near 294

nm, which indicates the conversion of TA from one species to another. Because pH is

the variable, these spectral features must be related to proton dissociation. Using the

data obtained from Figure I O , an estimation of the p& value for the carboxylic group in

TA was detemined to be 5.8. This value is within the range 3 s pK. s 6 reported for acid

dissociation constants typicai of simple cahoxylic acids and outside the typical range of

8 s p K 5 12 for phenolic groups 128. The pK. result of 5.8 is consistent with an eariier

reported pKp value of = 6 Since the p K for phenol is 9.9 while that for the phenolic

group in salicylic acid is 13.4. the phenolic group in TA is likely to remain essentialiy

unionized up to pH z 9.0. The reducüon in the absorbana of the peak near 278 nm and

subsequent increase in the absorbnce near 322 nm over the pH range shown would

then indicate a deprotonation proœss involving carboxylic fundonal groups (Figure 10).

This result therefore implies that some carboxylic groups are present in TA molecules.

Figure 10. Absorption spectra for TA (5.0 ppm): Influence of pH of the solution.

The absorption spectra of filtrates obtained after aging mixtures prepared by

reacting aluminum ion (4-3 ppm) and TA (t7.0 ppm) in the presenœ or absence of

phosphate ions are shown in Figure 11- The spectra indicate the disappearanœ of the

band at 278 nm and formation of a new absorption peak at 315 - 322 nm- Simitar

changes in the specba of TA as a function of increasing concentration of fenic and

aluminum ions have been previously reported '". Comparison of the spectra shown in

Figure il with those obtained for TA in alkaline media (Figures 9 and 10) indicates a

similar absorption peak in the region around 320 nm. Are the absorption peaks shown in

Figure i 1 due to deprotonated TA species or an aluminum-tannate cornplex? To answer

this question, an alkaline solution of sodium bicarbonate was used to prepare four

different aqueous solutions. The first was a solution of TA with a pH value of 9.0; the

second, a mixture of TA and alurninum ions with a pH value of 9.0; the third a solution

containing TA with an initial pH of 9.0 but acidified to a pH value of 5.6; and the fourth, a

solution containing TA and aluminum ions with an initial pH of 9.0 before acidification to

a pH of 5.6. Spectra of these solutions were collected and are depicted in Figures 12

and 13. Figure 12 shows the disappearance of the predominant peak (322 nm)

attnbuted to depmtonated TA species, which subsequently became a protonated

species peak at 278 nm. However, in the spectnim of a mixture of TA and

aluminum ions (Figure 13) the absorption peak at about 320 nm remained. These

findings suggest the formation of an aluminum-tannate complex at high pH and indicate

that this complex does not dissociate when the pH is lowered to 5.6- In these solutions,

the peak near 318 nm is due to the aluminum-tannate cornplex- Although not show

here, when the pH of the solution containing the complex was reduced to less than 3.0,

the complex was dissociated as evidenced by the reversion of the spectnim to one that

was characteristic of protonated TA. The absorption peaks shown in Figure 11 are Iikely

due ta an aluminum-tannate complex rather than deprotonated free TA species.

Figure 1 1. Absorption spectra for the filtrates obtained using 0.45 pn membrane

filters frorn 30 min aged samples of AlTA and AlPTA systems.

I I

L

Figure 12. Absorption spectra of TA at two different pH values.

Figure 13. Absorption spectra of a mixture of TA and aluminum ions at two different

pH values.

In previous work, ultraviolet absohnœ at 271 nm has been used for the

detenination of residual TA conœntration '". However. it is apparent from Figure 11

that the use of absorbance at 271 nm or 278 nm for the detemination of residual TA

would produce erroneous results when aluminum is also present in the solution and the

pH was in the usual range for the coagulation experiments described here. This problem

could be overcome by releasing the bound TA in the aluminum-tannate cornpiex via

acidification to pH values below 3.0, Below this pH, the total TA (mat in complexed fom,

which had been set free, plus the uncompfexed TA) would be determined-

Having established that audification of the filtrate was required in order to

prevent an underestimation of residual TA in the filtrate, standard solutions of acidified

(pH = 2.0 with sulfuric acid) TA with concentration varying from O to 20 ppm were

prepared. The absorbance values of the acidified standard solution were determined in a

1-cm cell at 278 nrn against similady acidified distilled de-ionized water blank. On

samples, direct measurement of residual TA was made after filtration of samples through

~ i l l i p o r e ~ ~ membrane filters (0.22 Pm and 0.45 pm pore size) and acidifying to pH 2.0.

The absorbanœ data obtained from the acidified filtrates were converted to residual

concentration of TA in pprn based on the calibration curves (Figure 14) generated from

the standard TA solution The analytical features of TA detemination using this method

are summarized in Table 7.

O 2.5 5 7.5 10 12.5 15 17.5 20 22.5 25 27.5 Concentration of TA, pprn

Figure 14. Calibration curve obtained fmm standard TA solutions acidifieci to a pH

value of 2-0-

Table 7. Analyücal features of the spectrophotometric rnethod for the

detemination of TA.

Equation y = 0.0482 x - 0.003

Regression Coefîicient (?) 00.99

Linear range O - 25.0 ppm

Detection limit (LOD) 0.14 ppm

Quantification Iimit 0.61 ppm

Precision (RSD)

tow level (1 -0 pprn) 0.88%

High level (25.0 pprn) 0.08%

2.1 0. Fourier transfonn infrared spectroscopy (FTIR) analysis

Approximately 0.7 * 0.1 mg of the sample and 200 mg of KBr (dned at 105 O C )

were mked with a mortar and pede and then transferred to a KBr disc press- The disc

was then pressed at 10 t total pressure for 2 min. to produce a KBr pellet. A KBr

reference pellet was prepared in the same manner as the sample and contained only

200 mg of the dned KBr. Fourier transfomi infrared spectroscopy was perfomed on a

Bomen MB-lW lnfrared Spectrometer (100 sans at a resolution of 4 cm"). Spectra

were recorded between 400 and 4000 cm -'on samples in KBr pellets against a KBr

reference disc-

2.1 1. Atomic force microscopy (AFM) analysis

Slurries of the dned sample (approximately 0.35 g L-l) in distilled de-ionized

water were sonicated in an ultrasonic bath for 60 min to produce an aqueous dispersion

of the sarnple. A 30 pL aliquot of the dispersion was deposited on a freshly cleaved mica

substrate 1 cm2 in area with the aid of a syringe. The sample was then spun at 4000

rpm on a spin water for 60 s to ensure an even distribution of particles over the

substrate surface. The sample was then allowed to dry for 1 h before the AFM imaging.

The AFM images were acquired under ambiant conditions using a PiooSPM

(Molecular Imaging, Tempe, Arizona) operated in MAC mode fitted with a Nanoscope IIE

controller (Digital Instruments, Santa Barbara, CA). The MAC mode is essentially the

same as tapping mode 13' but diffen in that the cantilever is magneticaliy mated and

dnven by an extemal oscillating magneüc field. Silicon Nitride (SirN3) cantilevers which

had a force constant of -0.5 N m-' and a asonance frequency of -100 kHz were used.

All images were acquired at the fundamental resonance frequency of the cantilevers and

at scan speeds of 1 - 2 luie s*' using a 30 pn x 3û pm scanner. 80th height and phase

shift data were recorded simultaneously as a fundion of both cantilever oscillation

amplitude (&) and set point ratio r, = &/ &- Quantitative evaluation of the images was

performed using the Nanoscope 11E off-line analysis software.

2.1 2. Interfacial force micmscopy (IFM)

The proœdure that was reported for AFM sarnple preparation was also used to

prepare samples for IFM analysis.

All the IFM images of the control hydrolytic p-itate of aluminum (Al), the

co precipitated materia[ (AIPTA) and the postprecipitated material (A15PTA) were

acquired at a constant repulsive load of = 85 nN using contact mode imaging. A tungsten

tip was used to detemine the nanomechanical properlies of the particles. Force

calibration of the IFM sensor was camed out using gofd as a calibration standard. An

indentation modulus of 75.8 + 10.6 GPa as wmpared to the tabulated value for bulk gold

of 78 GPa was obtained-

In order to obtain quantitative data regarding the modulus of a surface, IFM force

vs displacement (F-D) measurements were made dunng each indentation experiment

An indentation involves rnoving the tungsten tip to make contact with the sample up to a

set repulsive load and then back to its original position. During an indentation,

measurements of the load as a function of üp-sample deformation are recorded. From

these measurements, F-D curves are obtained; an example is shown in Figure 15- A plot

of Hertzian fit of the elastic loading data is camed out for each curve. From Hertz theory,

the relationship between force and displacement is given by

Where F is the applied force, E* is the reduœd modulus of the system (approximately

the same as the modulus of the surfaœ since the tungsten tip is very hard), R is the

radius of wrvature of the indenter and D refers to the total displacement Thus, from the

plot of force vs displacement, and knowing R, one can determine the reduced madulus

E* of the system.

Displacement (nm)

Figure 15. F-D curve of a surface. The directions of the amms indicate a typical

loading (sensor approach) and unloading cycles (sensor retract) during data acquisition.

The zem separation value represents the tum-around point. On the curve, 1 is the

elastic loading region (increasing tip-sample contact a m ) , 2 is both elastic and plastic

deformation region, 3 is the elastic unloading region (constant tip-sample contact area)

and 4 is the region of tip withdrawal (decreasing tip-sample contact area).

2.1 3. Solid-date =AI Magic angk spinning nuclear magnetic ratonance

(MAS NMR)

Prior to solid state 2 7 ~ 1 MAS NMR measurernent, the solid sarnple was ground in

a mortar and pestle into a fine powder and subsequentiy packed (61 I 18 mg) into a

7.5 mm diameter zirconia rotor until the sample contacted the rotor and cap upon

closing. Care was taken to ensure even distribution of the sampie in the rotor dunng the

sample loading. Solid state 2 7 ~ ~ NMR measurements were made at a frequency of

104.2 MHz on a Bruker AM400 spectrometer operating at a field of 9.4 Tesla. The

probe used was a 5-mm VT multinuclear ultra high-speed double-tuned MAS probe- The

sarnple spinning speed ranged from 7.8 - 10 kHz. The chernical shift of eaai nucleus

was measured relative to a sWc signal obtained from a 0.2 mol L" aqueous solution of

AW03)3.

2.1 4. Electron microprobe analysis

Electron microprobe analysis of dried solid samples from some of the systems

was canied out at the Department of Geological Sciences of Queen's University. The

sarnples were mounted on catbon holders using doublesided carbon tape and coated

with carbon under vacuum, with the aid of a Kinney KSE-2A-M vacuum evaporator. An

ARLSEMQ electron probe (take-off angie of 52.20) in wnjunction with an energy

dispersive X-ray analyzer (Tracor Northem, Inc. Model TN-5500) was employed for the

acquisition of the electron microprobe data- The operating conditions for the acquisition

of the electron microprobe data are summarized in Table 8.

Table 8. Operating conditions for electron microprobe data on select hydrous

alurninurn oxide speciesc

Accelerating potential 15 keV

Magnification 500 x

Emission wrrent 1WmA

Beam current 40 nA

Beam size 0.5-1 pm

3.0. Resutts and Discussion

3.1. The F@mn test and Fbun'er frsnstbnn infiaried (FTI9 specfmscopy

The reaction of the solids produœd under various wastewater conditions with

feron depended on the composition of the reacting mixture and on the aging peBod. The

solid phases formed in the presenœ of orthophosphate or TA or a mixture of both were

different fmm one another and from solid phases forrned in the absence of these

ligands. Data showing the % of fast reacting aluminum and the vatues of obtained

from the femn experiments are shown in Tables 9 and 10. In general, solids obtained

frorn coprecipitation of TA and aluminum were the most reactive while solids obtained

from postprecipitation of prehydrolyzed aluminum with a mixture of TA and phosphate

ions were the least reacüve. In addition, higher values for % fast-reacting alurninum

species were found in systems containing TA compared with systems in which the

hydrolysis was camed out in the presence of only phosphate or aluminum ions-

Table 9. Percent fast reacting aluminurn species (FR) and b values obtained frorn

ferron tests on coprecipitation systems after different aging periods.

Systems Aging time (min)

5 30 60 90

O h FR (s) % FR tS0 (s) % FR ta@) % FR (s)

Al 6-6 1715 4.5 29i8 13 3527 29.7 4254

AIP 7.0 2887 5 .O 4050 6.5 471 3 7.2 5134

AiTA 22.2 247 22-1 488 22-6 510 27.2 633

AiPTA 7.9 2453 12.3 2700 13-3 2448 15.9 2717

Table 10. Percent fast reacting aluminurn species (FR) and values obtained from

ferron test on postprecipitation systems after ditferent aging periods.

Systems Aging tirne (min)

-- -

AISP 3.7 41 08 3.7 5287 4.4 5740 1.5 6000

AlP5TA 5.5 9822 13.2 5550 11-3 5022 13.0 4599

AISTA 15-7 3824 21 -1 767 26.6 768 28.6 615

AITASP 15.2 954 14.8 1376 9.9 1521 18.2 1691

AISPTA 4.7 17100 3.1 11400 8 -7 7500 11.8 6300

The FTlR technique provided qualitative information on the composition of the

solids and the nature of the interactions between the wmponents in the reacting

solution- All the solid phases of the systems investigated showed a broad OH band in

the region 3491 to 3431 cm" indicative of amorphous materials. The spectnim of the

hydrolyzed aluminum in the presenœ of alkalinity alone was significantly different from

the spectra obtained for solids precipitated in the presenœ of TA and / or phosphate.

Unlike the spectrum of crystalline aluminum species such as bayerite and gibbsite,

w hich are characterized by well-defined absorption bands, the spectmm of freshly

precipitated hydrous aluminum oxides contained very broad absorption bands.

Significant differences were observed between the spectrum of TA and those of the

solids precipitated in the presence of TA in the region 7W - 1800 cm-'. Compared with

the spectrum of pure TA, those for solids precipitated in the presence of TA with or

without orthophosphates showed absorption band shifts (1712 cm" to 1700 + 4),

reduction in some band intensities (peaks at 1712 and 1200 cm-') and the presence of

new band at 1371 cm". Reduction in the intensity of the peak associated with the

carboxylic acid group (1713 cm") has been attributed to its dissociation H i l e the new

bands st about 1600 and 1375 cm-' were also reported to indicate formation of

carboxylate ions to Mich metal ions were bonded via electrovalent linkages 70v ?

Further details regarding FTlR spectra of selected solid samples are presented under

each subheading for the individual systems,

3.1.7. CoprecIpitation studies

Ai system (Prirc@itation of hydrous aîuminum oxide alone)

When alum was added to water containing alkalinity, a white precipitate of

hydrous aluminum oxide was immediately forrned. The rate for reaction of the solid with

ferron (Table 9) decreased as the solid aged in the suspension. Similar observations

were previously reported for hydrous aluminum oxides precipitated and aged for a

prolonged period of time The decrease in surface readnrity was attributad to a

reduction in the specific surface area of the precipitate due its gradua1 transition fmm

amorphous to a more crystalline afuminum hydroxide- While the tso value increased, the

amount of initial fast-reacting cornponent also increased somewhat as the sotid material

aged, accounting for benNeen 6.6% (after 5 min aging) and 29-7% (after 90 min aging) of

the total aluminurn- Parallel to the increase in highly reactive aluminum. the

concentration of residual aluminum increased from 4% (5 min aged) to

27% (120 min aged), wnfinning that one or more soluble species of aluminum was the

reactive form. While aging, the pH of the mixture increased from 6.78 to 7.90- Aluminurn

hydroxide, an arnphoten'c wmpound, exhibits minimum solubility in the pH 6.0 - 6.5

range and above these pH values, the solubility increases, with the preponderant

species on the alkaline side being AI(0H); (Figure 3).

The FTIR spectnim of the 'simple' hydrous aluminum oxide solid is shown in

Figure 16. Broad, relatively featureless infrared absorption bands in both the -OH

stretching and deformation regions indicate the amorphous nature of the freshly

precipitated solid phases la. An absorption band that can be attributed to AC0 vibrations

of aluminum in octahedral coordination is seen in the region between 400 - 750 cm-' ". It

is interesting to note the presence of two absorption bands at 1538 and 1404 cm-'.

These are characteristic of the presence of carbonate as a unidentate complex with

aluminum 133 . It has been observed that atmospheric carbon dioxide adsorbs strongly as

carbonate on moist geothite surface '" and thus the source of carbonate in the

precipitate from Al system could be attributed to adsorption of atmospheric carbon

dioxide during the drying of the unwashed sample.

1 1 I 1 1 1 1

4000 3500 3000 2500 2000 1SOO 1000 500

Wavenumber, cm -'

Figure16. Fouriertransfoninfrared(FTIR)spe~mofthe'simple'hydrous

aluminum oxide solid (AIS) in the region 4000 - 400 cm-'. [Al] = 430 mg L-'.

AIP systems

The surface reactivity of the solid precipitated when aluminum hydrolysis

ocwrred in the presence of orthophosphate was significantly lower Vian that of 'pure'

hydrous aiurninum oxide Fable 9). The sdid was very unreacüve to f e m . highly

insoluble, and residual concentrations of aluminum were generally less than 4% of the

initial concentration of aluminum, Like the pure material however, reactivity decreased

further wïth aging tirne.

The Rmared speanim (Figure 17) of this solid had an intense H-O-fi bending

vibration maximum around 1634 cm" and a very strong broad absorption at 1086 cm".

The presence of a strong band between 1100 and 1 WO cm" has been attributed to P-O

stretching vibrations in inorganic orthophosphates and the position of this band depends

on the species of phosphate and assodated ions ". The specûum of AIP was neariy

identical to that of wavellite lS Ala(OH)3(P04)2.5H20 (Figure 18) hdicating that this solid

is a fom of aluminum hydroxyphosphate combining both AI-OH-AI and AI-PO4-AI

linkages 52 rather than a mked material containing aluminum hydroxide and aluminum

phosphate individually. An absorption band was also observed near 543 cm", which is

indicative of aluminum in an octahedral coordination environment Furtherrnore,

absorption bands observed at 540 - 550 cm-' have been assigned to HPO,~- in

phosphate containing materials,'" which again implies that H P Q ~ ~ - species interacted

with aluminum in AIP system-

Figure 17. Fourier transfomi infrared (FTIR) spectrum of solid sample

obtained from AIP system (AIP5) in the region 4000 - 460 cm-'. [Al] = 430

mg L"; [Pl = 500 mg L-'.

40-

30 - 4

20 - P . ?Y, ,

1 O 1 1 I

4000 3500 3000 2500 2000 1500 1000 500

Wavenumber, cm ''

When TA was coprecipitated with hydrous aluminum oxide, the precipitate that

fomed exhibited very high readivity as ïndicated by the small value (Table 9). This

property can be attnbuted to the presence in the sotid of transiuonal pores and

macropores, which provide easy access for penetration and degradation of the bonds

between TA and aluminum. Previous studies '" have demonstrated that the specifïc

surface area of aged products of hydrolysis of aluminum was about five times larger

when camed out in the presence of TA, compared to hydrolysis without organic matter

present. Another study that also obsenred an increase in the surface area of

precipitation products of aluminum due to incorporation of TA noted that the disordered

nature of the precipitaüon products causes the exposure of the edge -OH groups in the

solid 'O. The ferron test also showed that there was more fast-reacting aluminum (greater

than 20% of the recovered aluminum) than in soiids obtained using either of the

precipitation conditions already discussed - The concentrations of residual aluminum

were similar (18 €0 27%) to the % fast reacting aluminum, again indicating that it is a

soluble fom of the rnetal that reacts 'instantaneously' with ferron.

The FTlR spectnim of pure TA is presented in Figure 19. This spectrum

compares very well with those reportad in the literature 'O* 75* and assignrnent of the

absorption bands is given in Table 11.

Wavenumber, cm -'

Figure 19. Fourier transfomi infrared (FTIR) spectnim of TA

in the region 4000 - 400 cm-'.

Table 1 1. Fourier transfomi infrared (FTIR) spectroscopy peaks characteristics of

TA-

Frequency Assignment

(cm")

O-H stretching vibration

C=O stretching vibration of carboxylic acid or / and ester

COO- C=C, C=O conjugated with double bonds, hydrogen- bonded C=O

Stretching vibration of aromatic C=C bonds

Bending vibration of aliphatic C-H groups

O-H deformation and C-O antisymmetric stretching of ester

C-O symmetric stretching of ester and syrnmetric C-O stretching of phenols

Symmetnc C-O stretching of phenols anNor alcoholic O-H groups

G O stretching of carbohydrate

The FTIR spectnim of the solid fomed from the coprecipitation of aluminum and

TA is illustrated in Figure 20. This spedrum reveals absorption band shifts and changes

that suggest the formation of solid products dïfîerent from either Al or AIP solids.

Compared with pure TA, the absorption band at 1712 cm" was shifted to a lower

frequency (1 704 cm-') and new absorption bands indicative of formation of carboxylate

ions appeared near 1595 and 1370 cm". These peaks are slightly shifted from the

corresponding ones in pure TA- Tannic acid has been used as a surrogate for organic

rnatter in previous studies "* 7s of interactions with aluminum minerals in soils. The series

of absorption bands reported between 1700 and 1000 cm" in the spectra of TA with

allophane and imogolite were also different from those of pure TA and were in positions

neariy identical to those observed in the present AmA system. Huang and co-workers se*

" attributed these bands, on samples obtaineâ after very long reaaion times. to the

presence of hydroxy-Al-tannate complexes-

An absorption band at 614 un" indicative of aluminum coordination in an

octahedral environment was also observed in the spedrum of the AITA solids- The FTIR

data therefore indicates mat these solids were hydroxy-Al-tannate corn plex and the

coordination environment of aluminum in the complex is octahedral,

I 1 1 I 1 ' 1 I I 1

4000 3500 3000 2500 2000 1500 1000 500

Wavenumber, cm "

Figure 20. Fourier transfomi infrared (FTIR) spectnim of solid sample

obtained from AITA system (AITAS) in the region 4000 - 400 cm-'.

[Al] = 430 mg L-"; FA] = 1700 mg L-'.

A IPTA systems

The values for fSo in the AIPTA system indicate a solid that is different from either

AITA or AIP. Not surprisingly, the for AIPTA was higher than that for the fast-reacting

AlTA but lower than that of AIP. Furthemore, unlike these two-component cases, the

reactivity of the AIPTA solids did not change significantly over a 90 rnin aging period.

The AIPTA solids were very fine. and large fractions passed through 0.45 TI (84 %) and

even 0.22 pm (64 %) membrane filters. For the other cases, most or al1 of the

precipitated materials was retained, even by a 0.45 pn filter. Figure 21 shows the

absorbance measurement (at 370 nm) versus time during the ferron test on a solution

containing AIPTA solids (120 rnin aged) before and after membrane filtration. The plot

shows that large amounts of aluminum were present either in a soluble fom or as

particles that had nominal dimension of less than 0.22 pm and 0-45 pm respectively.

Figure 21. Absorbance venus time for the ferron test on a solution aintaining AlPTA

solids (120 min aged) before and after filtration using membrane filters.

The small, mlloidal size of the partides when aluminum, TA and phosphorus

were precipitated together is likely to result in poor seffling with inefficient suspended

solid removal during wastewater treatment- However, m e n the ratio of aluminum to TA

was increased to 1.5, the precipitate that was fomied setüed readily, probably as a result

of more cornpiete neuutrelization of the negative charge associated with TA and

phosphate. A fermn test revealed that there was no aluminum in the filtrate. Therefore,

additional coagulant can be used to enhance removal of these fine particles.

When the concentration of TA in the AlPTA system was increased from 2.5 ppm

to 100 ppm, the percent fast-reacting alurninum increased from 6.9 % to 47. 7 % and the

solids mat were fomied became more readive (Figure 22) in the early stage of aging.

Figure 22. Influence of increasing concentration of TA on surface reactivity of solids

obtained from AIPTA systems with ferron reagent-

500 -

O

There was a significant difference in how surface reactivity of AIPTA changed with

time, depending on the concentration of TA. The surface reacüvity of systems containing

2.5 ppm decreased as the solution aged and this trend was similar to the trend observed

in Al and AIP systems. However, as the concentration of TA was increased to

100.0 pprn, an increase in surfa- reactivity with aging period was observed (Figure 22).

The surface reactivity of the solid aged for 120 min was lower than Al solid aged for 5

min but higher than AITA solid aged for 90 min Fable 9) suggesting that an AITA-like

material was evolving and the incorporation of phosphate into the solid network seerns

to be inhibited. ln addition, M i l e almost al1 the solids in AlPTA system containing 2.5

pprn of TA were retained on 0.45 pm membrane filter, in systems containing 100 pprn

1

5 15 30 60 120

Time, m in,

TA a large fraction passed through both the 022 pm and 0.45 pn filtes, These findings

suggest that as the concentration of TA increases. the colloidal size of the precipitated

matenai decreases. The presence of a peak at 31 8 nm in the absorption spedra of the

filtrates obtained from the 100 ppm TA system and the data descnbed in Section 2.9

suggest that the soluble aluminum woufd exist in the fomi of aluminum-tannate

complexes in the aged suspension. This TA concentration (400 ppm is equivalent to

53.6 mg C L-') represents a lower carbon content than was measured (soluble organic

matter between 60 and 96 mg C L-') in raw sewage feed samples at the treatrnent plant

in Kingston West,

Data from all the systems described above indicate that the reactions involving

hydroxyl ions, TA and orthophosphate wïth aluminum ions are parallel and cornpetitive.

At low TA concentrations, the TA and phosphate are incorporated into the network of the

hydrous aluminum oxide floc. However, as the TA concentration increases, a higher

concentration of soluble complexed aluminum species are fomed and a higher

concentration of residual afuminum remains unprecipitated

The interaction of organics to fom complexes with soluble metal species prior to

precipitation has been suggested as the first step for the removal of organic matter from

wastewater p. and from dnnking water '". Precipitation ocairs either when the binding

capacity of the natural organic matter (NOM) has been satisfied or the solubility of the

ACNOM complex is exceeded '=- The FTIR spectnirn of AIPTA (Figure 23) has peaks that are characteristic of

both aluminum tannate and aluminum phosphate Iinkages. The shift of the

orthophosphate band from 1086 cm-' (AIP system) to 1074 cm" (AIPTA system)

indicates the presence of tannate as an associated ion in the complex. Similar shifts in

an equivalent absorption in a metabphosphate-fulvic acid complex compareci to the

140 corresponding absorption in the pure metal phosphate have been reported .

Absorption bands at 555 and 618 cm'' indicative of aiuminum coordination in an

octahedral environment were obsenred in the spedrum of the solids- The FTlR data

therefote are consistent with a view that these solids were hydroxyphosphate-AI-tannate

cornpiex and the coordination environment of aluminum in the comptex is octahedrai,

Wavenumber, cm -'

Figure 23 . Fourier transfomi infrared (FTIR) spectnim of solid sample

obtained fmm AlPTA system (AIPTA5) in the region 4000 - 400 cm-'

[Al] = 430 mg L-'; [Pl = 500 mg L-'; F A ] = 1700 mg L-'.

In an atternpt to assess phosphorus removal by sludge components during the

recycling of sludge in the adivated process, different prehydrolyzed systems containhg

alurninum, aluminumphosphorus and aluminum-TA solid species were midiad. These

systems designated as AISP, AISTA, AISPTA, AIPSTA and AIP5TA were prepared by

initially aging the component(s) appeaflng before the numeral for 5 min and then adding

the component(s) after the numeral for an additionai aging penod ranging fmm

5 - 120 min.

Al5P system

This system is characterized by having the smallest amount of fast-reacting

aluminum and it had surface reactivity even lovver than that found in coprecipitated AIP.

The reactivity decreased still further while aging, This finding is consistent with

phosphates first binding to the surface of freshly precipitated hydrous aluminum oxide " and subsequentiy k i n g incorporated into the structure of the solid. The presence of

phosphate wncentrated at the surface of the precipitation product is likely to hinder the

ability of femn to degrade the solid phase product in order to react aluminum.

The infrared spectrurn of AlSP (Figure 24 ) is similar to that of AIP except for the

more intense H-O-H bending vibration with a maximum around 1640 cm-', which may

indicate a higher concentration of waters of hydration. The phosphate band was less

intense in A15P than in the AIP system indicating that the former contains less

phosphate. This is supported by measurements of residual phosphorus as reported in

Figure 31. The carbonate bands found in the original hydrolytic aluminum precipitates

were present but exhibited significantly reduced intensities suggesting that some surface

sites had been blocked dunng phosphate adsorption, thus reducing the amount of

carbon dioxide that could be sorbed from the air. Absorption bands that show aluminum

to be in an octahedral coordination environment were observed at 552 and 610 cm?

Like Me AIP system, the absorpüon band at 552 suggests that the HPO,*- species

interaded with prehydrolyzed aluminum solid, The FTlR data therefore indicates that

these solids were aluminurn hydroxyphosphate and the coordination environment of

aluminurn is octahedral.

Wavenumber, cm ''

Figure 24 . Fourier transfon infrared (FTIR) spectnim of solid sample

obtained from AlSP system (A15P5) in the region 4000 - 400 cm".

[Al] = 430 mg L"; [Pl = 500 mg L-'.

A15TA sysfems

As indicated by the large value of tsa. the AI5TA solid that had aged for 5 min

reacted only very slowly with ferron, In one sense, this is surprising because the

coprecipitated AITA solid was very readivs, for reasons diswssed above. However. we

suggest that the initial slow readivity could be due to a surface coating of TA ont0 the

prehydrolyzed aluminurn hydmde Roc; evidence for this is discussed under the section

describing AISPTA systems. Such a coating would inhibit the interaction bahNaan the

undedying aluminurn ions and the ferron reagent in solution. The increased reaavity as

the solid aged uiuld be due to the surface TA intefading slowly with inorganic material

beneath it, eventually forming a product similar to the stmcturally distorted material from

the AlTA coprecipitation case. The proportion of very fast reacting species Iikewise

increased with time, approaching that of AITA. The effect of prolongeci aging of AISTA

solids was evaluated. in order to confirm if such aging would eventually resuIt in an AlTA

type of material. The result of monitoring the t 50 of the precipitates during a 12-day aging

period is shown in Figure 25 . The figure shows an initial rapid increase in values over

the first 71 hours of aging and then a much slower rate of increase. At the end of the

12 d aging period, the value had decreased to 11 06 s, a value which is approaching

633, the value reported for AlTA system aged for 90 min. This result is consistent

with the view that the AISTA solids were slowly being transfomed into an AITA Iike

material.

O 2000 4000 6000 8000 1ûûûû 12000 14000 16000 18000

Time, min.

Figure 25 Influence of long-term period aging on surface reactivity of AISTA

solids,

The FTlR spedrum of AISTA (Figure 26 ) confinns the presenœ of functional

groups of tannate in the solid. The carbonate bands found in the original hydrolytic

aluminum precipitates as well as Ïn A15P precipitates were absent which is probably due

to the presence of organic wating inhibited absorption of carbonate. In order to

compare the degree of interaction between TA and the alurninum in AISTA and AlTA, the

height of the band near 1448 cm-' (bending vibration of aliphatic C-H groups) was used

to normalize absorption bands associated wïth binding of TA lo alurninum in the region

1800 -1000 cm-'. The data analysis reveals that the new bands (at about 1600 and

1375 cm") associated with the formation of carboxylate ions in the hydmlytic products,

had relatively higher intensities in AlTA systems [1595 and 1369 cm"] compared to

A15TA system [1595 and 1371 cm-']. These results suggest a higher degree of chernical

binding in the former case An absorption band indicative of aluminum coordination in

the solids being in an odahedral coordination environment was observed at 610 an-'.

Based on fenon and FTIR data for AISTA, therefore we propose that the initial

reaction of TA, a high molar rnass compound, with prehydrolyzed aluminum Ïnvolves the

formation of an organic coating on the solid. The hypothesis regarding coatings on solids

in similar situations has precedents in the environmental Merature- It has been shown,

for example, that large organic molecules, whether charged or not, are strongly retained

by physiso@on on the surface of clay minerals "'. It is interesting that slightiy more TA

was removed by postprecipitation than by coprecipitation. More will be said about mis

below-

Figure 26 . Fourier transform infared (FTIR) spectnim of sofid sample

obtained from A15TA system (A15TA5) in the region 4000 - 400 cm".

[Al] = 430 mg "; FA] = 1700 mg L-'.

AIPSTA, AITA5P and Al5PTA sysfems

Three other cases of post-precipitation were also studied. Reacüon of TA with

prehydmlysed aluminum hydroxyphosphate (AIPSTA) resulted in a preupitate which

exhibited very low reactivity, although the reacüvty of the solid încreased somewhat with

aging as had ocairred in the AISTA system (Table 10). The solid AIP5TA products were

much less reactïve than the already unteactive A15TA precipïtate irrespective of age-

Interestingly, the AIP5TA solids differed significantly in reactivity from those

obtained in the AITASP system. The latter rnaterial was much more reactive and

reactivity also showed an opposite trend with aging. As well, the % fast-reacüng

aluminum species in the latter system were higher, indicating that part of ttie aluminurn

ion remained in solution probably as an Al-tannate cornplex- The marked differences in

surface reactivity between AIP5TA and AITASP would appear to be due to differences in

the chemical modification of the aluminum on the surfaœ of the solid phases.

For AIPSTA as for AISTA, we postulate that the postprecipitation addition of TA

results in coating the initially forrned aluminum hydroxyphosphate solid followed by slow

structural distortion as the TA reacts with the precipitate, substituting for the original Al-

OH and ACO-P bonds. The displacement of phosphate from the solid by TA was shown

by increased soluble residual orthophosphate as aging proceeded (Figure 31a). On the

wntrary, the solid produced in the AITASP system showed decreases in surface

reactivity with aging and a simultaneous increase in phosphorus removal (Figure 31 )

and TA removal (Figure 32) from the simulated wastewater. However, phosphorus

rernoval in this system was the least efficient of al1 the cases. The low phosphate

removal in the AITASP mmpared to AIPSTA shows mat the surface of the initially forrned

hydroxy-aluminurn-tannate solid is apparently somewhat resistant to replacement by

phosphorus.

Solid AISPTA was by fa? the least reactive of the matefiais investigated;

moreover, Iike AISTA and AIP5TA where TA was removed by postprecipitation, there

was an increase in surface reacüvity with aging (Tables 9 and 10)- In addition, the

% fast-teacting aluminurn species also increased gradualIy *th aging- The increase in

soluble aluminum continu& with further aging and fast-reading aluminum made up

28 Oh of the total after 15 h, Twenty four percent of the aluminum either remained in

solution or was a fine enough solid to pass through a 0.22 prn membrane filter

(Figure 27 )_

(114 f

Figure 27 . Absorbante versus time for the fermn test on AISPTA solids (15 h aged)

before and after fiitration using membrane filters.

Again, we postulate that the low reactivity of the solid phases is due to the TA

coating, an effect that is in some way enhanced even more by the simultaneous

presence of phosphorus. The inhibitory effect of the wanig in various TA

postprecipitation systems was especially important over short aging periods and in

cases where the concentration of TA was large-

A cornparison of the peak positions in the absorption spectrum of the filtrate

obtained from AlSPTA with #ose in the absorption spactnim of the filtrate of AlPTA

indicates that they fomi identical soluble aluminum complexes (Figure 28 ).

250 300 350

Wavelength, nm

Figure 28 Wavelength scans for filtrates obtained from AISPTA and AlPTA systems

aged for 90 minutes (pH s 7.5).

In order to ensure that the high observed in AISPTA was not associated with

an interaction between components in the mixture of TA and phosphate pnor to the

aging period, wavelength scan of a mixture of TA and phosphate was obtained (Figure

29). As can be seen from Figure 29, the spectnim of the mixture is identical to that of TA

alone,

Figure 29. Cornparison of wavelength scan for a mixture of TA and phosphate with

that for a solution of TA alone: (a) TA (17.0 ppm) alone and (b)

Phosphate (5.0 ppm) and TA (17.0 ppm).

Both backscattered electron and elemental mapping images of samples from

AIPTA and A15PTA systems were obtained using the electron microprobe technique.

The images are given in Appendix A. The images did not provide any useful information

that could account for the difference in the reactivity of the solids. However, mapping of

the variation in the composition of phosphonis and aluminum in the solids showed that

the elements were homogeneously distributed in the solids.

Differenœs in the effect of the presence of TA or 1 and phosphate on the surface

reactivity of solids obtained from wprecipitation and postprecipitation reactions are

noteworthy and are summarized in Figure 30a and b- In most systems, increasing age

resulted in a decrease in surface reactivity of the solid hydrous aluminum species, which

probably indicates a reduction in the specifk surface area of the aging precipibte due to

its gradua1 transition to a more orderiy structure. Exceptions occurred M e n various

inorganic precipitates were subsequently exposed to TA - in particular, the systems

A15TA, AIP5TA and AISPTA. which became more reactlve as the precipitate aged. In

these latter systems, the presence of organic coatings on the surface of these partscies

may be postutated to explain the significantty different reactivity of the soiids. Further

evidence for this is presented in Section 3.2, in which the coating pmperties of the

particles were examined using AFM and IFM techniques-

I I S m i n Cl 30 min 160 min

Al AITA ASTA AIP AJSP System s

Figure 30a Cornparison of surface reacüvity (measured as b) of hydrous aluminum

oxide CO- and postprecipitated in the presenœ of either phosphate or TA.

Al N A 5 P AJPTA AiP5TA ASPTA

Systems

Figure 30b. Comparison of surface reactivity (measured as b) of hydrous aluminum

oxide CO and postprecipitated in the presence of both phosphate and TA-

3- 7.3. Residclal phosphoms and TA

Data on residual phosphorus and TA are presented in Figures 31 and 32-

Phosphorus was removed to a greater extent when coprecipitated rather than

postprecipitated in the presence of prehydrolyzeâ solid, For example, more than as

much phosphorus was removed as AIP (Figure 31a) compared to Ai5P (Figure 31b) at

the aging periods investigated. Where phosphorus and TA were present during

coprecipitation (AIPTA), the cornpetition between them interfered with the removal of

both carnponents but TA appeared to have a larger inhibitory effect on phosphorus than

did pttosphonis on TA. The cornpetition was evident in post-precipitaüon expen'ments as

well. When TA was added to the 5 min aged coprecipitate of aluminum and phosphate,

the % phosphorus removal declined, with more phosphorus going back into solution as

the precipitate aged.

O IO 20 30 40 50 60 70

Phosphorus remaval (%)

.AIP UAPSTA mAPTA

Figure 31a. Phosphoms removal (%) in coprecipitation systems. In the case of

AIPSTA, TA was added after phosphorus had been removed by

coprecipitation,

O 2 4 6 8 I O 12 14 16 18 20 Phosphorus remmal(%)

mN5P O AKPTA iAlfA5P

Figure 32a. Tannic acid removal ( % ) in coprecipitaüon systems. In the case of

AiTAoP, phosphate was added after TA had been rernoved by

coprecipitation,

tannic acid removal (%)

Figure 32b. Tannic acid removal (%) in postprecipitation systems.

While TA removal was adversely afiected when TA and phosphate were

simultaneously caprecipitated, removal of the organic material was enhanced when

precipitated ont0 an AIP ramer than aluminum solid. Furthemore, when phosphate was

added to a 5 min aged wprecipitate of aluminum and TA, the amount of TA that was

removed from solution reached a maximum after 30 min of aging the mixture.

60th phosphorus and organic removal were much less efficient in the AlPTA

system mmpared to the other phosphorus- and TA-containing systems. This, in large

part, is due to the fine nature of the precipitate. It was in this system that the femn test

indicated that some colloidal matter was present in the filtered sample (Figure 21).

In general, phosphorus was better removed by coprecipitation while TA was

usually more efficiently taken from solution by precipitation onto a prehydroyzed solid. Of

al1 the cases, the AIP5TA system gave best compromise results for phosphorus and

organic removal. This finding has practical implication in wastewater üeatment plants

where coagulants are used to remove both phosphorus and organic matter. The addition

of coagulant at the aerator exit where the concentration of organic matter is relatively low

should be effective in removing (by coprecipitation) any inorganic phosphate that

remains in the waste stream. The sludge produced at this point would consist in large

part of aluminum hydroxyphosphate and hydrous aluminum oxide- This sludge, when

recycled into the aerator, should be capable of efiïciently removing soluble organic

matter by postprecipitation ont0 the solid precipitate. The simultaneous biological

processes, in the aerator add to the effkiency of removal of both organic matter and

phosphorus.

The influence of concentration of TA on phosphorus removal was examined for

the AIPTA system. As shown in Figure 33, increase in the concentration of TA (decrease

in aluminum: TA molar ratio) resulîed in poorer removal of phosphorus from the

simulated wastewater (53% rernoval where there was little or no TA to 6% removal in the

presence of 17 ppm TA). In part, the poorer removal efficiency is due to the increase in

solubility of aluminum when complexed with the organic ligand. This was indicated by

spectmphotornetnc measurernents on the filtrate, which revealed a peak due to soluble

complexed aluminum ion at 318 nm- The peak at 3f 8 nm increased as the concentration

of TA in the system increased.

- 9 s o - E 2 4 0 - rn 3 ô 30- C P g 20-

10- s

O 0.02 0.04 0-06 0.08 O, 1 0.1 2 0-14 0-16 0.18

Alurninum : TA rndar ratio

Figure 33. Influence of increasing TA concentration on phosphonis

removal: AlPTA system, aged for 30 min.

3.1.4. Summary

The information obtained from the ferron test. FTlR spedmscopy. phosphonis and TA

removal data can therefore be surnmarized as follows:

1. Ferron Test

Significantly different solid phases were identified in the various precipitates. The

nature and surface reactivity of the products varied mnsiderably according to the

constituents of the solution, the presenœ and concentration of TA during aluminum

hydrolysis, the sequenœ of anion addition and the duration of aging of the precipitate

pnor to exposure to phosphorus and / or TA-

In most systems, increasing age resulted in a decrease in surface reactivity of the

solid hydrous aluminum specks. The decreased readivity observed was attributed

to a reduction in the specific surfaœ area of the aging precipitate due to its gradua1

transition to a more orderly solid. Exceptions occurred for systems such as AISTA,

AIPSTA and AISPTA, which becarne more reactive as the precipitate aged, AIPTA

systems on the other hand. did not show significantiy changed surface reactivity after

a 30 minute agîng p e M -

The great differences in surface reacüvity of postprecipitation systems such as A15TA

and AISPTA compared to the corresponding coprecipitation systems, AlTA and

AIPTA respectively, was ambuted to the presence of an organic mating on solids

obtained from the former systems, which prevents the feron reagent from directiy

contacting the aluminum atoms at the surface of the inorganic solid. The inhibitory

effect was especially important over short aging pefiods. For the coprecipitation

systems (AITA and AIPTA) it appears that TA interacted directly with the added

aluminum ion as hydrolysis was occumng, and formed soluble complexes, which

sewed as precursors for a solid phase-

2. FTlR spectroscopy

The broad relatively featureless infrared absorption bands in both the -OH stretching

and deformation regions indicate the amorphous nature of the freshly precipitated

solids.

Absorption bands that could be attributed to AI-0 vibrations of alurninum in

octahedral coordination were observed in the region between 400 - 650 cm".

Precipitates aged in the presence of phosphate showed P-O stretching vibrations

between 1100 and 1040 cm-'. Spectra of solids from AIP system were nearly

identical to that of wavel lite (A13(OH)3(P0&.5Hfl) indicating that the solid from AIP

systerns is an aluminum hydroxyphosphate.

Absorption bands indicating the asymmetric and symmetric stretches of carboxyiate

bound to aluminum appeared at about 1594 i 4 cm-' and 1368 t 4. This implies that

the COO- functional groups in TA were the active chemical coordination sites dumg

the interaction.

3. Phosphorus and TA removal data

The most efficient removat of phosphorus occurred during the coprecipitation of

phosphorus and alurninum ions without organic matter

The efficiency of TA removal is greater by postprecipitation ont0 prehydrolyzed Al

rather than by coprecipitating it with the coagulant

The presence of TA, especially at high concentration, inhibited phosphorus removal

dunng the coprecipitation of aluminum, TA and phosphate, which is synonymous

with the addition of alum before or in the aerator.

Relatively more phosphorus was removed when phosphate was added to

prehydrolyzed Al than when added as a mixture of TA and phosphate.

Tannic acid inhibited phosphorus removal more when added as a mixture with

phosphate ont0 prehydrolyzed Al than when added to a coprecipitate of aluminum

ion and orthophosphate. Aging of the system did not influence this trend-

Solids obtained from coprecipitation of aluminum and phosphate showed a greater

rernoval of TA than solids from prehydrolysis of aluminum ions

3.2. Atomic force microscopy (AFM) and interfacial force microscopy (IFM)

studies: Evidence for organic coatings on postprecipitation products

Atomic force microscopy and interfacial force microscopy were used to study

changes in the surface properties of the solids and to provide evidenœ for organic

coatings on the AISPTA solid.

3.2- 1- AFM sfudies on hydtwus afuminum oxides p119cr;Pifateâ un&r vanous

wastewaterconditions

For the AFM studies, AFM images of TA and of solid phase hydrolyüc

products from the following different systems were acquired:

Al system (hydrous alurninum oxide particles used as control)

AIPTA (coprecipitated particles)

AISPTA (postprecipitated particles)

In the AFM images, (i) the left image shows height mode data, m i le the right

image shows phase imaging data (ii) dark regions in the images correspond to lower

values of height and phase shift, white brighter regions correspond to higher values.

Tannic acid dispersed on mica

The AFM images of TA deposited on a mica substrate from methanol are

given in Figure 34. The images reveal that TA foms agglomerates of vanous sizes.

To gain insight into the physical characteristics of TA, a small region of the sarnple

was imaged at higher magnification (Figures 35 a - d). The images (Figure 34)

reveal that TA was present on the mica as varying sized agglornerate of particles.

The cross sectional profile indicates that the diameters of the small agglornerates

were about 15 nm while those of large agglomefates ranged from approximately

300 to 600 nm. Figure 36 depicts the cross sedionat profile of an isolated particle of

TA whose height is 1-96 nm. The Figure also shows a single large agglornerate,

probably consisting of isoiated particles of varying height

To assess the effect of decreasing TA concentration on the agglomerates,

images of various I owr concentrations of TA were also acquired- As the

concentration of TA was decreased, the frequency and size of the larger

agg lomerates decreased (images not shown). This observation seems to indkate

that agglomerates of TA consist of particles that can exist as isolated particles.

especially when deposited from dilute solution. Agglomerates of fulvic acid have

also been idenüfïed using tapping-mode AFM lQ and in transmission electron

microscope (TEM) images of cuncentrated solutions of fulvic acid. At low

concentrations, sponge-like structures consisting of rings about 15 nm in diameter

as well as small spheres ranging in diameter from 10 - 50 nm were obsenred in the

AFM images. Although the sponge-like structures were not reported in the TEM

study, the conflicting report has been attributed to variations in the methods of

sampie preparation

A closer look at the images acquired at setpoints between 0.90 and 0.30

while maintaining tip oscillation amplitude of lû6 nm shows contrast reversal

between height and phase imaging (Figure 35 a and b). Images acquired at high

setpoint (0.90) and lower tip oscillation amplitude show considerably less wntrast

(Figure 35 c and d). Contrast reversai, especially in AFM phase images, can be

related to stiffness variation between the sample and substrate or within

multicomponent materials 115, 116. 117 - A strong interaction between the tip of the

sensor and the sarnple was observed during the acquisition of the AFM images of

TA. Like fulvic acid partides '42, TA particles occasionally adhered b the sensor tip

and were subsequentiy deposited elsewhere on the surface of the substrate-

Explanations of these obsewations are presented during the discussion of

postprecipitated particles, where similar effects were obsenred.

Figure 34. AFM images of TA adsorbed on a mica substrate- Height (left)

(2 range = 25 nm) and phase imaging (right). The image is 2.0 pm

square and was acquired at A, = 106 nm. and r, = 0.90.

Agglomerate of particles of various sizes can be seen in the height

image-

Figure 35. AFM images of TA adsorbed on a mica substrate. Height (left)

(Z range = 10 nm) and phase imaging (right). The image is

500 nm square and was acquired at (a) A, = 106 nm, and r,, = 0.90 and (b) A, = 1 û6 nm, and r,, = 0.30.

Figure 35. AFM images of TA adsorbed on a mica substrate. Height (left)

(Z range = 10 nrn) and phase irnaging (right). The image is 500 nm

square and was acquired at (c) A. = 52 nm, and r, = 0.90. and (d)

A. = 26 nm, and r, = 0.90.

Distance, pn

Figure 36. Cross sectional profile analysis data for TA- Several small

particles appear as "spikes" on the cross sectional profile. The largest height

(1 -96 nm) is for the particle on the left side of the profile. The single large

agglomerate on the right side appears €0 be made up of small particles of

varying heights. The profile at the top is along the line shown in the image

below.

Al system particles (used as control) on mica

Particies fmm the Al system are hydrous aluminum oxide precipitated in the

absence of TA or phosphates, Images of these particies were acquired at setpoints

ranging from 0.90 - 0.15 and amplitudes from 20 - 105 nm- An AFM image of control

particles on the mica substrate at intermediate oscillation amplitude of 58 nm and set

point of 0-50 is shown in Figure 37, The particles are roughly spherical in shape-

Cross sectional analysis of the partictes was carried out The particles Vary widely in size

and height with diameters from 40 - 500 nm and height between 10 and 150 nm-

Figure 38 depicts a typical cross sectional profile-

The amtrol particles showed virtually no contrast with respect to the substrate in

the phase imaging mode at large to intemediate values of cantilever tip oscillation

amplitude and moderate setpoints. The slight shading at the particle boundaries is

probably due to the failure of the feedback loop to adequately follow the rapid changes in

sample height as the tip tracks across the particles. During the acquisition of the images,

tip-sample interaction was never observed,

Figure 37. AFM images of the control particles dispersed on a mica substrate. Height (left) (2 range = 200 nrn) and phase Ïmaging (right) The images

are 2 pm square and were acquired at A. = 58 nrn and r, 0.50.

Distance, pm

Figure 38 a. Cross sectional profile for particles from Al system used as contmi,

showing that the height of the particle highlighted on the image is 24.0

nm. The profile at the top is akng the line shown in the height mode

image below the profile data-

Distance, pm

Figure 38 b. Cross sectional profile for particles from Al system used as control,

showing that the height of the particle highlighted on the image is 124-3

nm. The profile at the top is along the line shown in the height mode

image below the profile data.

Larger scale AFM images of the control particles were also acquired and

consistentiy demonstrated that the pattern seen in Figure 38 is typical of that observed

across the surface-

The AFM images of coprecipitated AIPTA particles deposited on the mica

substrate at different set points ranging from 0.90 to 0.25 were acquired. Figures 39 a - c

show the images acquired at the same set point of 0.65 but at different oscillation

amplitudes of 116, 58 and 29 nm. The coprecipitated AIPTA parücles are approximately

spherical in shape- Cross secüonal analysis (Figure 40) of the partides was camed out

and show that these particles Vary widely in size and height The height of the particles

range from 17 - 150 nm and the diameters from 40 - 500 nm. These data are similar to

those obtained for the control particles-

Interestingly, as the amplitude was decreased, the height and phase images for

coprecipitated AIPTA particles appeared more similar to the images of the control

particles. At large to intermediate values of cantilever tip oscillation amplitude and

moderate setpoints, the coprecipitated AIPTA particles and the mica substrate show

sirnilar phase shift. No tip-sampie interaction was obsetved during the imaging of these

particles. These observations are similar to those reported for hydrous alurninurn oxides

particles used as control implying that both the coprecipitated materials and the hydrous

aluminurn oxide particles used as a control have similar viscoelastic pmperties as the

mica substrate. This finding is not surprising since similar surface structural groups are

present on both particles. In fact, octahedral alumina species are a basic sub-unit of both

h ydrous aluminum oxide and mica, Mich is an alumino-silicate mineral,

Figure 39. AFM images of wprecipitated AlPTA particles adsorbed on a mica

substrate. Height (left) (Z range = 200 nm) and phase imaging (nght).

The images are 2 pm square and were acquired at (a) A. = 116 nm, and

ri, = 0.65.and (b) A, = 58 nrn, and r, = 0.65.

Figure 39, AFM images of coprecipitated AlPTA particles adsorbed on a mica

substrate. Height (left) (2 range = 200 nm) and phase imaging (right).

The image is 2 pm square. It was aquired at A, = 29 nm, and r,, = 0.65.

Distance, pm

Figure 40 a. Cross sectional profile for coprecipitated AIPTA partides showing that the

height of the particle highlighted on the image is 54.6 nm. The profile at

the top is along the fine shown in the height mode image below the profile

data.

m C c -

Distance, pm

Figure 40 b. Cross sedional profile for coprecipitated AIPTA parücles showing that the

height of the parade highlighted on the image is 81.5 nm. The profile at

the top is along the Iine shown in the height mode image below the profile

data.

Postpreci@ifated AISPTA particles

Six AFM images showing postprecipitated AISPTA particles dispersed on

the mica substrate are displayed in Figures 41 - 43. These images were acquired

at different values of both cantilever tip oscillation amplitude (22 -104 nm) and

setpoint (0.60 - 0-90)- The image size and z scale (i-e- height or phase shift) are

the same for each image. In al1 cases, dark regions correspond to lower values

of height and phase shift, M i l e brighter regions correspond to higher values-

Larger sa le images, up to 20 pm square, demonstrate that the pattern seen in

Figures 41 - 43 is typical across the sunace.

A close examination of the AFM images indicates that in al1 cases the

contrast between height and phase shift data is reversed. However, the range of

contrast varies considerably for both data types- The greatest contrast is

discernible at large tip oscillation amplitude and intemediate set points (Figure

41 b) and the least contrast at low tip oscillation amplitude and large set points

(Figures 42 a and b). A simiiar trend in contrast intemediate ta those shown here

was observed for images acquired at intemediate set point values at each of the

tip oscillation amplitudes reported above.

Figure 42 a and b are images of the same area, taken at an interval of

about A5 min. Interestingly, the surface features indicated with the circles in

these figures change slightly from one image to another. The changes are

indicative of tip-sample interaction and this effect was observed in a number of

cases when data were acquired in sequenœ at exacüy the same set point and

amplitude. To investigate this effect further, a small region (500 nm) of the

sample was imaged using confacf mode AFM. The contact mode AFM data did

not show any evidence of the particles remaining on the surface.

Furthemore, reimagnig the same area in tapping mode revealed tha the

particles had been swept out during the contact mode imaghg. A similar effect

was observed with the TA sample but was not observed during the imaging of

the control and coprecipitated AlPTA particles-

b Figure 41 - AFM images of the postprecipitated AlSPTA particles dispersed

on a mica substrate. Height (Mt) (Z range = 20 nm) and phase

imaging (right) Images are 2 pm square and were acquired at A, of

104 nrn and at the following r,, (a), 0.90 and (b) 0.60.

Figure 42. AFM images of the postprecipitated AISPTA parücles dispened on

a mica substrate. Height (left) (2 range = 20 nm) and phase

imaging (right) images are 2 prn square and were aquired at A,

of 46 nm and at the following r.,: (a) 0.90 and (b) 0.75.

Figure 43 AFM images of the postprecipitated AISPTA particles dispersed on

a mica substrate. Height (left) (Z range = 20 nm) and phase

imaging (right). Images are 2 pm square and were acquired at A.

of 22 nm and at the following r,, (a), 0.90 and (b) 0-70,

In otder to evaluate the characteristics of the particles, tM, fundamental

questions rnust be addressed. Which component in the heigM images shown in

Figures 41 - 43 corresponds to the particles and which represent the substrate?

Secondly, how many layers of particles are on the surface of the substrate?

Contrast variation between particles and their surroundings in both height

and phase images has been shown to depend on experimental conditions such

as tip oscillation amplitude and setpoint ratio, although the relationship is

complex lrsn '"* l'? In tapping mode AFM, the height images acquired at hgh

values of setpoint should approximately reflect the actual sample topography Il6.

"7. lmaging at lower set points (in which the tip penetrates far into the sarnple)

can result in sample defomation, especially in the case of soft samples. When

this occurs, the feedback mechanism indicates that this more cornpliant region is

lower in the height mode image and such information is misleading. The images

presented in Figures 41 a, 42 a, and 43 a are al1 acquired at high setpoint (0.90).

It is obvious that the raised components in these images also appear raised in

the remaining images (Figures 41 b, 42 b, and 43 b). These obsewaüons

indicate that the raised objects are the real particles- Since the postprecipitated

AISPTA particles appear as raised components in al1 the height images of the

sample, the particle height andfor the sample stiffness rnust be suffîciently large

to cancel defomation effects at lower set points-

In order to answer the second question, we need to carefully examine the

phase imaging data. The phase shift data dernonstrate that the particles on the

surface of the mica substrate consist of a single, incomptete layer of particles, If

more than one Iayer of particies were deposited on the mica substrate, the phase

imaging data should show little contrast since the material on the surface is

expected to have the same sample stiffness. It can be seen from the phase

images presented in Figures 41 and 42 that the parocles show a negative

excursion, which means that they possess a lower modulus than the mica

substrate. Contrast inversion has been reported in studies on polymer surfaces

where they were generally obsenred at very low values of setpoint ratio 116. 144

Another interesting observation was the instability of the feedback loop at low set

points dunng the AFM data acquisition on the postprecipitated AISPTA samples.

This could be attnbuted to very strong tip-sample interaction as the tip penetrates

the surface layer- The effect was not obsenred during the acquisition of AFM

data on the control or coprecipitated AlPTA particles-

Having identified the particles in the images, cross sectional analysis was

then carried out. Examples of the results obtained are shown in Figures 44 a and

b below, Fmm the cross sectional analysis data on the samples, the

postprecipitated AISPTA particles were found to be uniforrn in height

(2.05 + 0.21 nm). The height rneasurement was taken at high set point and tip

oscillation amplitude, which should give the closest approximation to the actual

height of surfâce features. The particles Vary widely in their lateral dimensions

and are extremely flat They are irregular in shape and may be considembly

smaller in size than either the hydrous alurninum oxide particles used as the

control sample, or the coprecipitated AIPTA particles. The solids in these images

are probably not individual particles, but rather aggregated particles of relatively

uniform size, which are loosely bound to the substrate and to one another. AFM

data on the size and nature of these particles is consistent with earlier results

O btained using transmission electron microscope 13'.

Distance, pm

Figure 44 a. Cross sectional profile of postprecipitated APTA particles showing that

the height of the highlighted particle on the image is 1.8 nm. The profile at

the top is along the line shown in the height mode image below the profile

data.

Distance, pm

Figure 44 b. Cross sectional profile for postprecipitated AlSPTA system showing that

the height of the highlighted particle on the image is 2.1 nm. The profile at

the top is along the line shown in the height mode image below the profile

data-

As can be seen by comparing the phase shift images, the phase contrast

behavior of TA and AISPTA particles are similar and also quite different from the mntrol

and AlPTA particles. Tannic acid and the AISPTA partides show signifiant changes in

contrast both in the phase and height data as a function of setpoint and tip oscillation

amplitude. This observation indicates that the postprecipitaion particles are wated with a

layer of TA. The interaction between TA and the prehydrolysed aluminum oxide surface

probably involves the hydrophilic (phenolics, carûoxylic and ester fundional groups) part

of TA, thus resulting in the formation of a hydrophobie-like material. The interaction

between the postprecipitated material and the hydrophilic mica would be expected to be

weaker than the interaction between the hydrophilic mica substrate and hydroxyl

terminated coprecipitated AlPTA and contra1 particles which themselves appear to be

much more strongly bound to the substrate.

Influence of aging on postpmcipitated parficIes

From the ferron test results Fable IO), it was noted that the values for the

pre hydrol yzed particles onto which TA had been postprecipitated decrease with ag ing .

The enhanced reactivity as they aged suggests that the coating on the particles was not

a permanent feature. If the wating effect is temporary, how does the aging process

affect the viscoelastic properties of the solids? Does aging of the postprecipitation

particles in the original solution result in further growth of the particles? Does the

decrease in with aging observed in the ferron test results (Table 10) imply that the

particles would revert ta a fom of precipitate that is similar to the coprecipitated product?

An attempt was made ta find answers to these questions by using the AFM

technique to monitor changes in viscoelastic properties of AISPTA particles during aging.

AFM images were acquired on particles aged for 1 h and 24 h. The 5 pm image of

precipitate that had been aged for 1 h (Figure 45) shows aggregates that are sphericat in

shape and differ wldely in size. Images obtained at lower magnification are depided in

Figures 46 a and b- In these images, the small particles in the background appear to

fom patterns simifar to those observeci in images of the 5 min aged samples- Unlike the

particle aged for 5 min, however, these ones are not uniform in height and are larger in

average size. A cross sectional profile taken under conditions of high setpoint and tip

oscillation amplitude indicates that the particle height ranged from 4.9 - 55-8 nm with

diameters ranging from 31 - 117 nm. The difference in height and size between the 5

min and 1 h aged AlPTA partides suggests that stnictural transformation and growth

occurred during the 1 h aging period- Furthemore, the 1 h aged particles showed a

negative excursion in the phase image, indicating that they are still wated with an

organic layer of TA.

Figure45- AFM images of the postprecipitated AISPTA particles (1 h aged)

dispe~ed on a mica substrate. Height (left) (Z range = 70 nm) and phase

irnaging (nght). Images are 5 pm square and were aquired at A, of 1-30

nm and at r,, of 0.88.

Figure46. AFM images of the postprecipitated AISPTA particles (1 h aged)

dispersed on a mica substrate. Height (left) and phase imaging (right)-

The 2 pm square images are images of different regions of the sample

and were acquired under identical conditions. (2 range = 50 nm, hign A,

and r,,).

Results for the aging of postprecipitated particles for a period of 24 h are given in

Figure 47 and 48. These images were acquired at both low and high magnification.

Figures 47 a - b shows a 5 pm images of different region of the 24 h aged particles while

Figures 48 shows a 2 pn image typical of the surface of the coated mica substrate. One

important feature of the height images of the 24 h aged particles is the absence of the

small Rat particles in the background that could be observed in the images of the 1 h

sample- Ail the particles in the more aged material are well aggregated- However, the

figures indicate that the particles are coated as they still show a negative excursion in

the phase images indicating that the particles are softer than the mica substrztte. A cross

sectional profile taken under conditions of high setpoint and tip oscillation amplitude

indicates that the particle heights are in the range from 10.4 - 120.5 nrn with diameters

ranging from 46.9 - 136.7 nm- The values for height and diameter are still much lower

than those obtained for the controt and coprecipitated particles, although the particles

look similar in shape. In addition, the dimensions are slightly greater than those of the

1 h aged particles-

Figure 47 a- AFM images of the postprecipitated AISPTA particles (24 hour aged)

dispersed on a mica substrate. Height (left) (2 range = 40 nm) and phase

imaging (right)- Images are 5 pm square and were acquired at A,, of 1-12

nrn and at rsp of 0.92.

Fig i Ire 47 b. AFM images of the postprecipitated AISPTA particles (24 hour aged)

dispersed on a mica substrate. Height (left) (Z range = 40 nm) and phase

imaging (right)- images are 5 pm square and were aquired at A, of 1-12

nm and at r,, of 0.92.

Figure 48. AFM images of the postprecipitated AISPTA particles (24 hour aged)

dispersed on a mica substrate- Height (left) (2 range = 30 nrn) and phase

imaginç (right). Images are 2 prn square and were acquired at A, of

1.12 nrn and at r,, of 0.92.

3 2 l l Summary

In overview, using Atornic Force Microswpy it was possible to demonstrate that

the surface of AISPTA solid had remarkably different viscoelastic pmperties than the Al

and AlPTA solids- The latter two materials behaved in a manner similar to the mica

substrate, a relatively hard minerai- On the contrary, the AISPTA surface had

characteristics of a 'soff, organic material - characteristics that were similar to those of

pure TA. Ai5PTA showed sïmilar change in contrast both in phase and height data, as a

function of set point and tip oscillation amplitude, The particte size and shape of the Al

and AIPTA solids are similar, but are different from the particle size and shape of the

AIBPTA solid. Compared with the 5 min aged AIBPTA. the viscoelastic properties of the

sarne matenal aged for 1 and 24 h suggest the continuing presence of an organic

coating on the particles. However, there are differences in particle shape, height and

diameter that indicate that growth has been occurn'ng-

3.2.2, inferfacial force rnicmswpy (IFM) studies on hyîmus aluminum oxides

precr;Oitated under mtfous wasfewaterconditions

In order to obtain quantitative information on the viscoelastic properües

associated with the images obtained by AFM, several indentation experiments were

conducted on the controi, coprecipitated AlPTA and postprecipitated AISPTA particles

using IFM. Similar measurements were also made on the mica surface used as the

substrate for the particles. lnterfacial force microscopy contact-mode images of the

particles and mica substrate were acquired before, dunng and after the indentation

experiments. Unlike the AFM technique, duting IFM imaging of a sample, features of

interest can be selected and the viscoelastic properties subsequently determined with

nanoforce resolution l las ltl- le. The IFM contact mode topograp hic image of the particles

and mica surface acquired after an indentaion experiment are shown in Figures 49 a - d,

The particles in these images are not identical to those observed in the AFM images

presented in the Section 3.2.1 because the resoluu'on of IFM images are usually lower

than the resolution of AFM images. Factors that could account for the difference in

resolution include the radius of the IFM probe which is much larger than the radius of

tips used in the AFM tapping mode technique, the enhanced capability of the AFM

Nanoscope II€ controller to collect more Iines per scan during the acqusition of images

of the sample, and the effect of deformation of campliant particles under repulsive load

during contact mode IFM imaging '"- . It has been noted that the IFM and AFM

techniques are complimentary tools because IFM offers more superior quantitative

information on viscoelastic properties than AFM while on the contrary, the AFM offers

superior imaging capabilities

Figure 49. Contact mode IFM image (a) ffeshly cleaved mica surface

(6 pm x 6pm), (b) control particles (6 pm x 6pm)- (cl coprecipitated AIPTA

particles (12 pn x $2 pm) and (d) postprecipitated AISPTA particles -

(12 prn x 12 pm )-

Several force curve experiments were carried out on the substrate and the

particles and typical force displacement (F-D) cunres are presented in Figures 50 - 53.

The arrows in the Figures are used to show the full interaction cycle during the

experiments. This cycle starts with the movement of the sensor from a rest position

(sensor approach) until it is in contact with the sample and ends with the removal of the

load (sensor retract)- Force curves have been used to extract meaningful information on

adhesive monolayer film-tip interaction ''O* 12' and elastic moduli of different types of

samples 118. 146 - As illustrated in Section 2.12, the slope of the unloading portion of the

F-D curve can be modelled using Hertzian mechani~s"~~ '" and the elastic modulus

calculated- This procedure has been used in several studies 118. 145. 146.148.149

Inspection of the approach portion of the F-D curve in Figure 53 a for the

postprecipitated AISPTA particles reveal the presence of two distinct slopes. This feature

is typical of the postprecipitated particles but not observed in the F-D curves of control

and coprecipitated AIPTA particles. Figure 53 b is an enlargment of the region showing

cleariy the the siope at the point of contact between the tip of the sensor sample (points

between A and B in Figure 53 a ) which indicates that the tip was probably in contact

with a compliant layer. The presence of this feature in the postprecipitated particles but

not in the control particles suggest that the former particle surface is compnsed of a

bilayer. The small slope upon contact in the F-D curve of postprecipitation samples

means a low reduced modulus for the compliant layer and the large siope a high

reduced modulus for the prehydrolyzed inner core of the particle. This is because the

Hertzian model indicates that the applied load (F) is proportional to three-halves

power of displacement (D) when the tip makes contact with the sample.

F = E* D ( where E* is reduced rnodulus of the system) (1 9)-

In Figures 50 - 53, the shape of the approach portion of the F-D curves is

reminiscent of Herbian (or elastic) behavior; The Her&ian fits for the intial (or elastic)

cornponent of the F-D curves obtained were undertaken in order to detemine the

slope(s) for each curve- An example of Hertzian fit for the approach portion

corresponding to contact between the sensor and the cornpliant organic is shown in

Figure 53 c. Typical examptes of Herttian fk for the inner inorganic core of the

postprecipitated and the control particle are shown by the dark lines in Figures 54 d and

51 b respectively- From the values obtained for the siope of the fits, moduli of the

particles were detemined as descnbed in the experimental section of this thesis. The

reduced modulus of the organic layer was found to be 0-17 + 0.07 GPa while values

for the mica surface, control particles, coprecipitatbn particles and postprecipitation

particles were calculated to be 15 + 2, 12 t 3 , 18 t 2, and 15 + 4 GPa respectively.

These results indicate that the mica substrate and the particles have similar elastic

modulus. The only exception was the reduced modulus of the organic layer on the

postprecipitation particles, which was significantly lower.

Literature values of reduced modulus data for hydrous aluminum oxide were not

available for cornparison with our data. However, the calculation of reduced modulus for

aluminum oxide (a-Al& ) and biotite (one form of mica family) were camed out using

the information listed on Table 12 and the expression '":

Where Et and E, refer to Young's modulus, V, and V, indicate Poisson's ratio and

the subscipts t and s denote tip and sample-

Table 12 Young's modulus and Poisson's ratio data*

Material Young's modulus, GPa Poisson's ratio

Tungsten 350 - 400 (375) 0-3

AIuminurn oxide 228 - 280 (254) 0.236

Mica (Biotite) 40 0.225

( ) denote mean value used for the calculation.

'Sources: Handbook of Physid Quantites. Grigoriev, 1. S. and Meilikhov. E. 2. Eds.

CRC Press Inc. USA. 1997; Touloukian, Y.S. and Ho, C. Y. In Physical Pmperffes of

Rocks and Minerztls, McGraw-Hill, New York, 1981 ; Crotell, A, H- In The Mechanical

Propedies of Matfer; John Wiley and sons, New York, 1964,

The calculated reduced moduli for a-A1203 and mica were 162 and 38 GPa

respectively - values that are clearly higher than our experimental values. Unlike the

amorphous precipitated particles. however, a-Al203 is known to be very hard ''O. The

calculated modulus for mica is about twice as large as the experimental values. The IFM

tip used in this study has a parabolic geometry and dunng the indentation experiments,

both normal and parallel forces would influence the rneasurement of the modulus of the

sample. Indentation of the sample normal to the surface would require a much larger

force compared with indentation parallel to the sampIe surface. Thus, the additional

influence of parallel forces due to tip geometry could accaunt for the observed difference

between the experimental and calculated moduli for mica.

Distance (nm)

Figure 50. F-D curve obtained from indenting freshly cleaved mica surface-

1 I

30 40 Distance (nm)

Figure 51. F-D curve obtained from indenting control particles on a mica substrate

Figure 52.

Distance (nm)

F-D curve obtained from indenting coprecipitated AIPTA particles on a

mica substrate.

50 55 60 Distance (nm)

Figure 53 a. F-D cunre obtained from indenting postprecipitated AISPTA particles on a mica substrate.

50 Distance (nm)

Figure 53 b. F-O cuwe obtained from indenting postprecipitated AISPTA particles on a

mica su bstrate. In this enlarged figure the two distinct slopes indicating

that the particles consist of a bilayer is shown.

50 Distance (nm)

Figure 53 c. F-D curve of postprecipitated AISPTA particles on a mica substrate

showing the Hertzian fit for the dope of the organic layer.

50 Distance (nm)

60

Figure 53 d. F-D curve showing the Hertzian fit for the inner core of postprecipitatêd

AISPTA particles,

30 40 Distance (nm)

Figure 51 c. F-D curve showing the Hertzian fit for particles fiom AI system used as

3.2-2.1. Summary

Cornparison of changes in the force-displacement curves of three types of

particles similar to material found in wastewater and mica substrate shows that the

particles have sirnilar reduced modulus and suggest the presence of a bilayer in only the

postprecipitated particles, The bilayer feature implies that the postprecipitation particles

are coated with a layer of tannic acid, The F M results are consistent with the phase

imaging AFM data which indicate qualitative ditferences in the viscoelastic properties of

postprecipitated particles compared to mica substrate and the other particles

investigated. Thus, AFM and IFM have been used to provide for the fifst time both

qualitative and quantitative information about the surface morphology and viscoelastic

properties of hydrous aluminum oxides precipitated under wastewater conditions.

3.3. =AI MAS solid-state NMR spectroscopy

The influence of aluminum-phosphorus-organic matter interactions on alurninum

coordination in solids precipitated under various wastewater conditions was exarnined

using ''AI MAS NMR- Spectra of solids aged for 5 min and 30 d (43 200 min) were

acquired. As noted in the introductory part of this thesis, peaks wioiin chemical shift

ranges - 10 to + 20 pprn, and + 50 to + 80 ppm were associated with aluminurn

coordination in the octahedral and tetrahedral envkonments respectively lm* '03. '01 In

addition, peak(s) with intermediate chernical shïfts lm? '" '06 (+30 to +40 ppm) were

assig ned to either five-mordinate aluminum or distorted octahedraf aluminum

coordination, The preponderance of recent evidence points to these pea ks representing

five-coordinate aluminum and in the present work we shall use this assignment

Generally, the spectra obtained for al! the hydrolytic products revealed that

aluminum coordination was predominantly octahedral although two other minor

coordination environments, estimated collectiveIy as less than 15 % of total alurninum,

were evident. The rninor peaks were broad and strangly overiapping, and hence

attempts to resolve them using a deconvolution program called NUTS produced results

that must be considered only approximate.

The peak width of MAS NMR signals depends on the extent of molecular

order of the solid; the more ordered the soiid, the narrower the peak. Peak widths

determined for al1 the freshly precipitated and 30 d aged solids investigated were

generally greater than 10 pprn indicating that the soiids are amorphous. This result is

consistent with eariier work that reported peak vuidth greater than 10 ppm for amorphous

aluminum compounds and less than 10 ppm for crystalline cornpounds ".

Soiid phase pmducfs wntahing alumrnurn alone

The *'AI MAS NMR spectra of hydrous aluminurn oxide precipitated in the

presence of alkalinity after it had aged for 5 min (A15) and 30 d (A143200) are show in

Figures 54 a and b. Table 13 gives a summafy of the peak parameters. The chernical

shifts of the peaks in the spectnirn of A15 (Figure 54 a and Table 13) indicate that

aluminum is present in each of the airee coordination environments. The preàominant

signal is large and broad. It has a chemical shift of 4.2 ppm, characferisüc of

octahedrally wordinated aluminum (AiO& Two srnaIl broad peaks with chemical shifts

at 31 -9 and 63.3 pprn are atso present in the spectnrm. The signal a chernical shift

at 31 -9 pprn indicates aluminum coordination in a five-coordinate environment while the

signal at 63.3 pprn falls in the tetrahedral region of the 2 7 ~ ~ NMR chemical shift scale.

The tetrahedral and five-coordinate environments taken together aaount for less than

10% of the total aluminum content of the sample (Table 13)-

Figure 54 b presents the spectrum of the same precipitate aged for 30 d

(A143200). In this spectnim, a large and broad OctahedraI peak is located at a chemical

shift of 6.0 pprn and the visible 'remains' of a peak near 74.9 pprn strongly overlapping

the spinning side band can be seen. The broad octahedral peak appears to consist of

overfapping peaks indicating more than one kind of aluminurn coordination environment.

Figure 54 c displays the resolved overlapping peaks and the chernical shifts of the peaks

are indicated on Table 13. The two octahedral peaks in this spectnirn are at 4.4 pprn and

-8.9 ppm, which rnay show that the freshly precipitated solid is in the process of k i n g

transfonned into a gibbsite type of material. Gibbsite has been shown to exhibit two

octahedral peaks at 8.0 and -5.0 pprn 5'.

Figure 54 a. ''AI MAS NMR spectnim of hydrous aluminum oxide precipitated in the

presence of alkalinity and aged for 5 min (A15); [Al] = 430 mg L-'.

Figure 54 b. 2 7 ~ ~ MAS NMR spectrum of hydrous aluminum oxide precipitated in the

presence of alkalinity and aged for 30 d (AW3200); [Al] = 430 mg L".

Figure 54 c. 2 7 ~ 1 MAS NMR spectnim of A143200

coordination signals. [AI] = 430 mg L''.

-1 130 PP""

showing the resolved aluminum

Table 13 2 7 ~ ~ MAS NMR chernical shifts and peak anaiysis results for Al alone

system.

- -

SampIe Aging Peak Area C hemical Aluminum coordination

Perïod wïdth of shift environment

( P P ~ ) peak ( P P ~ )

(Oh) ---- - - - -

AIS 5 min 18.5 90.13 4.2

9-7 2.34 31 -9

40.0 7.53 63.4

Octahedral

Five-coordinate

Tetrahedral

A143200 30 d 6-0 Octahedral

11.0 31 -16 4.4 sh 'Octahedral

30.0 68.84 -8.9 'OctahedraI

- - - - - - - -

'Results from spectrum deconvolution program, Sh = sharp

Solid phase pmducfs mnfaining aluminum and phosphate

The ''AI MAS NMR spectra of solids obtained by coprecipitating aluminum and

phosphate for 5 min (AlP5) and 30 d (AlP43200) are show in Figures 55 a and b and

the peak parameters are presented on Table 14- 60th spectra reveal that aluminum is

predominantly octahedral, but again there is evidence of small amounts of tetrahedral

and fivecoordinate aluminum coordination. All the peaks in the spectra were shifted

upfield cornpared with those in the spectfa of aluminum alone- The major peak with a

chemical shift of 3 . 8 ppm is in a position that is typical for octahedrally coordinated

alurninum with phosphonis atoms as second neighbors- The second peak at a chemical

shift at 21.6 ppm is indicative of five-coordinate aluminum- The third peak with a

chemical shift at 44.9 ppm is indicative of tetrahedral coordinated alurninum with

phosphonis atoms as second neighbors- Similar NMR behavior has been observed in

AIP systems before ''. Upon aging, the three peaks were observed to shift downfield compared witti

those in the fresh material (Figure 55 b and Table 13). The downfield shift, a (possibly)

very srnall increase in the relative area of the octahedral peak caupled with a small

redudion in the relative areas of other peaks in the spectrum of AIP43200 ail suggest

that aging of the precipitate caused structural rearrangement and partial loss of the

incorporated phosphorus atom in the solid. This is supported by phosphate removal

data, which indicates that the percent residual phosphorus in the solution increased from

23.5 (AIP5) to 44.9 (AIP43200). Simultaneously, a rise ni pH of the aging solution from

7.6 (AIP5) to 9.6 (AIP43200) was observed. The higher residual phosphonis

concentrâtion on aging the sample for 30 d is probably pH related- However, this pH

value is much higher than typical pH of wastewater before and after treatrnent. During

aging, the peak width narrowed slightly from 19.6 pprn (AIP5) to 17.7 ppm (AIP43200)

consistent with transformation of the solid from a less ordered to a more ordered

matenal-

The chemical shifts of the peaks in the spectra of the AIP solids are in

significantly different positions from those reported for metavarisite (AlPO4.2Hfl) at

-13.2 ppm, and aluminum dihydrogen phosphate (AI(HzP04)3) at 46-6 ppm 51. 151- This

implies that the solids are not simple aluminum phosphates. The FTlR data presented

earlier in this work ako indicate the formation of solids similar to wavellite (a fonn of

hydroxyphosphate) rather #an aluminum phosphate. No NMR spectnim is available for

wavellite.

Solids fomed as a result of allowing phosphate to interact with prehydrolyzed

hydrous aluminum oxides (AISPS) also had three peaks (Figure 55 c and Table 14). The

chemical shifts of both the octahedral and tetrahedral aluminum in the spectmm of

A15P5 were somewhat downfield, away from chemical shifts obsewed for peaks in the

spectrurn of AIP. They are in positions intermediate between those for Al and AIP, which

would be expected since they contain less amount of phosphate, Phosphorus removal

data indicate that m i le the AIP system removed 76.5 % of the phosphorus in the

wastewater, only 43.0% was removed in the A15P system.

Figure 55 a. 2 7 ~ ~ MAS NMR spectnim of solids obtained from the coprecipitation of

aluminum and phosphate aged for 5 min (AIPS). [Al] = 430 mg 1-l;

[Pl = 500 mg L-'.

Figure 55 b. 2 7 ~ 1 MAS NMR spectrum of solids obtained from the coprecipitation of

aluminum and phosphate aged for 43200 min (AIP43200).

[Al] = 430 mg c': [Pl = 500 mg L-l.

Figure 55 c. '?AI MAS NMR spectrurn of solids obtained from postprecipitation of

prehydrolyzed aluminum and phosphate aged for 5 min (A15P5).

[AI] = 430 mg L-'; [Pl = 500 mg L-'.

Table 14, ?AI MAS NMR chernical shifts and peak analysis results for AIP systern.

Sample Aging Peak Area Chernical Aluminum coordination

Period width of shift environment

(PPW peak ( P P ~

- --

AIP5 5 min 19.6 86-43 -3-8 Octahedral

9-4 4.49 21 -6 Five-coordinate

9.2 9 .O8 44.5 Tetrahedral

AIP43200 3 d 17.7 88.56 -2.7 Octahedral

11.4 3.42 22.5 Five-coordinate

24.8 8.02 48-3 Tetrahedral

17.6 88.0 -0-4 Octahedrai

9.0 2.4 26 -0 Five-coordinate

32.4 9.6 53.2 Tetrahedral

Solid phase pmducts containing aluminum and TA

Figures 56 a and b show the "AI MAS NMR spectra of solids obtained from the

AlTA system aged for 5 min (AITAS) and 30 d (AITA43200) respectively. The peak

parameters are summarized on Table 15. Three peaks were observed in the spectnjm of

the AITA5 solid and these peaks were shifted slightly domfield compared with peaks in

the spectnim of A15 The peak width of the predominant octahedral peak of AITAS (23.2

ppm) was higher than the octahedral peak width for the A15 solid (18.5 ppm), which is

expected if the presence of TA in the solid hydrous aluminum oxide creates molecular

disorder in the precipitated material. The tetrahedral rasonance in AlTA system was

extremely weak.

When the material was allowed to age for 30 d, the spectnim of the solid

(AITA43200) showed signais similar to those reported for AITAS. (Figure 56 b and Table

15). The most significant effect of aging is the further broadening of the Peak- This trend

contrasts with the narrowing in peak wïdth observed on aging of solids obtained from Al

and AIP system - an effect expected m e n a solid is allowed to digest in the solution

from which it has been precipitated. The broadening obsewed in the peakwidth of peaks

in AlTA system on aging and the difference between the spectnim of AITA43200 and

A143200 are strong evidence that organic acids such as TA hinder the crystallization of

hydrous aluminum oxide. Another observable effect of prolonged aging was an increase

in the arnount of aluminum in the five-coordinate environment (Figures 56 a and b,

Table 15).

Figure 56 a. '?AI MAS NMR spectrum of solids obtained from coprecipitation of

alurninum and TA aged for 5 min (AITAS). [AI] = 430 mg L-';

Figure 56 b. '?AI MAS NMR spednim of solids obtained from coprecipitation of

aluminum and TA aged for 3 d (AlTA43200). [Al] = 430 mg L-';

Table 15. Z 7 ~ ~ MAS NMR chernical shifts and peak analysis results for AlTA system-

Sample Aging Peak Area Chernical Aluminurn coordination

Period width of shift environment

( P P ~ peak (PPW

AlTA 5 min 23.2 96.82 5.6 Octahedral

9.5 2-76 44.2 Five-coordinate

3-9 0-42 67.3 Tetrahedral - - -

AITA43200 30 27.0 93.23 5-3 Octahedral

14-0 4-88 40-1 Five-coordinate

17.2 1.89 70.0 Tetrahedral

The spectra of A15TA5 and A15TA432ûû are presented in Figures 57 a and b,

and Table 16 surnmarizes the peak parameters, The three peaks for AISTAS were

positioned between those reportai for AIS and AmAS- Aging caused a downfield shift of

the peaks, indicating a greater degree of interadion between TA and the inorganic

precipitate. After both AlTA and A15TA had b e n aged for 30 d. the NMR spectra were

very sirnilar. We take this as evidence to support our hypoüiesis regarding the

mechanism of rernoval of TA by post-precipitation wifh hydrous aluminum oxide- In the

first instance, the TA fomis a coaüng on the surface of the precipitate, and there is

minimal chernical interaction with the underlying inorganic material. As the precipitate

ages, however, it interacts more strongly wWIth the underiying material and becomes

chernically incorporated into the structure of the solid-

Figure 57 a. 2 7 ~ 1 MAS NMR spectnim of solids obtained from postprecipitation of

prehydrolyzed aluminum and TA aged for 5 min (Al5TA5).

[Al] = 430 mg L-'; F A ] = 1700 mg L-l.

Figure 57 b. *?AI MAS NMR spectnim of solids obtained from postprecipitation of

prehydrolyzed aluminum and TA aged for 30 d (A15TA43200).

[Al] = 430 mg L*'; F A ] = 1700 mg L-!

Table 16, 2 7 ~ ~ MAS NMR chemicai shifts and peak analysis results for AISTA

system-

Sample Additional Peak Area Chernical Aluminum coordination

aging width of shift environment

Period ( P P ~ peak ( P P ~ )

(%)

AISTA 5 min 21 -3 88.31, 5.1 Octahedral

12-1 9.83 34.7 Five-coordinate

24-9 1.21 67-3 Tetrahedral

AISTA43200 27.5 94-76 5.4 Octahedral

14.0 4.10 39.7 Five-coordinate

10.1 1.13 68.3 Tetrahedral

Solid phase pmducts confaining aluminum, phosphate and TA

As shown above, the presence of phosphate was associated with upfield

chemical shifts and narrowing of the octahedral peak width on prolonged aging whereas

the presence of TA was associated with downfield chemical shifts and broadening of the

octahedral peak width. In the system containing coprecipitated aluminum, phosphate

and TA, the s p W m of AIPTAS (Figures 58 a and b and Table 17) revealed three

peaks shifted downfield cornpared peaks in AIPS spednirn but upfield compared

with peaks in the spectra of both AI5 and AITA5. The upfield shift of the peaks cornpared

to peaks in AITA system and the phosphoms removal data suggest that phosphorus has

been incorporated into the network of the solid product The position of the peaks in the

spectrum of AIPTA5 relative to systems containing either phosphate or TA alone

suggests cornpetition between phosphate and TA for sites within the precipitate, with the

phosphate having the greater influence on the nature of the spectnim- However, on

allowing the sample to age for 30 d, the TA inhibited further incorporation of phosphates

into the solid. Evidently, the inhibitory effects of TA on phosphorus during the hydrolysis

are reflected in the significant downfield shift of the predominant octahedral peak from - 3.8 ppm (AIPTAS) to 0.9 ppm (AIPTA43200) and phosphoms removal data which

indicates increase in residual phosphorus from 33% (AIPTAS) to 61.32% (AIPTA432QO).

An increase in pH from 7.6 to 9.5 was also observed and mis is another factor that

favors the loss of phosphonis from the precipitate- Unlike other systems, aging did not

show any significant effect on the peak width.

Figure 58 a. 2 7 ~ ~ MAS NMR spectrum of solids obtained fmm the coprecipitation of

aluminum, phosphate and TA aged for 5 min (AIPTAS). [Al] = 430 mg L-',

[Pl = 500 mg L-' FA] = 1700 mg L-".

Figure 58 b. 2 7 ~ 1 MAS NMR spedrum of solids obtained fmm coprecipitation of

aluminum, phosphate and TA aged for 30 d (AIPTA43200).

[Al] = 430 mg L"; [Pl = 500 mg L*'; FA] = 1700 mg L".

Table 17, 2 7 ~ ~ MAS NMR chernical shifts and peak analysis results for AlPTA

system.

Sample Aging Peak Area Chernical Aluminum coordination

Period width of shift environment

( P P ~ ) Peak (PPm)

(%)

AIPTAS 5 min 21 -7 86.4 -3.6 Octahedral

9.2 9.08 47.8 Tetrahedral

AIPTA43200 3 d 21 -9 88.56 -0.9 Octahedral

20.8 3-42 25-4 Five-coordinate

26-7 8-02 42-1 Tetrahedral

Figures 60 a and b present the spedra for prehydrolyzed hydrous aluminum

oxide expos~d to a mixture of TA and phosphate for 5 min, (A15PTA5) and 30 d

(A15PTA43200). Table 18 summafizes the peak parameten. The spectmm of AISPTA5

(Figure 59 a, Table 18) shows three peaks and these peaks are shifted upfield

compared with chernical shiffs for peaks in the spectra of A15 and AITAS. This suggests

that some sorption sites on the solid were occupied by phosphate, a suggestion that is

supported by the observation that 20.3 % of the phosphorus had been rernoved h m

solution. Again wmparison of peak width data with that for the A15TA5 system implies

that TA did not interact as much as it did in the latter case. Furthemore, the similarity

between the A15PTA43200 spectnim and that of AfPTA43200 leads to a conclusion that

aging of the post precipitated solid resulted in its transformation to AIPTA Iike matefial.

Interestingly, the peak widths of the predorninant octahedral aluminum peak reporteci for

A15PTA5 were the lowest of any that have been observed in spectra of TA containing

material.

Figure 59 a. 2 7 ~ ~ MAS NMR spectrum of solids obtained from aging prehydrolyzed

aluminum in the presence of a mixture of phosphate and TA for 5 min

(A15PTA5). [Al] = 430 mg L-'; [Pl = 500 mg L-'; F A ] = 4700 mg L-'.

Figure 59 b. ''AI MAS NMR spectrum of solids obtained from aging prehydrolyzed

aluminum in the presence of a mixture of phosphate and TA for

43200 min (A15PTA43200). [Al] = 430 mg L-'; [Pl = 500 mg c'; F A ] = 1700 mg L*'.

Table 18, 2 7 ~ ~ W S NMR chernical shifts and peak analysis results for AISPTA system-

Sample Additional Peak Area Chemical Aluminum coordination

ag ing width of shift environment

Penod @Pm) peak ( P P ~

('w

AI~PTA 5 min 19-7 86.51 -1-3 Octahedra 1

14-4 6.02 25-7 Fivecoordinate

28.2 7.47 48.8 Tetrahedral

A15PTA43200 20.2 77.71 -0-7 Octahedral

37.8 ?4-41 19-9 Fivecoordinate

32.4 7.88 45-0 Tetrahedral

3.3- 1 Summary

By means of solid state MAS NMR, the coordination environments of

aluminum in solids precipitated under wastewater conditions have been identifieci.

Changes in the chernical shifts and peak widths of these solids due to aging of the

mixture in the presenœ of organic matter or f and phosphate have been monitored and

evaluated. The 2 7 ~ 1 MAS NMR results show that:

The predominant alumhum coordination environment in all the hydrous aluminum

oxide precipitates investigated was octahedral.

Generally the presence of phosphates Ri a solid produœs an upfield shift of the

peaks whereas Me presence of TA pmduces a downfield shift In the presence of

both phosphate and TA the peaks are initially shifted upfield and as the material is

allowed to age, the peaks are subsequently moved downfield due to the influence of

TA.

Aging is a vital pmcess that facilitates or inhibits the developrnent of order

depending on the composition of the matenal undergoing the aging proœss. The

presence of phosphate appears to enhance molecular order in the material whereas

the presence of TA inhibits molecular order.

The NMR spectra did not provide direct evidence for the presenœ of organic

coatings on A15TA5 and A15PTA5 materials. However, the results dernonstrated that

TA interacted less with the inorganic matenal in freshly precipitated solids than it did

after the precipitate had been aged for 30 d. The narrowness of the octahedral peak

width in the spectra of bath AI5TAS and AISPTAS solid compared with peak width of

the same peak in AITAS and AIPTAS solid suggest that TA had interacted less in the

former materials and this wuld be because it was adsorbed on the surface by .

physical forces.

3.4. Studies invofving municipal wastewater and recycled sludge samples

In order to compare the information obtained from the simulated wastewater

experiments with data based on experiments using influent wastewater, untreated

wastewater was obtained from Kingston West Sewage Treatrnent Plant ( M T P ) ,

Ontario, Canada. In this plant, alum is used to enhance biological phosphorus removal.

Two types of samples were obtained - untreated wastewater and recycled sludge

sampIes. The DOC concentrations of the wastewater and recycled sludge samples as

well as other parameters were measured (Table 19). The reduction in the DOC

concentration of the recycled sludge compared with raw water is typical of an activated

sludge process. The phosphorus concentration is lower than the typical value of 5 mg L-'

P used for the simulated wastewater experiments.

Table 19. Dissolved organic carbon and phosphorus concentrations of wastewater

samples obtained from an amvatecl sludge plant

Sample Parameters

Untreated Recycled sewage

wastewater sludge

PH 7.8 6.8 Turbidity, NTU Unfiltered 18.8

Filtered (0.45pm filter) 0.6

Phosphorus, mg L-' 3.0 +, 0-1

Dissolved Organic Content (DOC). mg 1-' 89.1 I 5.7 5-8 t 0.1

The FTlR spectrurn of the dried residue from the recycled sludge obtainsd fmm

the KWSTP after filtration is shown in Figure 60. The spednim is similar to those

reported for other sludge (Table 26) and the frequency assignments are those used for

Anderson. Frankfort and Marion sludge in Reference 152. However. their assignment for

the P-O stretching frequency is lower than values reported for phosphate-wntaining

materials by other workers 78.91.144 f 53.154

Wavenumber, cm -'

Figure 60. FTlR spectrurn of dtied recyded sludge obtained from KWSTP. Ontario.

Canada.

Table 20- lnfrared absorption spectra data for sludge samples in Reference 152 and in this study.

Frequency Assignment 'Sludge

Anderson FranMort Marion "Kingston "AISPTA

2960,2870 Aliphatic w w w 2958

methyl G H w

stretch

2920,2850 Aliphatic

methyfene G H

stretch

C=O stretch of sh

carboxyt or

carbonyl

groups

C=O stretch of s

secondary

amide (amide I

band)

N-H

deformation of

secondary

amide (amide

Il band)

1 000- 1040 P-O stretch in

phosphate-

containing

mate n'al

1040 C-O stretch of vs vs vs 1046

polysaccharide s

vs = very strong, s = strong, m = medium, w = weak vb = very broad.

" This study

The funcüonal groups involving carbon, oxygen and nitmgen indicate the

presence of organic wnstituents such as hurnic substances, proteins and

polysaccharide materials in the wastewater. The spedra of both the activated sludge

and TA possess oxygen-containing functional groups in their structures. Furthemore.

while bands characteristic of nitmgen-containing groups were evident in the sludge

samples, they were absent in the spectrum of AISPTA,

The alurninum-containing wmponent of the dried sludge sample from WSTP

plant was further characterized using '?AI MAS NMR and the spectnim compared wim

those obtained using simulated wastewater (Figure 61). Like the 2 7 ~ 1 MAS NMR spectra

of solids obtained from the simulated wastewater studies, this spectrum indicates that

the material is arnorphous and contains aluminurn predominantiy in an octahedral

coordination environment The octahedral peak is shifted upfield compared with the

analogous peak in the simulated experimental Al systern but downfield compared with

both A15P and AISPTA solids. These observations suggest the presence of less

phosphate in the KTWPC plant sample mmpared with sarnples obtained from AISP and

AISPTA systems.

Coprecipitation and postprecipitation experiments were also camed out on the

untreated wastewater. Samples of the water were filtered using a 0.45 pm membrane

filter and aliquots of the filtrate diluted DDW to give DOC cancentrations of 8.9 and

22.3 mg L-' , which were used for the experiments. The distilled de-ionized water (DDW)

used for the dilution was adjusted to the pH of the wastewater (7.8). The AIUS system

involved coprecipitation of components in the filtered and diluted wastewater by alum

whereas in the AISUS system, alum-generated solids aged for 5 min were allowed to

interact with the components in the same modified wastewater. The products were

characterized by the ferron test Tables 20 and 21 summarize the results.

Overall, the values for the solids forrned by either CO- or postprecipitation were

generally Iower than values reported for similar experiments using simulated

wastewater- The for coprecipitation experiments using samples containing a DOC of

approximately 9 mg L-' were 1124 s and 2453 s for AlUS and AIPTA respectively. For

the postprecipitation experiments, they were 1352s and 17100 s for AISUS and AISPTA

respectively. The differenœs could be due to the low phosphotus content of the

untreated wastewater (Table 18) which was diluted (to less than 1 mg L-') in order to

obtain the DOC concentrations used for the expenments- The low values coufd also be

due to the nature of wastewater DOC, a cornplex mixture of organics that Vary in size

and fundional groups (Section 1.6.5 page 31 ; Table 20 page 176).

The values for coprecipitated AlUS sarnples (992 - 1240 s) were lower than

those reported for Al systems (1715 - 3527 s), and this is consistent with the results

obtained for the AlTA systern (247 - 510 s). The values did not change significantly

with aging of the sarnple, which is similar to the trend observed during the aging of

AIPTA samples- Like the AiPTA systern (Table 9) where increases in TA concentration

were associated with increased reactivity of the solid, increasing the DOC concentration

caused an increase in the reactivity of the solids.

The values for postprecipitated A15US solids were higher than those obtained for

coprecipitated AlUS solids at the h o concentrations studied. This trend is similar to

those noted for systems containing tannic acid (AIPTA and AISPTA). The DOM removal

efficiency for both systems (AISUS and AIUS) was not remarkably different but removal

was better than in the analogous systerns involving the use of TA.

Figure 61. 2 7 ~ 1 MAS NMR spectmm of recycled sludge obtained frorn KWSTP,

Ontario, Canada.

Table 21 - Percent fast reacting aluminum species (FR) and tso values obtained from

ferron test on wprecipitation system (AIUS) using municipal wastewater

Mer d'fierent. aging periods,

DOC, Aging time (min)

concentration 5 15 30 60

mg L-'

% FR tso (s) % FR to (s) %FR b (SI % FR to (s)

Table 22. Percent fast readng aluminum species (FR) and tso values obtaïned from

ferron test on postprecipitation system (A15US) using municipal

wastewater after different aging periods.

DOC. Aging time (min)

concentration 5 15 30 60

mg L"

%FR b(s) %FR b(s) %FR b(s) %FR b,(s)

3.5. General conclusions

In this work, an attempt has been made to antribute to a batter understanding of

the interaction between aluminum, dissolved organic matter and phosphorus during

wastewater treatrnent- The major findings are:

Tannic acid, as a surrogate for natural organic matter, inhibited phosphorus removal

and the extent of inhibition increased with increasing concentration of the tannic acid.

Coprecipitation of aluminum, phosphate and tannic acid, which is synonymous with

the addition of alum before or in the aerator of a wastewater treatment plant,

produced a hydfoxy-aluminum-phosphate-tannate complex, in the form of a solid

with very small particle size. Separation of the suspended particies in this system,

either by settling or by filtration. is difficult and is likely to result in inefficient solid

removal during wastewater treatment This system gives high residual aluminum,

with much of the aluminum present in the colloidal precipitate.

The system where phosphate was coprecipitated with alum. and subsequently

exposed to tannic acid gave the best compromise results for both phosphorus and

organic matter removal. This finding suggests that it is advantageous to add at least

a portion of the aluminum coagulant at the exit of the aerator. At this point, residual

phosphorus is removed by coprecipitation under conditions where Me concentration

of organic matter is relatively low. The sludge produœd (consisting in large part of

aluminum hydroxyphosphate and hydrous aluminum oxide) when recycled into the

aerator, enhances removal of soluble organic matter by postprecipitation ont0 the

recycled SI udge.

For the above system, atomic and interfacial force micmscopy were used to verify a

rnodel based on the formation of an organic coating on the surface of the inorganic

solid generated when prehydrolyzed aluminum is used to treat wastewater. The

atomic force microscopy results indicated that the viscoelasüc properties of particles

where phosphorus and tannic acid had been removed by postprecipitation were

different from those of particles obtained by coprecipitation. The interfaclal force

microscopy data for the former solid showed clear evidence that the particles

consisted of a hard inorganic core with a sofi surface coating of an organic layer-

In many ways, TA showed properties similar to the DOM in the wastewater obtained

from KWSTP- The solids from the simulated wastewater experiments using TA and

the recycled sludge solids were both amorphous and the predominant aluminurn

coordination environment was octahedral. In both cases, the to values for

postprecipitated solids were higher than values for coprecipitated solids. Aging the

samples or increasing the concentration of TA showed similar effects in

coprecipitated systems using simulated wastewater and wastewater from KWSTP.

There were differences, however. in that higher organic fernoval and lower tao values

were observed in experiments conducted with wastewater fmm KWSTP cornpared to

the simulated wastewater experiments using TA. These differences should be the

subject of further investigations.

1. Blitz, E., Bohnke, B., Doetsch, P.. Dreschmann, P., Poppinghaus, Km, Siekmann, K and

Thomas, S. Waste water technology: origin, collection, treatment, and analysis of

waste water, edited by W. Fresenius, W. Schneider. K. Poppinghaus, and B.

Bohnke; mrnrnissioned by the Deutsaie Gesellschaft Fiir Technische

Zusammenarbeit (GTZ) GmbH, Springer-Verlag Berlin Heidelberg, Germany, 1989-

2. Tchobanoglous. G- Wastewater treatmenk in Water rasoumes handbook, Mays, LW-

(ed.), McGraw-Hill, USA, 1996.

3. Lamb, C.J. Water Quality and lts Control. John Wiley and Sons Inc.. New York. 1985.

4. Viessman, W. Jr. and Harnmer, M.J. Water Supply and Pollution Confrol, Harper Collins

Publishers lnc., New York, 1985,

5. Lind, B.C- (1998). Phosphorus inactivation in wastewater treatment Biological and

Chernical Strategies. Water/Engineering and Management, February, 18 - 21.

6. Boehnke, 6, Diering, B. and Zuckut, S. (1 997). Cost-effective wastewater treatment

process for removal of organic and nutnents. Part 1. Water / Engii?eennng and

Management, 30 - 35.

7. Heinke, G. W., Tay, A- J. and Qazi, M. A, (1980). Effect of chernical addition on the

performance of settling tanks. J. WPCF, 52(12): 2946 - 2954.

8. Viessman, W. Jr. and Welty, C. Water Management Technoîogy and Institutions. Harper

and Row Publishers, Inc. 7985-

9. Baes, C. F. and Mesmer, R. E. The hydmlysis of cations. John Wiley and Sons, New

York, 1976.

10. Brosset, C., Biedennam, G. and Sillen, L. G. (1 954). Studies on the hydrolysis of metal

ions. XI. The aluminum ion, Al ". Acta Chem., Scand., 8:1917-1926.

1 7 . Hsu, .P. H. and Bates, T. F. (1964). Formation of X-ray amorphous and crystalline

aluminum hydroxide. Minemi. Mag., 33: 749-768.

12- Akitt, .J. W. (1989)- Multinuclear studies of aluminum cornpounds. Pmg. NMR

S p e ~ f r ~ ~ c ~ p y , 21 : 1 -149.

13- Bertsch, P. M. (1 989)- Aqueous polynuciear afuminum species in The enHmnmenfal

chemisty of aiuminum. Sposito, G, (ed-)- CRS Press, Inc- Boca Raton, Florida-

1989.

14. Schofield, R. K. and Taylor, A. W. (1 954). The hydrolysis of aluminum sait solutions- J.

Chem- Soc. 4445 - 4448.

15, Kubota, H, (1956)- The hydrolysis of aluminurn in dilute solutions, Diss- Abstr- 16, 864.

16. Bronsted, J. N- and Volquartz, K., (1928). Acid dissociation of aquo ions, 2.. Phys.

Chem. 134: 97-1 34.

17. Ito, T. and Yui, N. (1954). Hydrolysis constant of the aluminum ion in chloride solutions.

Chem. Abstract, 4445.

18. Fflnk, C- R. and Peech, M., (1963). Hydrolysis of the aluminum ion in dilute aqueous

solutions, Inorg. Chem., 2: 473 - 478.

19. Raupach, M-, (1963)- Solubility of simple aluminum cornpounds expected in soils- 1.

Hydroxides and oxyhydroxides, Aust J- Soi1 Res., 1 : 28-35.

20. May, H. M., Helmke, P. A., and Jackson, M. L., (1979). Gibbsite solubility and

thermodynamic properties of hydroxyaluminurn ions in aqueous solution at 25 " C.

Geochem- Cosmochim. Acta. 43: 861 - 868,

21. Hem, J. D. and Roberson. C. E., (1967). Fom and stability of aluminum hydroxide

complexes in dilute solutions, U. S. Geol- Surv. Water-Supply Pap., 1827-A-

22. Bertsch, P. M., Barnhisel, R. I . , Thomas, G. W., Layton, W. J., and Smith, S. L., (1986).

Quantitative detemination of aluminum-27 by high resolution nuclear magnetic

resonance spedrornetry, Anal, Chem. 58: 258302585.

23. Akitt, J. W. and Farthing, A., (1978). New 27 Al NMR studies of the hydrolysis of the

alurninum (Ill) cation, J. Magn Reson, 32: 345-352.

24. Bottero, J. Y., Cases, J. M., Flessinger, F., and 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.

25. Bertsch, P. M. (1987) Conditions for AllB polymer formation in partially neutralized

aluminum solutions, Soi1 ScL Soc. Am, J. 51 : 825-828.

26. O'Melia, C R . Coagulation and flocailation. physicochemical processes for water quality

control. W.J. Weber Jr., (ed-). John Wiley and Sons, New York, 1972.

27. Letterman, R- D-, Quon, J- E- and Gemmell, R S. (1973). Influence of rapid-mix

parameters on Flocculation. Jour: A W A , 65: 716 - 722.

28. AWWA Coagulation Cornmittee. (1 989). Coagulation as an integral water treatment

process. Jour. A W A , 81 (1 0): 72 - 78.

29. Jenkins, D., Ferguson, J-F-, and Menar, A.B. (1971). Chernical proœsses for

phosphonrs removal. Waf. Res, 5: 369-389.

30. Engelbrecht, R. S. and Morgan, J.J. (1959)- Studies on the occurrence and degradation

of condensed phosphates in surface waters, Sew. and Ind. Wastes, 31 (4): 458 - 478.

31. Vinconneau, J, C-, Schaack, F., Chevalier, D., Villescot, D., Jaubert, M,, Laval, Ch,, and

Lambert, S., (1985). The problem of phosphorus in France-lts presence in naturai

waters and biological phosphorus removal, Wat- Sei. Tech., 17: 1-9.

32. Sawyer, N. C. (1965). Problern of phosphorus in water supplies, Jour. AWWA

,S7(ll):l431-l439-

33. Hudson, E. J. and Marson, H. W. (1970). Chemisfry and Industy, 1449-1458.

34. Weiss, C .M., (1969). Relation of phosphates to eutrophication, Jour. AWWA. August ,

387 -391

35. Nesbitt, J. B. (1969). Phosphorus removal-the state of the art. J- WPCF, 41(5): 701-712.

36. De Vries, H. P. and Rensink J. H. (1985). Biological phosphorus removal at low sludge

loadings by partial stripping. Paper presented to Int Conference for management

strategies for phosphorus in the environment, Lisbon, Portugal, July 1985.

37. Fulkase, T., Shibata, M. and Miyayi, Y. (1985). Fadors affectrng biological removal of p hosphorus, Wat- Scl fechnol, 17: i 87-198.

38. Cooper, P., Dee, T. and Yang, G. (1993). Nutrient removal-methods of meeting the EC

urban wastewater directive. Paper presented at the 4m Annual Conference on

Industria1 Wastewater Treatment. Esher, Surrey, 1 O March 1993-

39- Nue, S.G. (2991). A review of approacfies to achieve low effluent phosphorus

concentrations. Wafer Polhf- Res. J- Can. 26: 495.547-

40- Ockershausen, R. W. (1974). A t m Vs Phosphates: Ifs no contest Wafer and Waste

Engineenilg. 1 1 :Ml.

41. Bowker, R. P. G., and Stensel, H. D. (1990). Phosphorus removal fmm wastewater.

Noyes Data Corporation, USA-

42. Boisvert, J., Tor T., Benak, A. and Jolicoeur. C. (1997). Phosphate Adsorption in

Flocculation processes of Aluminum sulfate and Poly-aluminum-silicatesulfate. Wat.

Res. 31 (8): 1 939-1 946.

43. Hayden, P. L. and Rubin, A.J. In Aqueous Environmental Chemisfryof Metals. Rubin,

A. J. Ed. Ann Arbor Science, Michigan. 1976.

44. Lijklema, L. (1980). Interaction of orthophosphate with iron (III) and aluminum

hydroxides, Envimn. Sci. Tech- 14(5): 537- 541,

45. Sims, J. T. and Ellis, B. G. (1983). Changes in phosphorus adsorption associated with

aging of aluminum hydroxide suspensions. Soil ScL Soc. Am. J., 47:912 - 916.

46. Bowden. J.W. and Nagarajah, S., Barrow, N.J., Posner, A.M. and Quirk, J.P. (1980).

Descfibing the adsorption of phosphate, citrate and selenite on a variable-charge

mineral surface. Aust. J. Soil Res- 18: 49-60.

47. Hsu, P.H. (1989)- Aluminum hydroxides and Oxyhydroxides. p.331-378. In J.B- Dkon

and S.B Weed Eds. Minerais in Soii Envrionments- 2* Ed- SSSA Book Ser- No. i -

SSSa, Madison, W1-

48. Rajan, S.S.S., Perrott, K-W-, and Saunders, W-H-M, (1 974)- Identification of phosphaie-

reactive sites of hydms alumina from proton consumption during phosphate

adsorption at constant pH values. J, Soil Sci- 38: 21 1-217-

49. Chen. B. H. H. King, P. H- and Randall, C. W. (1974)- A pmposed stnrcture of an

alurninum-phosphate species of importance in wastewater treatment Environ- Letts,

6(2):lZg-I 38-

50. Leckie, J, and Stumm, W- Phosphate Precipitation. In E, F- Gloyna and WW. jr.

Eckenfelder Eds. Water quality lmprovement by physical. and chernical proœsses-

Water Res. Symp. No 3. University of Texas press, Austin, Texas, 1970, pp237-249.

51- Duff'y, S.J. and vanloon, G.W. (1995)- Investigations of aluminum hydroxyphosphates

and adivated sludge by 2 7 ~ ~ and P MAS NMR. Can. J- Chem. 73:1645-1659.

52. Hsu, P. H. (1975). Precipitation of phosphate fmm solution using aluminurn salt. Wat Res-, 9: 1155-1 161.

53. Galameau, E. and Gehr, R (1997)- Phosphonrs Rernoval from Wastewaters:

Experimental and Theoretical support for alternative rnechanism , Wat- Res. 3(2):

328-338.

54. Recht, H.L,. Ghassemi, M. Kinetic and rnechanisms of precipitation and nature of the

precipitate obtained in phosphorus rernoval from wastewater using aluminum and

iron salts, Rept 1701 01 €KI, US dept of the interior- 1 979.

55. Painter, H. A., Viney, M., and Bywaten, A., (1960). Composition of sewage and sewage

effluents. Metropolitan and Southem Branch Meeting, The Instïtute. Of Sewage

Purification. London- UK-

56. Bunch, R- L., Barth, E-F., and Ettinger, M.B. (1961). Organic material in secondary

effluents. J. Water Poliut Control Fed. 33(2): 122-126.

57. Manka, J- and Rehbun, M. (1971)- Classification of organics in sewndary effluents.

Environ- Sci. Technoi. 7(5): 606-609.

58. AWWA Research Cornmittee Report (1979)- Organic removal by coagulation: A review

and research needs. Jour, A W A - 7 1 : 588 - 603,

59. Hundt, T. R- and O'Melia, C- R, (1988)- Aluminum-fulvic acid interactions: mechanisms

and applications, Jour, A W A , 80(4):176 -186

60- Dempsey, B. A-, Ganho, R.M. and 0' Melia, C.R. (1984)- The coagulation of humic

substances by means of aluminum salts- Jour, A WWA- 76(4)r 141 4 50-

61. Van Benschoten, J. E. and Edzwald, J- K. (1990). Chernical aspect of coagulation using

aluminum salts- II. Coagulation of fulvic acid using alum and polyaluminum chloride.

Wat- Res. 24(12): 1 527 - 1 535.

62. Moore, G. N- (1977). Interaction of fulvic acid and Kaolinite during coagulation, P m

A WWA, 97 Th Ann. Conf, Anahalm, California.

63. Collins, M. R., Amy, G. L and Steelink, C- (1986). Molecular weight distribution,

carboxylic acidity, and humic substances content of aquatic organic mater:

lm plications for removal during water treatrnents. Envimn. Sci Technol., 20:1028-

1032.

64. Semens, M.J. and Staples, A. B. (1986). The nature of organics removed during

treatment of Mississippi River water. JOUE AWWA, 78(2): 76-81 -

65. Julien, F., Gueroux, B. and Mazet, M. (1994). Cornparison of organic compounds

removal by coagulation-flocculation and by adsorption ont0 preformed hydroxide

flocs. Water Res. 28: 2567-2574.

66. Cathalifaud, G., Ayele, J. and Mazet, M. (1997). Aluminum ions / organic molecules

complexation: Formation constants and stoidiiometry. Application to dtinking water

production. Water Res. 31 : 689-698-

67. Lawrence, J. (1980). Semi-quantitative deteminaion of fulvic acid, tannic acid and lignin

in natural waters. Water Res. 14: 373-377.

68. Alberts, J.J. (1982). The effect of metal ions on the ultraviolet spectra of humic acid,

tannic acid and lignosulfonic acid. Water Res. 16: 1273-1 276-

69. Violante, A-, and Huang, P. M. (1984). Nature and properties of pseudoboehrnites

formed in the presence of organic and inorganic ligands- Soi7 Sci- SOC- Am. J. 48:

1 193-1 201.

70- Goh, TB., Violante, A. and Huang, F.M. (1986). Influence of tannic acid on retention of

copper and zinc by aluminum precipitation products- Soi. ScL Soc. Am. J. 50: 826

825.

71. Goh, T.B. and Huang, P.M. (19û6). Influence of ci* and tannic acids on hydroxy-AC

interlayering in montmorillonife- Clays and Clay Minerais, 34: 37-44.

72. Inskeep, W.P. and Silvertooth, J-C. (1988). Inhibition of hydroxyapahite precipitation in

the presence of fulvic, hurnic and tannic acids. Soi/ Sci.. Soc Am- J. 52: 941-946-

73. Buondonno, A., Felleca, O. and Violante, A. (i989). Properties of organic mineral

complexes formed by different addition sequences of hydroxy-Al, montmorillonite,

and tannic acid. CIays and Clay Mliierals- 37: 235-242.

74. Buckingham, J.. Dictionary of organic compounds. Vol 6, 6th ed- Chaprnan and Hall, New

York. 1996.

75. lnoue, K-, and Huang, P. M. (1986). Influence of seleded organic ligands on the

formation of allophane and irnogolite. Soi1 Sci. Soc. Am. J. 50: 16234637.

76. Guy, D. R. and Chakrabarti, L. C. (1976). Studies of metal-organic interactions in mode1

systems pertaining to natural waters. Can, J. Chem. 54: 2600 - 261 1.

77. Humic and Fulvic acids: Isolation, structure, and environmental mle, Gaffney, J. S.,

Marley, N. A. and Clark, S. B. (eds.). American Chernical Society, Washington DC,

1996.

78. Kastelan-Macan, M. and Petrovic, M. (1996). The mie of fulvic acids in phosphorus

sorption and release from mineral particles. Wat. Res. 34(7-8): 259-265.

79. Swenson, R.M., Cole, C- V. and Sieling, (1949)- Fixation of phosphate by imn and

aluminum and replacement by organic and inorganic ions, Soil ScL 673-22-

80. Sibanda, H. M. and Young, S.D. (1986). Cornpetitive adsorption of humic acids and

phosphate on geothite, gibbsite. and €wo tropical soils. J. Soil Sci. 37:197-204.

81. Kwong, Ng Kee k F. and P. M. Huang. (1978). Sorption of phosphate by hydrolytic

products of alurninum. Natum (London). 271: 336 - 338- 82. Uwong, Ng Kee k F.. and P. M. Huang- (1 979). Surface reacüvity of aluminum

hydroxides precipitated in the presence of low molecular weight organic acids. Soi/

ScÏ- Soc. Am. J. 43: 1107 - 11 13,

83. Borggard, O. K Jorgensen, S. S. Moberg. J. P. and Raben-Lange, B. (1990). Influence

of organic matter on phosphate adsorption by aluminum and iron oxides in sandy

sails. J. Soil Sei, 41 : 443 - 449.

84. Gerke, J. (1993). Phosphate adsorption by humic / Fe-oxide mixtures aged at pH 4 and

7 and by poorly ordered Fe-oxide. Geodrema. 59: 279-288.

85. Duffy, S.J., Deutschman, J.E., and vanloon, G-W. (1994). Evaluation of phospho~s

removal in the activated sludge process. Water Pollut. Res- J. Can. 24(4): 487406.

86. Aayukaev, R. I . Methods of increasing granular filters for water purification. In Chernical

water and wastewater treatment Hahn, H. H. and Klute, R. (ed.)., Proceedings of the

4'h Gothenburg symposium 1990, Odober 1-3, 1990, Madrid, Spain. Springer-verfag

Berlin Heidelberg, Germany.

87. Smith, R.W. (1971). Reactions among equilibrium and non-equilibrium aqueous species

of aluminum hydroxy complexes- Adv. Chem. Ser. lû6:250 - 279.

88. Duffy, S.J. and vanloon, G-W. (1994). Characteriration of amorphous aluminurn

hydroxide by the ferron rnethod- Envrion- ScL TechnoL 28(ll): 1950 -1956.

89. Tarte, P. (1 967). lnfrared spectra of inorganic aluminates and characteristic vibrational

frequencies of A104 tetrahedral and A104 octahedral. Specfmchim. Acta. 23A: 2127-

2143.

90. Morterra, C., Emanuel, C., Cerrato, G.. and Magnacca, G- (1992)- Infrared study of

some surface properu'es of Boehmite (AIO0H)- J- Chem- Soc. Faraday Tians . 88(3):

339-348.

91. Corbridge. D.E.C., and Lowe, E. J. (1954)- The infrared spectfa of some inorganic

phosphorus conpounds- J. Chem- Soc. (London)_ 493-502.

92. Schnitzer, M. (W69). Reactions between fulvic acid, a soil humic cornpound and

inorganic soil constitutes, Soi1 ScL Soc- Am. Pmc. 33:75-81.

93- Tan, K- H,, King. L, D. and Mom-s, L. D- (1971)- Complex reactions of zinc with organic

matter extracteci from sewage sludge, Soi1 Sci- Soc. Am. PTOC. 35:748 -752,

94. Bloom, P. R. (1981) Metal-organic matter interactions in soil, In Chemisfs. in the soil

environment. Spec. Pub. 40. American Society of Agronomy, Madison, WI.

95. Yost, E., Tejedor-Tejedor, M. L, and Anderson, M.(1990). In-situ CIR-fTIR

charactenzation of salicylate complexes of the geothite / aqueous solution interface,

Envr'ron. Sei, Technol. 24:822-828-

96. Biber, M. V. and Stumm, W. (1994). An in-situ ATR-FTIR study: The surface

coordination of salicylic acid on aluminum and iron (1 II) oxide, Environ. Sci. Technol.

281763-768.

97. Muller, D., Berger, G., Grunze, 1.. Ladwid, G., Hallas, E. and Haubenreisser. U. (1983).

Solid-state high-resolution 27 Al nuclear magnetic resonance studies of the structure

Cao-AI2O3-P205 glasses, Phys. Chem. Glasses- 24(2): 37-42.

98. Kirkpatrick, R. J., Oestrïke, R., Weiss, R., Smith, C- A., and Oldfield, E. (1986). High

resolution 27 Al and 29 Si NMR spectroscopy of glasses and crystals along the joïn

CaMgSi206.CaAlzOs. American Minemlagist. 71: 705-71 1.

99. Klinowski, J. (1984). Nuclear magnetic resonance studies of zeolites. Pmgmss in NMR

Spectmscopy, 16: 237-309.

IOO. Venerna, R.F., Peters, J. A. and Herman, V. B. (1990). Multinuclear Magnetic

Resonance study of the coi>rdination of aluminum (III) with glycolic acid in aqueous

solution, wmpared with oxalic and malonic acid- J- Chem. Soc. Dalton Trans-

21 37-21 43-

Turner, G. L., Kirkpatn'ck, R. J., Risbud, S-H- and Oldfield, E_ (1987). Multinuclear magic-

angle sample spinning nuclear resonance spectroscopie studies of crystalline and

amorphous ceramic materials. Amencan Ceramii Sociefy Bulletin, 66: 656-663,

Gessner, M. D., Behrens, HJ. and Scheler, G. (1981)- Determination of the aluminum

coordination in aluminum-oxygen compounds by solid-state high resolution 27 Al

NMR, Chem. Phy. Lett. 79(i): 59-61 .

Fitzgerald, J-J-, Dec, S. F. and Hamza, A. 1- (1989)- Observation of fivecoordinated Al in

pyrophyllite dehydroxylate by solid-state 27 AI NMR spectroscopy at 14 T. Amenin

minemlogist. 74: 1405-1 408.

Engelhardt, G and Michel, D. High-riesolution solid-state NMR of silicates and zeolites,

John Wiley and Sons Ltd., 1987.

Alemany, L. B. and Kirker, G- W. (1986). Fint observation of Scoordinate aluminum by

MAS '?AI NMR in well-characterized solids, J. Am chem Soc 108: 61 58 - 6162.

Alemany, L. B., Timken, H. K.C. and Johnson, 1. D. (1988). Aluminum-27 NMR study of

AiP04-21 and andalusite- Advantages of high field and very fast MAS. J- Mgn Reson

80: 427 - 438,

Magonov, S. N and Whangbo, M. H. Surface Analysis with STM and AFM, VCH

Publishers, Inc. Weinheim, 1996.

Magonov, S and Heaton, M. G* (1998). Atomic force microscopy, part 6: Recent

developments in Afm of polyrners. Amer. Lab- May, 9-16.

McGhie, A. J., Tang, S.L. and Li, S.F .Y. (1995). Expanding the uses of AFM, Chemtech,

20-26 -

Radmacher, M., Tillrnan, R.W. and Gaub, H. E. (1993). Imaging viscoelasticity by force

modulation wifh the atomic force microscope. 6iophysys J. 64: 735 - 742.

Toa, N. J., Lindsay, N. M. and Lees, S. (1 992). Measuring the microelastic pmperties of

biological matenals. Biophys. J- 63: 1 165 - 1 169.

Weisenhom, A- L, Maivald, P-, Butt, H. J. and Hansma. P. K. (1992). Measuring

adhesion, attraction, and repulsion between surfaces in liqoids with an atomic force

microscope. Pbys Rev. 8 (Condensed Matfer;). 45: 1 1226 - 11232.

Zhong, Q., Innis, D.. Kjoller. K and Elings, V. B. (1993). Fradured polyrner l silica fibre

surface studies by tapping-mode atomic force rnicroscopy. Suhœ Sci. Lett. 290:

L688-L692.

Hoper R-, Gesang T., Possart W., Heinemann 0.0. and Boseck S., (1995). Imagnig

elastic properties with an atomic force microscope operatirtg in the tapping mode.

Ulframicmscopy 6û: 1 7-24.

Whangbo, M. H., Bar. G. and Brandsch, R. (1997). Description of phase imaging in tapping mode atomic force microscopy by hannonic approximation. Sufi Sci. 41 1 :

L749-L801.

Magonov SN.. Elings V. and Whangbo M.H. (1997). Phase imaging and stiiess in

tapping-mode atomic force microscopy. Surf. Sci 375: L3û5-L391.

Bar, G., Thomann. Y.. Brandsch, Cantow, H. J. and Whangbo, M. H. (1997). Factors

affecting the height and phase images in tapping mode atomic force rnicroscopy.

Study of phase-separated polymer blends of poly (ethene-CO-styrene) and poly (2.6-

dimethyl-, 4-phenylene oxide), Langmuic 13: 3807.3812.

Warren, O.L.. Graham, J. F. and Norton, P. R. (1998). Interfacial force microscopy: A

novel scanning probe technique for imaging and quantitative measurement of

interfacial forces and nanomechanical properües. Physcs in Canada, 122-136.

Houston, J. E. and Michalske, T. A. (1992). The interfacial-force microscope. Natoie, . 356,266 - 267-

Joyce, S. A., Thomas, R.C., Houston, J. E., Michalske. T. A. and Crooks, R.M. (1992).

Mechanical relaxation of organic rnonolayer films measured by force rnicroscopy.

Phys. Rev. Lett 68(18): 2790 - 2793.

Joyce, S. A., Houston, J. E- and Mkhalske, T. A- (1 992)- Differentiaüon of topographical

and chernical structures using an interfacial force microscope. Appl. Phys. Le& 60(10): 1175 - 1177,

Warren, 0-L, Graham. J. F. and Norton, P. R, (7997) Tapping mode imaging with an

interfacial force microscope, Rev- Sci- lnsfrurn. 68<11): 41 24-

Microprobe Technique in Earth Sciences. Pott, P. J-, Bawles, J. F- W., Reed, S- J- B. and

Cave, M. R. (ed.). The rnineralogical Society Series, Chapman and Hall, UK, 1995.

Hoimes, G.. Singh, B. R. and Theodore. L. Handbmk of Envfmnmnfal Management

and TechnoIogyY John Wiley and Sons Inc- New York- 1993.

Edmald, J. K, Becker, W. C. Wattier, K L. (1985) Surrogate parameters for monitoring

organic matter and THM precursors, Jour. A WWA- ï7:i 22-1 32.

Amy, G. L., Chadik, P. A. and Chowdhury, Z (1987). Developing models for predicting

THM formation potential and kinetics, Jour, A WWA. 79: 89-97.

Marrow, C. M. and Minear, R. A. (1987). Use of regression models to link raw water

characteristics to THM concentrations in drinking water, Wat. Res. 21: 4148.

Edwards, M., Benjamin, M. M. and Ryan, N. J. (1996). Role of organic acidity in sorption

of natural organic matter (NOM) to oxide surfaces- Coiioidal Surfaces A.

Physiochern. Eng. Aspects. 107: 297 - 307. Alberts, J.J. (1982)- The effect of metal ions on the ultraviolet spectra of humic acid,

tannic acid and Iignosulfonic acide Wafer Res. 16: 1273-7 276.

Dentel, S. K. Bottero, J. Y.. Khatib, K., Demougeot, H., Duguet, J. P. and Anselme, C.

(1 995). Sorption of tannic acid, phenol, and 2,4,5trichlorophenol on organodays.

Water Res. 29: 1273-1 280-

Han. W. Lindsay, S. M. and Jing, T. (1996). A magnetically driven oscillating probe

microscope for operation in liquids. Appi- Phys. Lett. 69: 41 1 1-41 13.

Nail, S. L., White, J. L, and Hem, S. L (1975). Cornparison of IR spectroscopie analysis

and x-ray diffraction of aluminum hydroxide gel- J. Pham ScL 64 (7): 1 166-1 1 69.

133. Sema, C- J-, White, J. L-, and Hem, S. L- (1977)- Anion-aluminum hydroxide gel

interactions- Soi1 Sei. Soc- Am. J- 41: 1009-1013.

Russell, J. D., Paterson, E. Fraser, A. R. and Famer, V. C- (1975) Adsorption of carbon

dioxide on geothite (FeOOH) surfaces, and its implications for anion adsorption. J-

Chem. Soc. Faraday Tram- 171 :l62Sl63O,

The Sadtler infmmd specfra handbook of minerais and clays. Sadtler Research

laboratones, USA, 1 982-

Rye, C., Shimizu, m. Collins, B- And Glimchcher, MI J- (2990) Resolution-enhanced

fourier transfomi infrared spectroscopy study of the environment of phosphate ions

in the early deposits of a soiid phase of calcium-phosphate in bone and enamel, and

their evolution with age. Calcif TSssue Int -46: 384 - 394-

Ng Kee Kwong, KF-, and Huang, P.M. (1981) Cornparison of the influence of tannic acid

and selected low-rnolecu lar weig ht organic acids on precipitation products of

alurninum- Geodema- 26: 179-193,

Pouchert, C. J. (1985) The Aldrich Libray of FT-IR Specfra, Edition 1, Volume 1, Aldrich

Chernical company lnc. Spectrum No. 544C.

Gregor, LE, C. J. Nokes and E. Fenton (1997). Optimizing natural organic matter

removal from low turbidity waters by controlled pH adjustrnent of aluminum

coagulation. Water Res. 31 (2): 2949-2958.

Levesque, M. and Schnitzer, M. (1967). Organo-rnetallic interactions in soils: 6.

Preparation and properties of fulvic acid-metal phosphates. Soii Sci 1 O3(3): 183-

190.

Greenland, D.O. (1965). Interaction between clays and organic compound in soils, 1.

Mechanisms of interaction between clays and define organic compwnds. Soils Fert 28: 41 5-425.

Narnjesnik-Dejanovic, Ksenija and Maurice, Paûïcia. (1997). Atomic force microscopy of

soli and Stream fulvic acids. Colloids and Su-ces, A: PhysicochemiW and

Engineering Aspects 120: 77 - 86.

Stevenson, 1, L and Schnitzer, M- (1982)- Transmission electrcrn microscope of extracted

fulvic and humic acids- Soi! Sci- 133: 1 79-1 85-

Magonov, S. N, And Reneker, D- (1997). Charaderkation of polymer surfaces with

atomic force rnkroscopy, Ann Revs Mat ScL. 27: 175-180-

Joyce, S. A. and Houston, J. € (1991). A new force sensor incorporating force-feedback

control for interfacial force microscop y- Rev. ScL lnstrum. 62, 71 0-

Graham, J- F-, Kovar, M., Pappalando, P., vanLmn, J. and Warren, O. L. (1998).

Quantitative nanoscale mechanical properties of a phase segregated homopolymer

surface. J. Mafer. Res. f 3(12): 3565 - 3570,

Timoshenko, S- D. and Goodier, J. N. Theoryof elasfi'city, McGraw-Hill, New York, 1970.

Johnson, K- L. Contact mechanics. Cambridge University, New York, 1994.

Thomas, R.C., Houston, J- E., Michalske, T- A. and Crooks, R-M, (1993)- The mechanical

response of gold substrates passivated by self-assembling monolayer films.

Science. 259: 1 883 - 1 885.

Cotton, F. A. and Wilkinson, G. Advanœd inofganic chemistfy- John Wiley and sons,

New York, 1980-

Blackwell, C S and Patton, R- L (1984) Alurninum-27 and Phosphorus-31 Nuclear

Magnetic Resonance Studies of Aluminuophosphate Molecular Sieves. J. Phys.

Chern, 88: 61 35-61 39-

Boyd, S. A., Sommers, L E. and Nelson, D. W. (1979) lnfrared spectra of sewage sludge

fractions: Evidence for an amide metal binding site. Soi1 Sci. Soc. Am. J. 43: 893 - 899.

Williams, D. H. and Fleming, Specfmscopk Methods in Olganic Chemistry, Mc-Graw Hill,

London, 1973-

Nyquist, R. A., Putzig, C. L- and Leugers, M. A. The handbook of inframd and Raman

spectm of lnorganic Compounds and Organic Salts (in 4 volumes), lnfrared and

Raman spectral atlas of inorganic wrnpounds and organic salts: Text and

explanations- Academic Press, New York- 1997.

5.0. Appendix A

Microprobe analysis images for AIPTA and AISPTA solid sarnple. Discussions on the data are presented in section 3-1 page 94.

Figure 62 a Backscattered electron image of AlPTA system aged for 5 min.

Figure 62 b. Elemental distribution of aluminum and phosphorus in AlPTA system aged

for 5 min,

Figure 63 a Backscattered electron image of A15PTA system aged for 5 min.

Figure 63 b. Elemental distribution of aluminum and phosphate in AISPTA system aged for

5 min.