atomic level understanding of orthophosphate adsorption by

University of Southern Denmark

Atomic Level Understanding of Orthophosphate Adsorption by Magnesium Aluminum-Layered Double Hydroxides - A Multitechnique Study

Lundehøj, Laura; Cellier, Joel; Forano, Claude; Nielsen, Ulla Gro

Published in:Journal of Physical Chemistry C

DOI:10.1021/acs.jpcc.9b05891

Publication date:2019

Document version:Accepted manuscript

Citation for pulished version (APA):Lundehøj, L., Cellier, J., Forano, C., & Nielsen, U. G. (2019). Atomic Level Understanding of OrthophosphateAdsorption by Magnesium Aluminum-Layered Double Hydroxides - A Multitechnique Study. Journal of PhysicalChemistry C, 123(39), 24039-24050. https://doi.org/10.1021/acs.jpcc.9b05891

Go to publication entry in University of Southern Denmark's Research Portal

Terms of useThis work is brought to you by the University of Southern Denmark.Unless otherwise specified it has been shared according to the terms for self-archiving.If no other license is stated, these terms apply:

• You may download this work for personal use only. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying this open access versionIf you believe that this document breaches copyright please contact us providing details and we will investigate your claim.Please direct all enquiries to [email protected]

Download date: 14. Feb. 2022

https://doi.org/10.1021/acs.jpcc.9b05891

https://doi.org/10.1021/acs.jpcc.9b05891

https://portal.findresearcher.sdu.dk/en/publications/c4e57204-4c49-4263-9be2-0f7aa2ea520a

Subscriber access provided by University Library of Southern Denmark

is published by the American Chemical Society. 1155 Sixteenth Street N.W.,Washington, DC 20036Published by American Chemical Society. Copyright © American Chemical Society.However, no copyright claim is made to original U.S. Government works, or worksproduced by employees of any Commonwealth realm Crown government in thecourse of their duties.

C: Surfaces, Interfaces, Porous Materials, and Catalysis

An Atomic Level Understanding of OrthophosphateAdsorption by Magnesium Aluminum LayeredDouble Hydroxides – a Multi-Technique Study

Laura Lundehøj, Joel Cellier, Claude Forano, and Ulla Gro NielsenJ. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.9b05891 • Publication Date (Web): 10 Sep 2019

Downloaded from pubs.acs.org on September 11, 2019

Just Accepted

“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are postedonline prior to technical editing, formatting for publication and author proofing. The American ChemicalSociety provides “Just Accepted” as a service to the research community to expedite the disseminationof scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear infull in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fullypeer reviewed, but should not be considered the official version of record. They are citable by theDigital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,the “Just Accepted” Web site may not include all articles that will be published in the journal. Aftera manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Website and published as an ASAP article. Note that technical editing may introduce minor changesto the manuscript text and/or graphics which could affect content, and all legal disclaimers andethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors orconsequences arising from the use of information contained in these “Just Accepted” manuscripts.

1

An Atomic Level Understanding of Orthophosphate Adsorption by Magnesium Aluminum Layered Double Hydroxides – A Multi-Technique Study

Laura Lundehøj1, Joel Cellier2, Claude Forano2, and Ulla Gro Nielsen1,*

1 Department of Physics, Chemistry, and Pharmacy, University of Southern Denmark, Campusvej 55,

5230 Odense M, Denmark

2 Université Clermont Auvergne, CNRS, SIGMA Clermont, ICCF, F-63000 Clermont-Ferrand,

France

* Corresponding author: email: [email protected] phone +45 6550 4401

Abstract

The phosphate (P) adsorption properties of three magnesium aluminum layered double hydroxides

intercalated with nitrate and/or carbonate (MgAl-NO3/CO3 LDH) with targeted Mg:Al ratios of 2, 3,

and 4 were investigated to understand how phosphate interact with the LDH using powder X-ray

diffraction, solid state NMR, vibrational spectroscopy, scanning electron microscopy, and ICP-OES.

The solid products after phosphate exposure contained four different phosphate species which were

quantified by 31P MAS NMR: Phosphate intercalated in the interlayer (A), phosphate adsorbed to the

surface of the LDH particles (B and C), and phosphate adsorbed to amorphous aluminum hydroxide

(D). Their relative concentrations depend on the Mg:Al ratio, the phosphate loading, and the

intercalated anion in the parent LDH. Surprisingly, phosphate intercalated by anion exchange, which

has been assumed to be the dominant mechanism of P adsorption, is less than 30 %. Instead phosphate

adsorbed to the surface of the LDH particles is the predominant site (52-88 %). Moreover, a non-

negligible fraction (8-39 %) of phosphate is adsorbed to an amorphous aluminum hydroxide (AOH)

impurity (Site D) especially at high phosphate concentration and exposure time. AOH is a known

Page 1 of 36

ACS Paragon Plus Environment

The Journal of Physical Chemistry

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

mailto:[email protected]

2

impurity in LDH prepared by co-precipitation at constant pH, the most commonly used method in

phosphate removal studies. This indicates that MgAl-LDH have higher affinity for phosphate than

AOH. Furthermore, no significant dissolution of the LDH followed by precipitation of phosphate

minerals was observed by PXRD, 27Al and 31P MAS NMR.

Introduction

Phosphorus in the form of inorganic orthophosphate (P) is essential for plant growth and extensively

used as a fertilizer. Currently, phosphate rocks obtained from mining activities is the major source,

but the known deposits may be depleted within 50-100 years.1 Hence, phosphate rock is considered

a critical resource by the European Union.2 One of the most promising alternative phosphate sources

is domestic wastewater, which has a high concentration of phosphorus (5-20 mg/L). Furthermore,

recovery of phosphate from wastewater treatment plants (WWTP) would benefit the phosphorus

circular economy3 and simultaneously reduce the ecological impact (eutrophication) of the excessive

P discharge into surface water reservoirs. Currently, phosphate present at WWTP is mainly

concentrated in sludge, which is applied to agricultural fields. However, there is a growing concern

about contamination from pathogens, toxic organic compounds, and heavy metals. Consequently,

some European countries have banned application of sludge to fields.3-4 Alternative technologies to

sludge application include the recovery of phosphate by controlled precipitation of, e.g., struvite

(NH4MgPO46H2O) or by phosphate selective adsorbents.5-7 Layered double hydroxides (LDH), a

group of inorganic anion exchange materials related to the mineral hydrotalcite

(Mg6Al2(OH)16CO34H2O), are one of the most promising sorbent materials due to their unique anion

exchange properties and their high affinity for phosphate.8-11 The general formula for a LDH is [M2+1-

xM3+x(OH)2]x+(An-

x/n)yH2O, abbreviated as M2+M3+-A, where M2+, M3+, and A represent cations and

anions, respectively.12 Their structure consists of stacked, positively charged edge-sharing octahedra

Page 2 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

3

planes ([M2+1-xM3+

x(OH)2]x+) separated by hydrated anionic species (An-x/nyH2O) as illustrated in

Figure 1. Inorganic phosphate anions have been fully intercalated by anion exchange reaction in LDH

materials leading to pure phosphate containing LDH.13-16 Moreover, phosphate adsorption by LDH

have extensively been studied in the laboratory using pure phosphate solutions as well as solutions

from wastewater, sewage sludge filtrate, and seawater.9-11, 13, 15, 17-31 Furthermore, recyclability of

LDH materials has been demonstrated although a lower capacity has been reported after

regeneration.10, 18, 20-21, 32

Figure 1. Crystal structure of a Mg2Al-NO3 LDH and the different possible binding sites: Interlayer

phosphate (A) and surface adsorbed (B+C) related to the LDH, and phosphate adsorbed to

aluminum hydroxide (D).

Most studies have focused on the phosphate adsorption properties of the LDH under different

conditions (pH, phosphate concentration, LDH composition, exposure time and competing anions)

with emphasis on quantification of the amount of phosphate removed from the aqueous phase.8-10, 17,

19-21, 24-25, 32 In contrast, detailed structural studies of the phosphate adsorption mechanism by LDH,

i.e., exactly where and how phosphate species interact with the LDH have been limited, due to the

low crystallinity of P containing LDH, which is caused by structural disorder of the P anions in the

Page 3 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

4

interlayer domains. Powder X-ray diffraction (PXRD) of LDH intercalated with phosphate (M2+M3+-

P LDH) display only a few reflections of low intensity, which prevents refinement of the PXRD

profile. Thus, no crystal structure has been published for any of these M2+M3+-P LDH to our

knowledge. Several phosphate adsorption mechanisms including intercalation by anion-exchange,

surface adsorption, ligand complexation, and dissolution of the LDH followed by formation of

different phosphate minerals such as hopeite (Zn3(PO4)24H2O) and newberyite (MgHPO43H2O)

have been proposed based only on thermodynamic studies.11, 18, 23, 26-27, 31 Adsorption is commonly

understood as the overall uptake phenomenon irrespective of mechanism, whereas the term

theoretical adsorption capacity (TAC) refers to the maximum anion exchange capacity both in the

interlayer and on the surface of the LDH particles. The relative contribution from the suggested

mechanisms is essential to design an efficient and recyclable filter material but have to our knowledge

not been quantified. PXRD has extensively been used to probe intercalation of phosphate in the

interlayer, seen as a shift of the (00l) reflections towards low 2θ values and always associated with a

loss in crystallinity.13, 15-16, 18, 22-23, 27-28 Moreover, PXRD clearly shows that the CaAl-LDH structure

is destroyed upon exposure to phosphate resulting in the formation of calcium phosphates such as

brushite (CaHPO42H2O) and hydroxyapatite (Ca5(PO4)3(OH)).23, 29-30 However, it has also become

evident that the presence of amorphous phases are not detected by PXRD26, 33 and sorption to the

LDH surface cannot be probed by PXRD. Furthermore, the LDH used for phosphate adsorption are

normally prepared by co-precipitation at constant pH or at increasing pH using thermal hydrolysis of

urea (the so-called “urea method”).34 These two synthesis methods result in variable quantities (10-

60 %) of amorphous aluminum hydroxide (AOH) impurities.33, 35 Such AOH has clearly been

identified by 27Al SSNMR for MgAl- and ZnAl-LDH phases used for phosphate adsorption.28, 36-38

Moreover, the presence of amorphous aluminum hydroxides is critical as these have a high affinity

for phosphate39-41 and will lower the actual Al content in the LDH,33, 35-36, 42 i.e., result in a lower

Page 4 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

5

TAC. Thus, experimental techniques with atomic level resolution, which do not require long-range

order such as solid-state NMR spectroscopy (SSNMR), X-ray extended absorption spectroscopy

(EXAFS), vibrational spectroscopy (FT-IR and Raman) is needed to probe surface complexation and

amorphous phases.

SSNMR has provided valuable information about adsorption of phosphate by LDH and can be used

for quantification of the different phosphate sites26-27, 29, 36 and has as well given detailed insight into

phosphate sorption onto aluminum hydroxides.40-41 For example, SSNMR quantified phosphate

adsorbed to AOH and aluminum phosphate formed by dissolution reprecipitation36 in our recent paper

on removal of phosphate from acidified sludge by ZnAl-NO3 LDH. Moreover, SSNMR of LDH

identified ordering of the M(III) ions, structural defects, and amorphous impurities.33, 35, 43-48

Few studies have investigated the phosphate adsorption by LDH using SSNMR. These report partial

dissolution of the LDH at a pH ≤ 7, which is accompanied by precipitation of phosphate minerals,

e.g., aluminum phosphate (AlPO4), bobierrite (Mg3(PO4)28H2O), newberyite or brushite.27, 36, 49

Phosphate adsorption experiments under basic conditions show no precipitation of phosphate

minerals according to SSNMR and/or PXRD.23, 50 However, quantification of the different binding

sites on the LDH as well as the contribution from the amorphous AOH phase has not been investigated

in detail.

Here we report a detailed study of the phosphate adsorption properties of a series of MgRAl-NO3

LDH (R = 2, 3, 4) using a combination of phosphate adsorption experiments with detailed

characterization of the solid LDH products by PXRD, FT-IR, Raman, ICP-OES, and multi-nuclear

(1H, 25Mg, 27Al, and 31P) SSNMR as a function of phosphate loading to gain an atomic level

understanding on how phosphate interact with the LDH as well as identify and quantify the different

phosphate species. PXRD was used for identification of the intercalated anions and to monitor the

Page 5 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

6

crystallinity of LDH as a function of phosphate loading, whereas 25Mg and 27Al MAS NMR provided

information on the stability of the LDH and 31P MAS NMR was used for identification and

quantification of the phosphate species (P-species).

Experimental

The following reagents were used: magnesium nitrate hexahydrate (Mg(NO3)26H2O), aluminum

nitrate nonahydrate (Al(NO3)39H2O), dipotassium phosphate (K2HPO4), and sodium hydroxide

(NaOH). All chemicals were of reagent grade. K2HPO4 (pKa = 7.21 for HPO42-) was used as the

source for phosphate to ensure alkaline condition and avoid dissolution of the LDH.

Table 1. Synthesis parameters (pH, aging time, and Mg:Al ratio of the metal salt solutions) and the

properties of the parent MgRAl-LDH (Al content determined by ICP-OES, the Mg:Al ratio

determined from ICP-OES (bulk) as well 27Al and 1H MAS NMR (LDH), the intercalated anion based

on PXRD (intercalated), FT-IR, and Raman, the AOH content from 27Al MAS NMR, and the

experimental adsorption capacity (EAC).

Synthesis parameters

Sample Characteristics

Sample Mg:Al pH Bulk Al (%)

Mg:Al LDH (27Al)

Mg:Al LDH(1H)

Anion (PXRD)

Anion (FT-IR, Raman)

AOH (%Al)

EAC (mg P/g])

M2A 2.0 9.0(4) 34.6 2.7(6) 2.6(6) NO3- CO3

2-, NO3- 30(6) 47

M3A 3.0 9.5(3) 26.7 3.9(6) - NO3- CO3

2-, NO3- 30(6) 46

M4A 4.0 10.0(3) 22.4 5.2 (6) 4.9(6) CO32- CO3

2-, NO3- 33(6) 17

MgAl-LDH Syntheses:

MgAl-NO3 LDHs with a targeted Mg:Al ratio of 2:1, 3:1, and 4:1, respectively were synthesized by

coprecipitation at constant pH,33 as summarized in Table S1. The pH of the synthesis solution was

increased with increasing Mg:Al molar ratio in order to insure a total precipitation of both Mg2+ and

Page 6 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

7

Al3+ cations. A solution of the metals salt with a total metal concentration of 1 M and the desired

Mg:Al ratio was added by a peristaltic pump at a of rate of 0.1 ml/min to deionized water in a reactor.

A nitrogen (N2) atmosphere was used to minimize carbonate contamination. The pH of the synthesis

solution was kept constant by simultaneous addition of a 2 M NaOH solution by a peristaltic pump.

Subsequently, the synthesis mixtures were aged for 61-62 h and then collected by centrifugation

before the solid product was washed thrice with deionized water and dried at room temperature.

Aging was chosen as post synthesis treatment over hydrothermal treatment to mimic large scale

production of LDH. The parent materials were named according to the Mg:Al ratio of the metal salt

solution: M2A, M3A, and M4A.

Phosphate adsorption isotherms: The adsorption experiments were carried out in 2 ml Eppendorf

vials, which contained 10 mg LDH and 1 mL phosphate solution. The appropriate amount of

phosphate was taken from a 0.05 M K2HPO4 stock solution and deionized water was added to a total

volume of 1 ml. The samples were shaken overnight after which the supernatant was separated by

centrifugation. The supernatant was diluted by a factor of 50 prior to quantification of the HPO42-

concentration by ICP-OES. The equilibrium phosphate adsorption content, Cs, was determined and

plotted as a function of the equilibrium concentration, Ceq.

(1)Cs = (C0 ― Ceq) ∙vm

where C0 (mol·L-1) is the initial P concentration, v (L) is the total volume, and m (g) is the mass of

the LDH sample.

Preparation of phosphate loaded samples for characterization by PXRD, Raman, FTIR, ICP-OES,

and SSNMR:

Page 7 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

8

0.300 g LDH was suspended in 50 ml deionized water in which a 1 M K2HPO4 solution was added

in amount corresponding to 25, 50, 75, 100, and 400 % of the theoretical HPO42- anion exchange

capacity (TAC) of the MgRAl-NO3 LDH. The TAC was calculated under the assumption that nitrate

can be completely exchanged for hydrogen phosphate (HPO42-) by anion exchange and that the LDH

contain two water molecules. See Tables S1 and S2 for further information. The 25, 50, 75, and 100

% samples were exposed to phosphate for five hours whereas the 400 % samples were exposed to

phosphate for 62 h to ensure full saturation of the LDH. All adsorption experiments were performed

under a N2 atmosphere. The solid was collected by centrifugation and washed thrice with deionized

water before dried at room temperature. The samples were named according the parent LDH (M2A,

M3A, and M4A) and the initial P loading (Y): MRAYP. For example, M2A100P was synthesized

with a Mg:Al ratio of two and exposed to a phosphate concentration corresponding to 100 % of the

TAC.

Characterization of the solid samples:

Powder X-ray Diffraction (PXRD): The powder X-ray diffractograms were recorded on a Rigaku

Miniflex 600 with Cu Kα radiation (λ = 1.5418 Å) in the 5-70 ° range using a step size of 0.02 ° and

a speed of 10 °min-1. In-situ powder X-ray diffractograms were recorded with a Philips X'Pert Pro

diffractometer using CuKα radiation (λ = 1.5405 Å) over the 3–75 ° 2θ range at an accelerating

voltage of 40 kV, a current of 30 mA and a scan speed of 2 °min−1 using the XRD flow cell described

elsewhere.29

FT-IR and Raman spectroscopy: FT-IR spectra were recorded from 4000-400 cm-1 on a Nicolet-5700

spectrophotometer by the KBr pellet method. The pellet was made by grinding approx. 2 mg of the

sample and approx. 198 mg KBr. The Raman spectra were recorded from 150 to 1500 cm−1 at room

temperature using a confocal micro-Raman spectrometer (T64000 Jobin Yvon) with an excitation

Page 8 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

9

wavelength of 514.5 nm (argon-ion laser-line). Spectra were acquired at a spectral resolution of 1

cm−1 with charge coupled device (CCD) multichannel detector cooled by liquid nitrogen to 140 K

coupled with an Olympus confocal microscope.

Scanning Electron Microscopy: SEM images were recorded using a JSM-7500F FESEM operating

at 3 KV with the samples mounted on conductive carbon adhesive tape and coated with a thin gold

layer. Images were acquired at magnifications of x10,000 and x25,000 For each sample, SEM images

did not show variation in size and morphology of particles then SEM images were selected in term

of image quality (best contrast).

SSNMR: Single pulse 27Al and 31P MAS NMR spectra were recorded on a JEOL ECZ500CR MHz

NMR spectrometer equipped with a 11.7 T Oxford magnet and a 3.2 mm double resonance MAS

NMR probe using a spinning speed of 13 kHz. The 31P MAS NMR spectra were recorded with a 45

° pulse and a relaxation time of 30-60 s, the latter was optimized for each sample to ensure full

relaxation. The 27Al MAS NMR spectra were recorded with a relaxation time of 1 s and a 10 ° pulse

angle to ensure quantitative spectra. The 27Al 3QMAS NMR spectra were recorded with a z-filter and

a spinning speed of 14 kHz. 25Mg MAS NMR spectra were recorded on a Varian 600 MHz NMR

spectrometer (14.1 T) with a 5 mm double resonance MAS NMR probe with a Hahn echo DFS pulse

sequence,47 a relaxation delay of 1 s, and a spinning speed of 7-9 kHz. 1H MAS NMR spectra were

recorded on Varian 600 MHz NMR spectrometer (14.1 T) with a 1.6 mm triple resonance MAS NMR

probe. 1H MAS NMR spectra were recorded with a relaxation delay of 4-8 s and with 30-35 kHz as

spinning speed. The 1H, 25Mg, 31P, and 27Al MAS NMR spectra were referenced from water (δiso(1H)

= 4.6 ppm), MgO in the rotor (δiso(25Mg) = 25 ppm51), 1 M AlCl3 solution (δiso(27Al) = 0 ppm), and

phosphoric acid H3PO4 (δiso(31P) = 0 ppm), respectively. The 25Mg and 27Al MAS NMR spectra were

analyzed by Quadfit52 whereas the 1H and 31P MAS NMR spectra were analyzed by MestReNova.

Page 9 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

10

Results and Discussion

Characterization of the parent LDH materials prior to phosphate adsorption

Figure 2. PXRD for the three series a) M2A, b) M3A, and c) M4A at the P loadings of 0P (parent

LDH), 25P, 50P, 75P, 100P, and 400P. Reflections from the aluminum background in the sample

holder (#) are seen for some samples due to limited sample quantities.

The parent LDH were characterized by 27Al MAS NMR, ICP-OES, FT-IR, Raman and PXRD prior

to phosphate adsorption to assess the sample composition and structure. The powder X-ray

diffractograms of M2A, M3A, and M4A (Figures 2 and S1a) are characteristic of a LDH with a

hexagonal lattice and rhombohedral symmetry (R-3m).53 Only reflections that could be assigned to

the LDH is seen, i.e., no crystalline impurities are detected. The linear (R2 = 0.998) increase of the

M-M distance (a) confirms the change in Mg:Al composition of the layer while the decrease of the

basal spacing (c) with Mg:Al indicates a partial substitution of NO3- by CO3

2- (Figure S1b).53 FT-IR

and Raman spectra (Figures 3, 4 and S2) confirm the formation of MgAl LDH and exhibit the typical

infrared vibration bands of the interlayer anions with (NO3) and (CO3) at 1384 cm-1 and 1361 cm-1

(FT-IR) and at 1055 cm-1 and 1044 cm-1 (Raman). However, the (NO3) band for M4A has almost no

Page 10 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

11

intensity in the Raman spectrum in agreement with the lower basal spacing for M4A as compared to

M2A and M3A (Figure S1b). Hence, carbonate is mainly intercalated in M4A. Moreover, the MgRAl

lattice vibrations ((MO6) and (OMO)) are in the range 667-586 cm-1 and 451-400 cm-1 (FT-IR) and

545-557, 478-467 cm-1 (Raman) and varies in agreement with the Mg:Al ratio.8, 13, 15, 22, 33, 35

Figure 3. An expansion of the FT-IR spectra of the region within anion vibrations (900 to 1500

cm-1) for the three series: a) M2A, b) M3A, and c) M4A at P loading of 0P (parent LDH), 25P, 50P,

75P, and 100P. The bands v3(NO3-), v3(CO3

2-), and v3(PO43-) are marked with blue lines.

Page 11 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

12

800 850 900 950 1000 1050 1100 1150 1200

Abs

orba

nce

(a.u

.)

Raman shift (cm-1)

M2A400P

M2A00P

M2A75P

M2A50P

M2A25P

M2A0P

931987

10631(HPO4

2-)

1000 1010 1020 1030 1040 1050 1060 1070 1080 1090 1100

4(HPO42-)

as(NO3)

M2A400P

M2A00P

M2A75P

M2A50P

M2A25P

M2A0P

as(CO3)

Abs

orba

nce

(a.u

.)

Raman shift (cm-1)400 450 500 550 600 650 700 750 800

M2A400P

M2A100P

M2A75P

M2A50P

M2A25P

M2A0P

556

482

Raman shift (cm-1)

Abs

orba

nce

(a.u

.)

MO6

OMO

400 450 500 550 600 650 700 750 800

Abs

orba

nce

(a.u

.)

Raman shift (cm-1)

M3A400P

M3A100P

M3A75P

M3A50P

M3A25P

M3A0P

551

476

MO6

OMO

800 850 900 950 1000 1050 1100 1150 1200

Abs

orba

nce

(a.u

.)

M3A400P

M3A100P

M3A75P

M3A50P

M3A25PM3A0P

Raman shift (cm-1)

1062

931982

400 500 600 700 800

Abs

orba

nce

(a.u

.)

Raman shift (cm-1)

M4A50P

M425P

M4A0P

468546

512

800 850 900 950 1000 1050 1100 1150 1200

Abs

orba

nce

(a.u

.)

M4A50P

M425P

M4A0P

Raman shift (cm-1)

M4A75P

1000 1010 1020 1030 1040 1050 1060 1070 1080 1090 1100

Abs

orba

nce

(a.u

.)

M4A50P

M425P

M4A0P

Raman shift (cm-1)

as(CO3)1044

1000 1010 1020 1030 1040

1050

1060 1070 1080 1090 1100

Abs

orba

nce

(a.u

.)

Raman shift (cm-1)

M3A400P

M3A100P

M3A75P

M3A50P

M3A25P

M3A0P as(NO3)

as(CO3)

1062

1055

1044

a) M2A

b) M3A

c) M4A

M4A75P

M4A400P

M4100P

M4A75P

M4A100P

M4A400P

M4A100P

M4A400P10584(HPO4

2-)

1(PO43-)

1(HPO42-)

1(PO43-) 4(HPO4

2-)

4(HPO42-)

4(HPO42-)

4(HPO42-)

1062

1044

1055

Figure 4. Raman spectra for the three series: a) M2A, b) M3A, and c) M4A at P loadings of 0P

(parent LDH), 25P, 50P, 75P, 100P, and 400P.

Whereas PXRD imply highly crystalline samples without crystalline impurities, analysis of the 27Al

MAS NMR spectra (Figure 5a) show the presence of amorphous aluminum hydroxide (AOH),33, 35,

42 which is evident as a broad signal with iso(27Al) 10.2-11.3 ppm in the 27Al MAS NMR spectra.

The AOH is also easily identified in the 27Al 3QMAS NMR spectra (Figure S3). Hence,

deconvolution of the single pulse 27Al MAS NMR spectra were performed with to sites – one for the

LDH with CQ = 1.6 MHz and η = 0 and one for the AOH phase with CQ = 3.3 and η = 0.2, as reported

earlier.42 The presence of AOH was confirmed by the Al content determined by ICP-OES for M2A

(34.6 %) which is above 33 %, the maximum Al content in an LDH considering the Al-O-Al

Page 12 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

13

avoidance rule.43-44 From the simulations it was determined that Al in the LDH phase constitutes

approximately two thirds of the total Al content. From this and the ICP-OES results, the Mg:Al ratio

of the LDH can be calculated (Table 1). Deconvolution of 1H MAS NMR spectra of M2A and M4A

(Figure S4) as well confirmed the presence of AOH.44 Moreover, the Mg:Al ratio was found to 2.6(6)

for M2A and 4.9(6) for M4A, which matches the Mg:Al ratio determined by 27Al MAS NMR

combined with ICP-OES within the experimental uncertainty (Table 1). See Table S3 and Figure S4

for further details.

Page 13 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

14

Figure 5. a) Simulation of the 27Al MAS NMR spectrum of M2A, b) Relative intensity of the LDH

resonance for the three series: M2A, M3A, and M4A at the different P loadings, and c) δiso(27Al) for

the LDH resonance for the three series: M2A, M3A, and M4A at the different P loadings.

The 25Mg MAS NMR spectra M2A, M3A, and M4A (Figure 6) match earlier reported 25Mg MAS

NMR spectra of MgAl-LDH.43-44 An ideal Mg2Al-LDH contains only a single 25Mg resonance: a

Page 14 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

15

25Mg with three Mg and three Al in the first metal ion coordination sphere (Mg(OAl)3(OMg)3).44

However, the 25Mg MAS NMR spectrum of M2A cannot be satisfyingly simulated with a single 25Mg

site (Figure S5) implying a lower Al content (Mg:Al > 2) in the LDH.44 This is in good agreement

with the Mg:Al of 2.6(6) (Table 1) obtained from deconvolution of the 1H MAS NMR spectrum

(Figure S4).

Simulations of the 25Mg MAS NMR spectra of M3A and M4A were not attempted, but visual

inspection shows that the two samples have less Al than the M2A sample seen as the high frequency

brucite like Mg(OMg)6 site observed at approximately 0 ppm.44 This site is most intense for M4A

implying that the Al content in the LDH phase is M2A > M3A > M4A, in agreement with the Mg:Al

ratio of the synthesis solutions.

Page 15 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

16

Figure 6. 25Mg MAS NMR spectra for M2A0P (parent LDH, spinning speed = 8 kHz), M2A100P

(spinning speed = 8 kHz), M3A0P (parent LDH, spinning speed = 7 kHz), M3A100P (spinning speed

= 9 kHz), M4A0P (parent LDH, spinning speed = 9 kHz), and M4A100P (spinning speed = 9 kHz).

* and # marks spinning side band and MgO from the rotor, respectively.

Phosphate adsorption capacity

The TAC for a MgRAl-LDH with nitrate in the interlayer under the assumption of anion exchange as

the only mechanism for phosphate adsorption, is directly proportional to the aluminum content in the

LDH. It also depends on the pH-controlled speciation of the orthophosphate anions (H2PO4-/HPO4

2-

/PO43, Table S2). For example, the exchange reaction for intercalation of HPO4

2- species from a

solution at pH = 9.79 ((pKa2+pKa3)/2) is:

Page 16 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

17

𝑀𝑔1 ― 𝑥𝐴𝑙𝑥(𝑂𝐻)2(𝑁𝑂3)𝑥 ∙ 2𝑥𝐻2𝑂 +𝑥2 ∙ 𝐻𝑃𝑂2 ―

4 → 𝑀𝑔1 ― 𝑥𝐴𝑙𝑥(𝑂𝐻)2(𝐻𝑃𝑂4)𝑥2

∙ 2𝑥𝐻2𝑂 + 𝑥 ∙ 𝑁𝑂 ―3

(2)

For each Al half a monohydrogen phosphate ion, HPO42-, can theoretically be intercalated and one

nitrate is then expelled. The TACs were estimated under the assumption of two water molecules per

Al. However, the experimental adsorption capacity (EAC) should be lower than TAC if carbonate is

partly intercalated, since carbonate will not be completely replaced by phosphate due to the higher

affinity of LDH for carbonate.8 How the degree of protonation of the phosphate ion (PO43-, HPO4

2-,

or

H2PO4-) , which affects the TAC is reported in Tables S2 and S4. For example, twice as much H2PO4

-

can be intercalated as HPO42-.

The HPO42- adsorption isotherms of the three LDH samples (M2A, M3A, and M4A; Figure S6)

correspond to a H type isotherms according to Giles classification54 with a strong adsorption at low

HPO42- concentration followed by a continuous increase of adsorption at medium and high

concentration indicating several mechanisms of adsorption. The phosphate adsorption isotherms did

not completely reach a plateau even at high phosphate loading implying two (or more) adsorption

mechanisms, which are hard to model with either the Langmuir or the Freundlich model. We ascribe

this phenomenon in part to the presence of AOH in the samples. Thus, the adsorption capacities

corresponding to the anion exchange reaction were estimated as the kink of the isotherms (Figure S6)

and are summarized in Table 1 (EAC). As already mentioned by Everaert et al. for similar MgRAl-

NO3 phases, LDH with Mg:Al ratio equal to 2 and 3 display similar adsorption behavior and

capacities toward HPO42- anions while Mg4Al has a much lower efficiency to remove HPO4

2-.55 The

adsorption capacities for our three LDH are in the same range as earlier reported adsorption capacities

for LDH.8, 11, 21, 31-32 Furthermore, the EAC for M2A and M3A but not M4A match the TACs (Table

Page 17 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

18

S4): 50mg P/g, 40 mg P/g, and 34 mgP/g for a Mg2.6Al-NO3, Mg3.9Al-NO3, and Mg5.2Al-NO3 LDH,

respectively. However, taking into account the presence of 30% and 33% AOH in M2A, M3A and

M4A (Table 1), the corresponding TACs for the LDH are reduced to 35 mgP/g, 28 mgP/g, and 23

mgP/g, respectively. Hence, EAC for M2A and M3A are greater than the expected TAC (134 % and

164 % respectively), indicating that anion exchange may not be the only phosphate uptake

phenomenon and/or that the identified AOH contributes. The adsorption capacity for M4A is 74 %

of the TAC due to the presence of mainly carbonate in the parent M4A LDH.

Characterization of the phosphate loaded LDH samples

The phosphate loaded samples were characterized by a combination of ICP-OES, PXRD, FT-IR and

Raman spectroscopy as well as 1H, 25Mg, 27Al, and 31P MAS NMR to determine the phosphate

adsorption mechanism.

Phosphate content in the LDH samples:

The Mg:Al ratio (ICP-OES) and LDH content in the bulk samples (27Al MAS NMR) showed no

significant changes upon phosphate exposure, c.f., Table S5 and Figure 5b, respectively. Moreover,

the (110) reflection, which reflects the Mg content in the cation layer,12 is not shifted implying that

the Mg:Al ratio of the LDH is also unchanged (Figure 2). Thus, there is no indication of dissolution

or chemical transformation of the LDH upon phosphate exposure even at a P loading of 400 % for 62

h. The phosphate concentration at the different loadings for the three series is given in Figure 7.

Generally, the phosphate content in the solid phase increases with the phosphate concentration in the

solution for the three series. The phosphate content for the three series at the two highest phosphate

loadings (100P and 400P) matches very well the adsorption capacity determined from the phosphate

Page 18 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

19

adsorption isotherms given in Table 1 except for M4A400P (39 mgP/g), which is more than twice the

EAC (17 mgP/g) due to the high phosphate concentration and exposure time (62 h).

Figure 7. The amount of adsorbed P of for the three series: M2A, M3A, and M4A determined by

ICP-OES. The 25P, 50P, 75P, 100P were exposed to phosphate for 5 h. The 400P samples were

exposed for 62 h. The experiments were performed under alkaline conditions (pH 9-10).

Identification and quantification of the phosphate species by 31P MAS NMR: The 31P MAS NMR

spectra contains up to four different local environments (resonances): A (δiso(31P) = 8.3-8.7 ppm), B

(δiso(31P) = 2.3-3.4 ppm), C (δiso(31P) = -1.9 to -0.7 ppm), and D (δiso(31P) = -7.0 to -5.1 ppm), as

illustrated in Figure 8. The δiso(31P) for all sites decreases with the Mg:Al ratio (lower Al%) (Table

2). The relative intensity depends on the P loading and the Mg:Al ratio of the parent LDH, as

illustrated in Figures 8 and 9. The remaining 31P MAS NMR spectra (50P and 75P loading) are

presented in Figure S7. A fifth resonance with δiso(31P) 6 ppm from K-struvite,37 which constitutes

less than 5 % of the total intensity may be present, but cannot be unambiguously confirmed as it

overlaps with the A and B sites. The potassium could originate from the K2HPO4 stock solution. The

δiso(31P) and the relative intensity of the four resonances obtained from deconvolution of the 31P MAS

NMR spectra are reported in Table 2 and Figure 9, respectively. A chemical exchange between

Page 19 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

20

carbonate and bicarbonate has been reported for LDH.56 Thus, 31P-31P EXSY MAS NMR spectra

were recorded for M2A100P and M4A100P to probe dynamic exchange, e.g., between two phosphate

species in the interlayer, but no sign of exchange were observed (Figure S8).

Figure 8. 31P MAS spectra for the three series; M2A, M3A, and M4A with P loadings of 25, 100 and

400. The sites A, B, C, and D are marked. The top line is the experimental data, the second is the

simulated and the bottom line is the residual.

Assignment of site A: Site A is only present for the M2A (11-29 %) and M3A (0-17%) series and not

M4A series. Thus, site A must be due to a distinct difference between M4A and the other two. A

comparison of the PXRD, FT-IR, and Raman as well as the 25Mg and 27Al MAS NMR data for the

Page 20 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

21

three series reveal that phosphate intercalates in the M2A and M3A LDH, but not in the M4A LDH

as explained below.

Table 2. δiso(31P) for A, B, C, and D obtained from deconvolution of the 31P MAS NMR spectra and

assignment of the P-species.

δiso(31P) [ppm]Site

M2A M3A M4A

Assignment

A 8.5-8.7 8.3-8.6 - Intercalated

B 3.0-3.4 2.6-3.2 2.3-2.5 Surface

C -1.3 to -0.7 -1.4 to -0.9 -1.9 to -1.1 Surface

D -5.9 to -5.1 -6.1 to -5.3 -7.0 to -6.2 AOH

The basal spacing of the LDH and thereby the intercalated anion can be examined from the position

of the (003) reflection in the powder X-ray diffractograms. Previous PXRD studies suggest that both

H2PO4-, HPO4

2-, and PO43- can be intercalated in MgAl- and ZnAl-LDH depending on the pH

conditions.13, 15-16, 36, 50 In addition, the drying temperature of the sample affect the position of the

(003) reflection, which complicates a comparison with earlier studies, as details about sample drying

is often missing. For example, a decrease of the interlayer distance from d003 = 10.9 Å to d003 = 8.2

Å was seen for ZnAl-HPO42- LDH when drying temperature was raised from 20 to 60 °C.16

Upon phosphate exposure the (003) reflection for M2A and M3A shifts from 2 = 10.2 ° (8.7 Å) to

11.1 ° (8.0 Å) and 2θ = 10.5 ° (8.5 Å) to 2θ = 11.2 ° (7.9 Å), respectively, corresponding to the

exchange of nitrate by carbonate in the parent sample (Figures 2a and 2b).57-58 At a higher phosphate

loadings the (003) reflection for M2A first broadens (50P) and subsequently split in two reflections

at 2θ = 8.8 ° (10.0 Å) and 11.1 ° (8.0 Å), which are assigned to monohydrogen phosphate (HPO42-)

Page 21 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

22

and carbonate intercalated in the LDH interlayer, respectively.15, 58 For M3A, the (003) reflection

only broadens and in agreement with that the content of site A in the 31P MAS NMR spectra is lower

than for M2A, c.f. (Figure 2b and 9). The shoulder at 2 = 8.8 ° (10.0 Å) at the higher P loading as

well implies intercalation of some HPO42- into the LDH,15 but carbonate remains the predominant

interlayer anion based on PXRD (Figure 2b). M4A contains only carbonate in the interlayer based on

the location of the (003) reflection at 2θ = 11.2 ° (7.9 Å; Figure 2c)58 and a unique band at 1044 cm-

1 in the Raman spectrum (Figure 4) from carbonate.33, 35 The position of the (003) reflection is

unchanged upon exposure to phosphate implying no intercalation of phosphate at any P loading.

In addition, a kinetic study of HPO42- adsorption by M2A by in-situ PXRD (Figure S9) clearly shows

the progressive exchange of the nitrate by HPO42-. Exchanged M2A-HPO4

2- LDH display (003), (006)

and (009) diffractions lines with dhkl respectively at 2 8.17 °, 16.39 °, and 24.52 °, respectively

resulting in a basal spacing of 10.84 Å.

The v4 band for phosphate at ~1070 cm-1 in the FT-IR spectra (Figure 3) and 1062 cm-1 Raman spectra

(Figure 4) is observed for the three series after phosphate exposure with the v1 band at 982 – 987

cm-1 assigned to HPO42- species with C3v symmetry (Raman).59 However, a broadening of the v(OH)

vibration at ~ 3000 cm-1 in the FT-IR spectra (Figure S10), which earlier has been assigned to

hydrogen bonding between the water in the interlayer and the intercalated phosphate15, 22 is only

observed for M2A and M3A and not for M4A. In addition, for the M2A and M3A series (Figures 3a

and 3b), the intensity of the phosphate band increases with the P loading and the intensity of the

nitrate band simultaneously decreases. Whereas only a small increase in the v4(HPO42-) band is

observed as the P loading is increased for the carbonate containing M4A LDH (Figures 3c and 4c),

indicating less phosphate adsorption in agreement with the adsorption isotherms (Figure S6). Hence,

mainly nitrate and not carbonate is exchanged with phosphate upon phosphate adsorption.

Page 22 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

23

Furthermore, the 25Mg and 27Al MAS NMR spectra imply that phosphate is intercalated in the M2A

and M3A LDH and not in the M4A LDH. No changes are seen in the 25Mg MAS NMR spectra of

M4A upon phosphate exposure (100P), whereas the line shapes in the 25Mg MAS NMR spectra of

M2A and M3A series broadens (Figure 6). The δiso(25Mg) depends on the intercalated anion.44

Consequently, the broadening of the signal is assumed to be caused by a distribution in δiso(25Mg) due

to the intercalation of phosphate and/or the loss of crystallinity (vide infra) and therefore only

observed for the M2A and M3A series. Similarly, the δiso(27Al) shifts from 11.7 ppm (parent LDH)

to 10.5 ppm (400P) for the M2A series and from 11.0 ppm to 10.6 ppm for the M3A series, whereas

no change is seen for M4A (Figure 5c). iso(27Al) for MgAl-LDH also depends on both the intercalated

anion as well as metal ratio.33 The change in δiso(27Al) for the M2A (1.2 ppm) and M3A (0.4 ppm)

series is therefore assigned to the exchange of nitrate with phosphate in agreement with PXRD, FT-

IR, and Raman data.

Thus, the difference between the M2A, M3A, and M4A series is that phosphate only intercalates in

M2A and M3A and not M4A. Hence, site A, which is only present for M2A and M3A, is assigned to

phosphate intercalated in the LDH interlayer. Intercalation is not the dominating mechanism for the

phosphate adsorption, as site A constitute 0-30 % of the total intensity.

Furthermore, it is observed that the intercalation results in a broadening and significant decreases in

the relative intensity of the (003) reflection for M2A and M3A (especially M2A) compared to the

(110) reflection. Thus, the crystallinity of the LDH is lowered along the c-axis implying a high degree

of stacking faults of the LDH layers and/or thinner LDH particles upon phosphate intercalation, which

suggests strong structural disorder of the HPO42- anions in the interlayer galleries, in agreement with

the increased distribution in the 25Mg MAS NMR spectra.

Page 23 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

24

Figure 9. Relative integrals for A, B, C, and D obtained from deconvolution of the 31P MAS NMR

spectra for the three series; M2A, M3A, and M4A at the different P loadings.

Assignment of site D: Site D is observed in all 31P MAS NMR spectra with relative concentration of

13-27 % for M2A, 8-28 % for M3A and 12-39 % for M4A. It is assigned to phosphate adsorbed to

AOH based on its δiso(31P).39, 60 The samples contain approx. 30 % AOH based on 27Al MAS NMR,

c.f., Table 1. Moreover, two weak reflections at 2θ = 18.5 ° and 20.4 ° from gibbsite (Al(OH3), ICSD

240781) are present for the M2A series (Figure 2a). This 31P resonance was also seen for a Zn2.5Al-

Page 24 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

25

LDH with the same AOH phase exposed to phosphate rich acidified sludge.36 Hence, AOH

contributes to 8-39 % of the total P adsorption capacity in our samples (Figure 9). The relative

concentration of site D is largest at the highest P loading (400P) for all three LDH series, indicating

that LDH has higher affinity for phosphate than AOH.

Assignment of site B and C: The two remaining sites, B and C must then be HPO42- and/or PO4

3-

phosphate adsorbed to the surface and/or edges of the LDH particles or alternatively, poorly defined

magnesium phosphate precipitates. However, there is no indication of dissolution of the LDH (vide

supra). The 27Al 3QMAS NMR spectra for M2A and M3A slightly changes upon phosphate exposure

(100P), which are probably caused by the low crystallinity and change in δiso(27Al) (Figure S3). Two

sites are still present and simulations of the single pulse 27Al MAS NMR spectra show that the LDH

content is identical within the experimental uncertainties for all samples (Figure 5b) and no significant

changes is observed in the Mg:Al ratio upon phosphate exposure (Table S5). The single pulse 27Al

MAS NMR spectra for the parent LDH, 25P, and 400P for the three series are given in Figure S11.

In addition, the δiso(31P) does not match common magnesium phosphates, e.g., newberyite (δiso(31P)

= -7.2 to -8.0 ppm), bobierrite (δiso(31P) = 4-4.7 ppm), farringonite (Mg3(PO4)2, δiso(31P) = -0.5 ppm),

or cattiite (Mg3(PO4)222H2O, δiso(31P) = 1.1 ppm).27, 61-63 Hence, dissolution of LDH and

reprecipitation of magnesium phosphates can be excluded.

To identify B and C we instead focus on the M4A series. No significant loss in the intensity for the

(003) and (006) reflection for the M4A series is seen upon phosphate exposure. Consequently, no

exfoliation occurs for this series and neither B nor C is phosphate adsorbed to exfoliated single sheets.

Similarly, as earlier mentioned no change is seen in the 25Mg and 27Al MAS NMR spectra. Thus, B

and C are not associated with changes in the overall structure (Mg:Al ratio, stacking, crystallinity

etc.) or in the local environment around Mg or Al in the bulk LDH. 1H MAS NMR spectra of

M4A100P and M4A were recorded to investigate changes in the local environment around the protons

Page 25 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

26

(Figure S3). The same five resonances as for M4A with δiso(1H) = 0.7 ppm (Mg3OH), 2.1 ppm

(Mg2AlOH), 3.1 ppm, 4.6 (H2O) ppm, and 6.10 ppm are observed. Deconvolution of the 1H NMR

spectra showed no change in the Mg:Al ratio of the LDH in agreement with PXRD. However,

comparison of the spectra show an increased intensity in the region δiso(1H) 5-10 ppm region, where

resonances from protonated phosphate is expected.44, 64 The 1H MAS NMR spectrum for M2A100P

exhibit a larger intensity increase in this region indicating a higher concentration of protonated

phosphate than in M4A100P in agreement with the higher EAC for M2A (Table 1). However, the 1H

MAS NMR spectra do not provide more information on the binding mode of phosphate for sites B,

C, or D.

This leaves phosphate adsorbed to the surface of the LDH particles as the only binding site for the B

and C P-species, which is not likely to change the bulk properties of the samples. Whatever their

chemical composition, the surface of the MgAl-LDH particles will be positive due to a point of zero

charge (PZC) at 12 or above.65-66 Our adsorption experiment were performed below the PZC

implying a highly positive surface, which will favor adsorption of phosphate on the surface of the

LDH particles.

B and C are two sites with different binding mode (monodentate or bidentate), different binding sites

on the surface of the LDH particles, and/or different degree of protonation. Site B is the major P-

species in all samples with a relative concentration of 45-52 %, 56-74 %, and 54-78 % for M2A, M3A

and M4A, respectively, whereas the relative concentration of C is 5-25 %, 6-20 %, and 8-20 % for

M2A, M3A and M4A, respectively. Thus, surface adsorption and not anion exchange in the interlayer

is clearly the dominating pathway for phosphate adsorption. This can be explained by the particle

morphology observed by SEM giving a higher surface area than a smooth surface (Figure S12). The

as-prepared LDH display a homogeneous distribution of small hexagonal platelets, 15-25 nm thick

and 100-150 nm large, leading to high concentration of reactive surface layers.

Page 26 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

27

Comparison with earlier combined PXRD and 31P SSNMR studies

Phosphate adsorption by LDH have previously been studied by 31P SSNMR,26-27, 29, 36, 38, 49-50, 67-68 but

detailed analysis (deconvolution) of the 31P MAS NMR spectra, which is necessary for quantification

and precise determination of δiso(31P) has only been reported in few studies.29, 36 Most studies report

only the 31P MAS NMR spectrum with distinct spectral features marked, which prevent a detailed

comparison with our data. The formation of phosphate minerals, e.g. newbeyrite,27, 49 bobierite,27, 67

aluminum phosphate,36 calcium phosphates29, 68 and iron phosphates,26 have been detected by 31P

SSNMR and PXRD after exposure of LDH to phosphate. Magnesium phosphate minerals were the

predominant phase identified from adsorption experiment with a starting pH of 7 and a final pH of

12 except for a broad resonance in the region 0-5 ppm, tentatively assigned to phosphate affiliated

with the MgAl-LDH.27 Subsequently, a broad 31P resonance at δ(31P) = 2.0 ppm was assigned to

phosphate strongly interacting with the commercial Mg3Al-LDH,67 but detailed analyses of the

spectra was not performed. The δiso(31P) is slightly lower than observed for M3A, but a fair agreement

is seen. Phosphate bound to a Mg2Al-LDH was reported at δiso(31P) = 3.2 ppm (major site) and -3.8

ppm. The first match well with site B and the latter is in between C and D.37 Hou at al performed

phosphate adsorption at pH = 9 and 12 for a Mg3Al-LDH. The maximum intensity was observed at

2.4 ppm (pH 9) and 3.7 ppm (pH 12), which was assigned to monohydrogen phosphate, whereas a

small shoulder near 6 ppm in the pH = 12 samples was assigned orthophosphate (PO43-). This was

also supported by PXRD and chemical analyses. We note that Hou et al. only considered intercalation

as the mechanism of phosphate adsorption.50 Guo et al. assigned the two dominating resonances at

δiso(31P) = -14.5 and -7.5 ppm in the 31P MAS NMR spectrum to phosphate associated with the LDH.49

However, δiso(31P) = -14.5 ppm are in the same range as aluminum phosphate synthesized under acid

condition36, 39, 60 and δiso(31P) = -7.5 ppm match newberyite.27, 61, 69 Both minerals are also clearly

identified by PXRD.27, 70 The phosphate adsorption experiments were performed at pH = 4.3, which

Page 27 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

28

explain the dissolution of the LDH and precipitation of phosphate minerals. For ultrathin (3-5 nm)

MgAl-LDH the same four resonance with slightly different δiso(31P) were observed with a lower

relative concentration of A (3-7 %) and a higher relative concentration of C (34-43 %) confirming

the assignments of the P-species.38

For a Zn2.5Al-LDH exposed to phosphate under neutral and acidic condition, the 31P MAS NMR

spectra showed four resonances and two of these with δiso(31P) = 4.2-5.2 ppm and δiso(31P) = 0.5-1.0

were assigned to phosphate in association with the LDH.36 The different δiso(31P) for the resonances

is ascribed to the presence of Zn instead of Mg. In the current study, which was performed under

alkaline condition, three sites (A, B, and C) have been assigned to phosphate associated with the

LDH. Hence, the number of P-species adsorbed by the LDH might depend on the pH, but further

studies are needed.

Conclusion:

Detailed insight into the interactions between phosphate and MgAl-LDH was obtained by a powerful

combination of phosphate adsorption studies along with multi-nuclear SSNMR, PXRD, FT-IR,

Raman, and ICP-OES for characterization of the LDH product. Four different P-species (A, B, C, and

D), were identified by analyses of the 31P MAS NMR spectra. These were assigned to phosphate

intercalated in the LDH (A), phosphate adsorbed on the surface of the LDH particles (B and C), and

phosphate adsorbed to an amorphous AOH (D) by combining information obtained from the different

characterization techniques above. The importance of amorphous AOH formation during LDH

synthesis has only recently been realized.33, 35 The relative concentration of the different P-species

depends on phosphate concentration, exposure time, the anion in the parent LDH, and the Mg:Al

ratio. Surprisingly, surface adsorption (B+C) is the major P adsorption mechanism (52-88 %),

whereas intercalation constitutes only up to 29(4) % of the total adsorbed phosphate. In addition, the

Page 28 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

29

amorphous AOH phase has a significant contribution to the phosphate adsorption capacity especially

at high P concentrations and long exposure times suggesting that LDH has a stronger affinity for

phosphate than AOH. Thus, the mechanism for the phosphate adsorption by MgAl-LDH prepared by

co-precipitation is a combination of phosphate intercalated in the interlayer (anion exchange), surface

adsorption, phosphate adsorbed to the amorphous AOH with surface adsorption to the LDH being the

dominating P-species. A lower crystallinity along the c-axis was seen after phosphate intercalation

indication of strong disorder of the HPO42- anions in the interlayer. In contrast, anion exchange of

carbonate with phosphate did not occur and the crystallinity is preserved along the c-axis. Moreover,

no significant dissolution of the MgAl-LDH is seen during phosphate adsorption. The high phosphate

adsorption capacities (17-47 mgP/g) make MgRAl LDH a promising phosphate adsorbent. However,

for application for wastewater treatment further studies regarding the optimum particle size and

morphology, as well as implication of phosphate adsorption to amorphous AOH on the recyclability

of LDH have to be investigated.

Supporting information:

Additional information about the phosphate adsorption studies (sample preparation, theoretical

adsorption capacities, adsorption isotherms, ICP-OES, FT-IR, in-situ PXRD) as well as SEM and

SSNMR (1H MAS NMR, 25Mg MAS NMR, 27Al 3QMAS NMR, 31P MAS NMR, and 31P-31P EXSY

NMR) data for selected samples is available in the Supporting Information.

Acknowledgements:

The authors are grateful for the financial support from The Innovation Found Denmark through the

project “Recovery of phosphorous from wastewater treatment system (RecoverP)” (grant 4106-

00014B), from the Villum Foundation through “Villum Young Investigator Programme” (grant

VKR022364; UGN, LL), and 600 MHz Spectrometer provided by Villum Center for Bioanalytical

Page 29 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

30

Services. The following are acknowledged for valuable experimental assistance: Elodie Petit

(Raman), Carina Lohman (ICP-OES), Dr. Vanessa Prevot and 2MATECH, Aubière, France (SEM).

Professor Poul Norby, Dr. Tae-Hyun Kim, Professor Hans Chr. Bruun Hansen, and Dr. Changyong

Lu are thanked for valuable discussions. We thank the anonymous reviewers for insightful comments,

which improved the manuscript.

Page 30 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

31

References:

1. Cordell, D.; Drangert, J.-O.; White, S., The Story of Phosphorus: Global Food Security and Food for Thought. Global Environ. Change 2009, 19, 292-305.

2. Commission, E., Report on Critical Raw Materials and the Circular Economy. Comm. Staff Work. Doc. 2018.

3. Tarayre, C.; De Clercq, L.; Charlier, R.; Michels, E.; Meers, E.; Camargo-Valero, M.; Delvigne, F., New Perspectives for the Design of Sustainable Bioprocesses for Phosphorus Recovery from Waste. Bioresource Tech. 2016, 206, 264-274.

4. Schoumans, O. F.; Bouraoui, F.; Kabbe, C.; Oenema, O.; van Dijk, K. C., Phosphorus Management in Europe in a Changing World. Ambio 2015, 44, 180-192.

5. Le Corre, K. S.; Valsami-Jones, E.; Hobbs, P.; Parsons, S. A., Phosphorus Recovery from Wastewater by Struvite Crystallization: A Review. Crit. Rev. Environ. Sci. Techno. 2009, 39, 433-477.

6. Ye, Y.; Ngo, H. H.; Guo, W.; Liu, Y.; Li, J.; Liu, Y.; Zhang, X.; Jia, H., Insight into Chemical Phosphate Recovery from Municipal Wastewater. Sci. Total Environ- 2017, 576, 159-171.

7. Melia, P. M.; Cundy, A. B.; Sohi, S. P.; Hooda, P. S.; Busquets, R., Trends in the Recovery of Phosphorus in Bioavailable Forms from Wastewater. Chemosphere 2017, 186, 381-395.

8. Goh, K.-H.; Lim, T.-T.; Dong, Z., Application of Layered Double Hydroxides for Removal of Oxyanions: A Review. Water Res. 2008, 42, 1343-1368.

9. Das, J.; Patra, B.; Baliarsingh, N.; Parida, K., Adsorption of Phosphate by Layered Double Hydroxides in Aqueous Solutions. Appl. Clay Sci. 2006, 32, 252-260.

10. Cheng, X.; Huang, X.; Wang, X.; Zhao, B.; Chen, A.; Sun, D., Phosphate Adsorption from Sewage Sludge Filtrate Using Zinc–Aluminum Layered Double Hydroxides. J. Hazard. Mater. 2009, 169, 958-964.

11. Yang, K.; Yan, L.-G.; Yang, Y.-M.; Yu, S.-J.; Shan, R.-R.; Yu, H.-Q.; Zhu, B.-C.; Du, B., Adsorptive Removal of Phosphate by Mg–Al and Zn–Al Layered Double Hydroxides: Kinetics, Isotherms and Mechanisms. Separation and Purification Techno. 2014, 124, 36-42.

12. Evans, D. G.; Slade, R. C., Structural Aspects of Layered Double Hydroxides. In Layered Double Hydroxides, Springer: 2006; pp 1-87.

13. Badreddine, M.; Legrouri, A.; Barroug, A.; De Roy, A.; Besse, J., Ion Exchange of Different Phosphate Ions into the Zinc–Aluminium–Chloride Layered Double Hydroxide. Mater. Lett. 1999, 38, 391-395.

14. Alibakhshi, E.; Ghasemi, E.; Mahdavian, M.; Ramezanzadeh, B.; Farashi, S., Fabrication and Characterization of PO4

3− Intercalated Zn-Al-Layered Double Hydroxide Nanocontainer. J. Electrochem. Soc. 2016, 163, C495-C505.

15. Legrouri, A.; Badreddine, M.; Barroug, A.; De Roy, A.; Besse, J.-P., Influence of pH on the Synthesis of the Zn–Al–Nitrate Layered Double Hydroxide and the Exchange of Nitrate by Phosphate Ions. J. Mater. Sci. lett. 1999, 18, 1077-1079.

16. Costantino, U.; Casciola, M.; Massinelli, L.; Nocchetti, M.; Vivani, R., Intercalation and Grafting of Hydrogen Phosphates and Phosphonates into Synthetic Hydrotalcites and Ac-Conductivity of the Compounds Thereby Obtained. Solid State Ionics 1997, 97, 203-212.

Page 31 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

32

17. Cheng, X.; Huang, X.; Wang, X.; Sun, D., Influence of Calcination on the Adsorptive Removal of Phosphate by Zn–Al Layered Double Hydroxides from Excess Sludge Liquor. J. Hazard. Mater. 2010, 177, 516-523.

18. He, H.; Kang, H.; Ma, S.; Bai, Y.; Yang, X., High Adsorption Selectivity of Znal Layered Double Hydroxides and the Calcined Materials toward Phosphate. J. Colloid Interface Sci. 2010, 343, 225-231.

19. Drenkova-Tuhtan, A.; Mandel, K.; Paulus, A.; Meyer, C.; Hutter, F.; Gellermann, C.; Sextl, G.; Franzreb, M.; Steinmetz, H., Phosphate Recovery from Wastewater Using Engineered Superparamagnetic Particles Modified with Layered Double Hydroxide Ion Exchangers. Water Res. 2013, 47, 5670-5677.

20. Chitrakar, R.; Tezuka, S.; Sonoda, A.; Sakane, K.; Ooi, K.; Hirotsu, T., Adsorption of Phosphate from Seawater on Calcined MgMn-Layered Double Hydroxides. J. Colloid Interface Sci. 2005, 290, 45-51.

21. Triantafyllidis, K. S.; Peleka, E. N.; Komvokis, V. G.; Mavros, P. P., Iron-Modified Hydrotalcite-Like Materials as Highly Efficient Phosphate Sorbents. J. Colloid Interface Sci. 2010, 342, 427-436.

22. Badreddine, M.; Khaldi, M.; Legrouri, A.; Barroug, A.; Chaouch, M.; De Roy, A.; Besse, J., Chloride-Hydrogenophosphate Ion Exchange into the Zinc–Aluminium–Chloride Layered Double Hydroxide. Mater. Chem. Phys. 1998, 52, 235-239.

23. Radha, A. V.; Vishnu Kamath, P.; Shivakumara, C., Mechanism of the Anion Exchange Reactions of the Layered Double Hydroxides (LDHs) of Ca and Mg with Al. Solid State Sci. 2005, 7, 1180-1187.

24. Ashekuzzaman, S.; Jiang, J.-Q., Study on the Sorption–Desorption–Regeneration Performance of Ca-, Mg-and CaMg-Based Layered Double Hydroxides for Removing Phosphate from Water. Chem. Eng. J. 2014, 246, 97-105.

25. Zhou, J.; Yang, S.; Yu, J.; Shu, Z., Novel Hollow Microspheres of Hierarchical Zinc–Aluminum Layered Double Hydroxides and Their Enhanced Adsorption Capacity for Phosphate in Water. J. Hazard. Mater. 2011, 192, 1114-1121.

26. Du, Y.; Rees, N.; O'Hare, D., A Study of Phosphate Absorption by Magnesium Iron Hydroxycarbonate. Dalton Trans. 2009, 8197-8202.

27. Bernardo, M. P.; Moreira, F. K.; Colnago, L. A.; Ribeiro, C., Physico-Chemical Assessment of [Mg-Al-PO4]-LDHs Obtained by Structural Reconstruction in High Concentration of Phosphate. Colloids Surfaces A, 2016, 497, 53-62.

28. Seftel, E.; Ciocarlan, R.; Michielsen, B.; Meynen, V.; Mullens, S.; Cool, P., Insights into Phosphate Adsorption Behavior on Structurally Modified ZnAl Layered Double Hydroxides. Appl. Clay Sci. 2018, 165, 234-246.

29. Bekele, B.; Lundehøj, L.; Jensen, N. D.; Nielsen, U. G.; Forano, C., Sequestration of Orthophosphate by Ca2Al-NO3 Layered Double Hydroxide – Insight into Reactivity and Mechanism. Appl. Clay Sci. 2019, 176, 49-57.

30. Xu, Y.; Dai, Y.; Zhou, J.; Xu, Z. P.; Qian, G.; Lu, G. M., Removal Efficiency of Arsenate and Phosphate from Aqueous Solution Using Layered Double Hydroxide Materials: Intercalation Vs. Precipitation. J. Mater. Chem. 2010, 20, 4684-4691.

31. Baliarsingh, N.; Parida, K.; Pradhan, G., Influence of the Nature and Concentration of Precursor Metal Ions in the Brucite Layer of LDHs for Phosphate Adsorption–a Review. RSC Adv. 2013, 3, 23865-23878.

32. Kuzawa, K.; Jung, Y.-J.; Kiso, Y.; Yamada, T.; Nagai, M.; Lee, T.-G., Phosphate Removal and Recovery with a Synthetic Hydrotalcite as an Adsorbent. Chemosphere 2006, 62, 45-52.

Page 32 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

33

33. Pushparaj, S. S. C.; Forano, C.; Prevot, V.; Lipton, A. S.; Rees, G. J.; Hanna, J. V.; Nielsen, U. G., How the Method of Synthesis Governs the Local and Global Structure of Zinc Aluminum Layered Double Hydroxides. J. Phys. Chem. C 2015, 119, 27695-27707.

34. Costantino, U.; Marmottini, F.; Nocchetti, M.; Vivani, R., New Synthetic Routes to Hydrotalcite‐Like Compounds− Characterisation and Properties of the Obtained Materials. Eur. J. Inorg. Chem. 1998, 1998, 1439-1446.

35. Staal, L. B.; Pushparaj, S. S. C.; Forano, C.; Prevot, V.; Ravnsbæk, D. B.; Bjerring, M.; Nielsen, U. G., Competitive Reactions During Synthesis of Zinc Aluminum Layered Double Hydroxides by Thermal Hydrolysis of Urea. J. Mater. Chem. A 2017, 5, 21795-21806.

36. Lundehøj, L.; Jensen, H. C.; Wybrandt, L.; Nielsen, U. G.; Christensen, M. L.; Quist-Jensen, C. A., Layered Double Hydroxides for Phosphorus Recovery from Acidified and Non-Acidified Dewatered Sludge. Water Res. 2019, 153, 208-216.

37. Borges, R.; Wypych, F.; Petit, E.; Forano, C.; Prevot, V., Potential Sustainable Slow-Release Fertilizers Obtained by Mechanochemical Activation of MgAl and MgFe Layered Double Hydroxides and K2HPO4. Nanomater. 2019, 9, 183.

38. Liu, C.; Zhang, M.; Pan, G.; Lundehøj, L.; Nielsen, U. G.; Shi, Y.; Hansen, H. C. B., Phosphate Capture by Ultrathin Mgal Layered Double Hydroxide Nanoparticles. Appl. Clay Sci. 2019, 177, 82-90.

39. Lookman, R.; Grobet, P.; Merckx, R.; Van Riemsdijk, W. H., Application of 31P and 27Al MAS NMR for Phosphate Speciation Studies in Soil and Aluminium Hydroxides: Promises and Constraints. Geoderma 1997, 80, 369-388.

40. Li, W.; Feng, X.; Yan, Y.; Sparks, D. L.; Phillips, B. L., Solid-State Nmr Spectroscopic Study of Phosphate Sorption Mechanisms on Aluminum (Hydr)Oxides. Environ. Sci. Technol. 2013, 47, 8308-15.

41. Li, W.; Feng, J.; Kwon, K. D.; Kubicki, J. D.; Phillips, B. L., Surface Speciation of Phosphate on Boehmite (-AlOOH) Determined from NMR Spectroscopy. Langmuir 2010, 26, 4753-4761.

42. Jensen, N. D.; Bjerring, M.; Nielsen, U. G., A Solid State Nmr Study of Layered Double Hydroxides Intercalated with Para-Amino Salicylate, a Tuberculosis Drug. Solid State Nucl. Magn. Reson. 2016, 78, 9-15.

43. Sideris, P. J.; Nielsen, U. G.; Gan, Z.; Grey, C. P., Mg/Al Ordering in Layered Double Hydroxides Revealed by Multinuclear NMR Spectroscopy. Sciene 2008, 321, 113-117.

44. Sideris, P. J.; Blanc, F.; Gan, Z.; Grey, C. P., Identification of Cation Clustering in Mg–Al Layered Double Hydroxides Using Multinuclear Solid State Nuclear Magnetic Resonance Spectroscopy. Chem. Mater. 2012, 24, 2449-2461.

45. Cadars, S.; Layrac, G.; Gérardin, C.; Deschamps, M.; Yates, J. R.; Tichit, D.; Massiot, D., Identification and Quantification of Defects in the Cation Ordering in Mg/Al Layered Double Hydroxides. Chem. Mater. 2011, 23, 2821-2831.

46. Pushparaj, S. S. C.; Jensen, N. D.; Forano, C.; Rees, G. J.; Prevot, V.; Hanna, J. V.; Ravnsbæk, D. B.; Bjerring, M.; Nielsen, U. G., Structural Investigation of Zn(II) Insertion in Bayerite, an Aluminum Hydroxide. Inorg. Chem. 2016, 55, 9306-9315.

47. Sene, S.; Bégu, S. Gervais, C.; Renaudin, G; Mesbah A.; Smith, M.E.; Mutin, P.H.; van der Lee, A.; Nedelec, J.-M.; Bonhomme, C.; Laurencin, D.;Intercalation of Benzoxaborolate Anions in Layered Double Hydroxides: Toward Hybrid Formulations for Benzoxaborole Drugs. Chem. Mater. 2015, 27, 1242-1254.

Page 33 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

34

48. Jensen, N. D.; Forano, C.; Pushparaj, S. S. C.; Nishiyama, Y.; Bekele, B.; Nielsen, U. G., The Distribution of Reactive Ni2+ in 2D Mg2−XNixAl-LDH Nanohybrid Materials Determined by Solid State 27Al MAS NMR Spectroscopy. Phys. Chem. Chem. Phys. 2018, 20, 25335-25342.

49. Guo, Y.; Wang, J.; Li, D.; Tang, P.; Leroux, F.; Feng, Y., Micrometer-Sized Dihydrogenphosphate-Intercalated Layered Double Hydroxides: Synthesis, Selective Infrared Absorption Properties, and Applications as Agricultural Films. Dalton Trans. 2018.

50. Hou, X.; Bish, D. L.; Wang, S.-L.; Johnston, C. T.; Kirkpatrick, R. J., Hydration, Expansion, Structure, and Dynamics of Layered Double Hydroxides. Am. Miner. 2003, 88, 167-179.

51. Fiske, P. S.; Stebbins, J. F.; Farnan, I., Bonding and Dynamical Phenomena in Mgo: A High Temperature 17O and 25Mg NMR Study. Phys. Chem. Miner. 1994, 20, 587-593.

52. Kemp, T. F.; Smith, M. E., Quadfit—a New Cross-Platform Computer Program for Simulation of NMR Line Shapes from Solids with Distributions of Interaction Parameters. Solid State Nucl. Magn. Reson. 2009, 35, 243-252.

53. Forano, C.; Costantino, U.; Prévot, V.; Gueho, C. T., Layered Double Hydroxides (LDH). In Developments in Clay Science, Elsevier: 2013; 5, 745-782.

54. Giles, C. H.; Smith, D.; Huitson, A., A General Treatment and Classification of the Solute Adsorption Isotherm. I. Theoretical. J. Colloid Interface Sci. 1974, 47, 755-765.

55. Everaert, M.; Warrinnier, R.; Baken, S.; Gustafsson, J.-P.; De Vos, D.; Smolders, E., Phosphate-Exchanged Mg–Al Layered Double Hydroxides: A New Slow Release Phosphate Fertilizer. ACS Sustainable Chem. Eng. 2016, 4, 4280-4287.

56. Di Bitetto, A.; André, E.; Carteret, C. D.; Durand, P.; Kervern, G., Probing the Dynamics of Layered Double Hydroxides by Solid-State 27Al NMR Spectroscopy. J. Phys. Chem. C 2017, 121, 7276-7281.

57. Everaert, M.; Slenders, K.; Dox, K.; Smolders, S.; De Vos, D.; Smolders, E., The Isotopic Exchangeability of Phosphate in Mg-Al Layered Double Hydroxides. J. Colloid Interface Sci. 2018, 520, 25-32.

58. Yun, S. K.; Pinnavaia, T. J., Water Content and Particle Texture of Synthetic Hydrotalcite-Like Layered Double Hydroxides. Chem. Mater. 1995, 7, 348-354.

59. Rudolph, W. W., Raman-and Infrared-Spectroscopic Investigations of Dilute Aqueous Phosphoric Acid Solutions. Dalton Trans. 2010, 39, 9642-9653.

60. Wang, X.; Phillips, B. L.; Boily, J.-F.; Hu, Y.; Hu, Z.; Yang, P.; Feng, X.; Xu, W.; Zhu, M., Phosphate Sorption Speciation and Precipitation Mechanisms on Amorphous Aluminum Hydroxide. Soil Systems 2019, 3, 20.

61. Scrimgeour, S. N.; Chudek, J. A.; Lloyd, C. H., The Determination of Phosphorus Containing Compounds in Dental Casting Investment Products by 31P Solid-State Mas-NMR Spectroscopy. Dental Mater. 2007, 23, 415-424.

62. Aramendı́a, M. a. A.; Borau, V.; Jiménez, C.; Marinas, J. M. a.; Romero, F. J.; Ruiz, J. R., Characterization by XRD, DRIFT, and MAS NMR Spectroscopies of a Mg2P2O7catalyst. J. Colloid Interface Sci. 1998, 202, 456-461.

63. Aramendı́a, M. A.; Borau, V.; Jiménez, C.; Marinas, J. M.; Romero, F. J.; Ruiz, J. R., XED and Solid-State NMR Study of Magnesium Oxide–Magnesium Orthophosphate Systems. J. Solid State Chem. 1998, 135, 96-102.

64. Jarlbring, M.; Sandström, D. E.; Antzutkin, O. N.; Forsling, W., Characterization of Active Phosphorus Surface Sites at Synthetic Carbonate-Free Fluorapatite Using Single-Pulse 1H, 31P, and 31P CP MAS NMR. Langmuir 2006, 22, 4787-4792.

Page 34 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

35

65. Han, S.; Hou, W.; Zhang, C.; Sun, D.; Huang, X.; Wang, G., Structure and the Point of Zero Charge of Magnesium Aluminium Hydroxide. Faraday Trans. 1998, 94, 915-918.

66. Halajnia, A.; Oustan, S.; Najafi, N.; Khataee, A.; Lakzian, A., The Adsorption Characteristics of Nitrate on Mg–Fe and Mg–Al Layered Double Hydroxides in a Simulated Soil Solution. Appl. Clay Sci. 2012, 70, 28-36.

67. Bernardo, M. P.; Guimarães, G. G.; Majaron, V. F.; Ribeiro, C., Controlled Release of Phosphate from Layered Double Hydroxide Structures: Dynamics in Soil and Application as Smart Fertilizer. ACS Sustainable Chem. Eng. 2018, 6, 5152-5161.

68. Bernardo, M. P.; Moreira, F. K.; Ribeiro, C., Synthesis and Characterization of Eco-Friendly Ca-Al-LDH Loaded with Phosphate for Agricultural Applications. Appl. Clay Sci. 2017, 137, 143-150.

69. Ma, N.; Rouff, A. A.; Phillips, B. L., A 31P NMR and TG/DSC-FT-IR Investigation of the Influence of Initial pH on Phosphorus Recovery as Struvite. ACS Sustainable Chem. Eng. 2014, 2, 816-822.

70. Palacios, E.; Leret, P.; De La Mata, M. J.; Fernández, J. F.; De Aza, A. H.; Rodríguez, M. A., Influence of the pH and Ageing Time on the Acid Aluminum Phosphate Synthesized by Precipitation. Cryst. Eng. Commun. 2013, 15, 3359.

Page 35 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

36

TOC graphics

Page 36 of 36



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

atomic level understanding of orthophosphate adsorption by

Documents