Subscriber access provided by MCGILL UNIV

Ethylenediamine-enabled sustainable synthesis of

mesoporous nanostructured Li2FeIISiO4 particles from
Fe(III) aqueous solution for Li-ion battery application
Huijing Wei, Xia Lu, Hsien-Chieh Chiu, Bin Wei, Raynald Gauvin, Zachary Arthur,
Vincent Emond, De-Tong Jiang, Karim Zaghib, and George P. Demopoulos
ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/
acssuschemeng.8b00090 • Publication Date (Web): 22 Apr 2018
Downloaded from http://pubs.acs.org on April 28, 2018

Just Accepted

“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital 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. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W.,

Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 25 ACS Sustainable Chemistry & Engineering

1
2
3
4
5
6
7
Ethylenediamine-enabled sustainable synthesis of
8
9 mesoporous nanostructured Li2FeIISiO4 particles from
10
11
12 Fe(III) aqueous solution for Li-ion battery application
13
14 Huijing Wei1, Xia Lu1, 2, Hsien-Chieh Chiu1, Bin Wei1, Raynald Gauvin1, Zachary Arthur3, Vincent Emond3, De-
15
16 Tong Jiang3, Karim Zaghib4, and George P. Demopoulos1,*
17
18
19 1. Materials Engineering, McGill University, 3610 University Street, Montréal, Québec H3A 0C5, Canada.
20 2. State Key Laboratory of Organic-Inorganic Composites, College of Energy, Beijing University of Chemical
21
Technology, Beijing 100029, China.
22
23 3. Department of Physics, University of Guelph, Guelph, Ontario N1G 2W1, Canada.
24 4. Centre d’excellence-ETSE, Hydro-Québec, Varennes, Québec J3X 1S1, Canada.
25
26 *Corresponding author: Prof. G. P. Demopoulos, Email: george.demopoulos@mcgill.ca
27
28 Abstract: Engineering of nanostructured lithium iron silicate (LFS) particles is pursued via a
29
30 novel benign synthesis approach seeking to understand the crystalline particle formation process
31
32
and its impact on energy storage capacity. Specifically, mesoporous Li2FeSiO4 nanostructured
33 particles are synthesized via a novel dual-step process involving organic-assisted hydrothermal
34
35 precipitation from concentrated Fe(III) (1 mol/L) aqueous solution followed by reductive (5 vol.%
36
37 H2) thermal transformation of the precipitate at 400°C (LFS400) and 700°C (LFS700). Scanning
38 and transmission electron microscopy revealed the formation of secondary submicron sized
39
40 porous agglomerates of unitary primary nanocrystals (~ 50nm for LFS400 and ~200nm for
41
42 LFS700). Both ethylene glycol and ethylenediamine are used as crystallization control additives.
43
44 It is demonstrated that formation of LFS from Fe(III) precursor is made possible only by the
45 action of ethylenediamine. The obtained LFS particles are found to be predominantly
46
47 monoclinic as per X-ray diffraction and Rietveld refinement and bear an in situ formed N-doped
48
49 carbon coating layer as characterized by X-ray photoelectron spectroscopy. TEM coupled with
50
selected area electron diffraction (SAED) analysis confirmed the Rietveld Refined XRD phase
51
52 compositions. The reductive annealing-induced phase transformation sequence leading to LFS
53
54 crystallization is characterized and the enabling role of ethylenediamine is discussed. Initial
55
56 galvanostatic charging-discharging and cyclic voltammetry measurements indicate the annealing
57
58
59 1
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 2 of 25

1
2
3 temperature of LFS formation to influence the Li-ion storage profile as it shifts from two-phase
4
5 reaction in LFS700 to solid solution in LFS400-this being attributed to nanostructural changes.
6
7
Keywords: Li2FeSiO4, Hydrothermal synthesis, Reductive annealing, Crystallization control,
8
9 Ethylenediamine, Li-ion batteries
10
11
12 Introduction
13
14 Rechargeable lithium-ion batteries (LIBs) have powered over the past two decades the
15
16 tremendous proliferation of portable electronics and now have started playing a key role in
17
powering a shift towards electric vehicle transportation.1-3 This shift imposes new LIB
18
19 requirements, namely high energy density (for longer driving range), good rate performance (for
20
21 higher power), long cycle life, but also high safety factor, and low cost scalable and sustainable
22
23 synthesis.3-4 Towards this end, lithium metal orthosilicates are of great interest because of their
24 two lithium-ion storage per formula unit (Li2FeSiO4) that could provide a theoretical capacity
25
26 that is double (330 mAh/g) of that of LiFePO4 (LFP),5-6 corresponding to a specific energy of
27
28 1,120 Wh/kg surpassing all current cathode materials.7 Although Li2FeSiO4 has similar
29
polyoxyanion framework as that of LFP,8 the existence of different polymorphs of lithium metal
30
31 silicates have hampered their development and performance as cathode materials in Li-ion
32
33 batteries.6 For example, in the case of Li2FeSiO4 (LFS) the different polymorphs are
34
35 predominantly associated with two main crystal structures, monoclinic and orthorhombic, which
36 due to very close formation energies tend to coexist,9-10 not to mention their significant degree of
37
38 crystal disorder and defects by which they are plagued from.11 As such their synthesis and
39
40 subsequent phase-dependent evaluation of their electrochemical performance is highly
41
42
complicated.10, 12-15 For making progress with the development of these potentially high energy
43 density Li-ion storage materials thus it becomes important to understand the underlying
44
45 crystallization mechanism of LFS phase formation and ultimately control via the advancement of
46
47 new synthesis routes. Equally important in this pursuit is that the new synthesis routes are
48
sustainable in nature to allow for ultimate scalable clean production.
49
50
51 Over the last 10 years, different methods have been applied to synthesis of lithium iron
52
silicates among which notable are solid state synthesis14, 16 and several solution methods.8, 13, 17-18
53
54 Conventional solid-state synthesis,16 although simple, tends to produce rather non-homogeneous
55
56 micrometer-sized particles,7 a problem partly solved by mixing the precursor chemicals by high-
57
58
59 2
60 ACS Paragon Plus Environment
Page 3 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 energy planetary ball milling prior to calcination.14 Solution synthesis (or “soft chemistry”)
4
5 methods are better suited to achieve the necessary nanocrystal control.7 It is important however
6
7 as argued by Zaghib et al.5 to ensure the feasibility of the synthesis method in producing
8
9
“…many tons of these nanostructured materials to prepare batteries for electric vehicles at a
10 reasonable price”. In this context complex solution synthesis routes like supercritical sol-gel
11
12 reaction producing nanosheet crystals19 may not offer a practical option. Alternatively,
13
14 hydrothermal synthesis5, 8 as discussed by Toyota’s J. Chen20 offers a truly sustainable process
15
option. Use of functional organic additives is a powerful way to regulate the nucleation,
16
17 aggregation and growth of the hydrothermally formed particles.21 Furthermore combining
18
19 solution precipitation with thermal transformation/annealing offers flexibility in producing
20
21 different phases,13 and eliminating Li-Fe anti-site defects.8 Also, because of the very low
22 conductivity of lithium metal silicate crystals, carbon coating is commonly pursued via the use of
23
24 organic chemicals during the solution synthesis22 or during post heat-treatment.23. The properties
25
26 of resultant particles, such as crystal structure, size, habit, purity, and porosity are greatly
27
influenced by the path of crystallization. As a consequence, the method of synthesis–particle
28
29 formation, can directly impact the electrochemical performance, i.e. intercalation properties of
30
31 resultant Li2FeSiO4 cathode material. To date, most of the previous solution synthesis studies
32
33 have made use of ferrous (Fe(II)) salts as the iron precursor followed by high-temperature
34 annealing in an inert atmosphere. The use of Fe(II) as precursor, necessitates oxygen-free
35
36 handling that greatly complicates the synthesis process. Developing a method starting with ferric
37
38 (Fe(III)) salts as precursor–ferric salts are significantly less expensive than their ferrous
39
40
counterparts, can greatly simplify the process rendering it more amenable to large scale
41 production. To our knowledge no solution based method for the synthesis of LFS from a ferric
42
43 salt as precursor has been reported other than some rather complex processing routes involving
44
45 the use of tri-block copolymers as in situ template that are not suitable for large scale synthesis.22,
46 24
47
48
49 It is the scope of the present work to report on a novel ferric salt-based synthesis method for
50
lithium iron silicate involving ethylenediamine-controlled hydrothermal precipitation of an
51
52 intermediate, which upon annealing in a reducing atmosphere converts to mesoporous C-coated
53
54 nanostructured LFS particles suitable for fabrication of Li-ion battery cathodes.
55
56
57
58
59 3
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 4 of 25

1
2
3
4
5
6 Experimental
7
8 Hydrothermal precipitation: The synthesis method involved hydrothermal preparation of
9
10 an intermediate precipitate in the presence of ethylene glycol and ethylenediamine that was
11
12
subsequently transformed thermally to LFS via reductive annealing. The method was adapted
13 from the work of Goodenough’s group on the synthesis of LiFePO4 microspheres.25 The
14
15 precursors were initially added in deionized water (194 mL) in stoichiometric amounts, typically
16
17 in the following concentrations: 0.1936 mol of Fe(NO3)3.9H2O (Sigma-Aldrich 99.95%), 0.4065
18
mol of CH3COOLi (Sigma-Aldrich: 99.95%), and 0.1935 mol of fumed silica (Sigma-Aldrich,
19
20 0.2-0.3 µm average particle size). After the solution was well homogenized the two organic
21
22 reagents were added, namely, 10.4 mL (0.1865 mol) of ethylene glycol (EG, Sigma-Aldrich,
23
24 99.8% anhydrous) and 10.4 mL (0.1557 mol) of ethylenediamine (EN, Sigma-Aldrich 99.5%).
25 Upon their addition and thorough mixing in ultrasonic bath, a brownish suspension with pH of 8
26
27 – 9 was obtained, which was then transferred to a Teflon-lined stainless-steel autoclave (Parr
28
29 4567) stirred at 300 rpm. The detailed precursor solution preparation procedure is given in the
30
31
Supplementary Information. Hydrothermal reaction was carried out at 180°C for 12 hours. The
32 resultant slurry was dried in air (i.e. water got evaporated) on an 80°C stirring hotplate to obtain
33
34 an amorphous/poorly crystalline intermediate product that was subsequently used in the
35
36 annealing experiments.
37
38 Reductive Annealing: The annealing step was carried out in a tube furnace (Carbolite GHA)
39
40 under a constant flow of reducing Ar/H2 gas mixture (95:5). The initial step involved purging
41
42
with the reducing gas for one hour at room temperature. Then the temperature was raised to
43 200°C at a rate of 3°C/min, where it was kept for 3 hours before the temperature was increased
44
45 again at 3°C/min to a target temperature of 400°C or 700°C (namely, LFS400 and LFS700) and
46
47 kept for a pre-selected amount of time (typically 6 hours) followed by natural cooling process to
48
room temperature.
49
50
51 Characterization: The characterization analyses included: X-ray diffraction (Bruker D8
52
53 Discover) with Co Kα source of radiation (λ ~ 1.78901 Å). Additionally, some XRD spectra were
54 collected with a Philips diffractometer with Cu source and some were collected using
55
56 synchrotron hard X-rays at Canadian Light Source Ltd. (λ ~ 0.6886 Å). The XRD spectra were
57
58
59 4
60 ACS Paragon Plus Environment
Page 5 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 analyzed using Rietveld refinement by TOPAS Total Pattern Analysis for determination of the
4
5 phase composition (i.e. the weight fraction of monoclinic vs. orthorhombic phase). Other
6
7 characterization methods included scanning electron microscopy (Hitachi SU-8230) and
8
9
transmission electron microscopy (FEI F20); X-ray photoelectron spectroscopy (K-Alpha XPS,
10 Thermal Fisher Scientific Inc.) and Raman spectroscopy (Bruker Senterra dispersive Raman
11
12 microscope) for surface characterizations; thermogravimetric analysis for determining the carbon
13
14 weight content (TA500 thermogravimetric analyzer) in a mixture of N2/O2 (60:40) atmosphere;
15
inductively coupled plasma optical emission spectroscopy (ICP-OES, Thermo Scientific iCAP
16
17 6500 ICP Spectrometer) for solution analysis of the acid-digested powders; and Fe2+/Fe3+ redox
18
19 titration.
20
21 Electrode preparation and cell testing: The electrode paste was prepared by mixing the
22
23 active material (LFS), carbon black and binder polyvinylidene fluoride (PVDF) in a typical ratio
24
25 of 8:1:1, using N-Methyl-2-pyrrolidone (NMP) as a solvent. The pristine LFS400 nanomaterial
26
was used as-is for making the paste. By contrast the LFS700 material because of its micron-sized
27
28 particles was firstly ball milled for three hours at 250 RPM in a high energy planetary ball
29
30 milling machine (Fritsch, Pulverisette 7) in the presence of 10 wt.% of carbon black, using ZrO2
31
32 beads as grinding media and isopropanol as a solvent. The ball milled LFS700 material was
33 obtained after filtration and drying under vacuum (namely, LFS700BM). After mixing the LFS
34
35 with carbon black and PVDF, the slurry paste was then cast on aluminum foil and dried in
36
37 vacuum oven overnight before being punched for further drying and coin cell assembly. In the
38
39
half cells, lithium metal was used as the anode, LiPF6 in EC/DMC (1:1) was used as the
40 electrolyte, and polypropylene film (Celgard2200) was used as the separator between two
41
42 electrodes. The coin cells were then tested using the Arbin multi-channel cycler for cycling and
43
44 Bio-Logic VSP potentiostat for cyclic voltammetry (CV). Charging-discharging of the two
45 electrodes was done at a rate of C/10 (1C=166 mAh/g) and CV at a scan rate of 0.04 mV/s over
46
47 the first three cycles between 1.50 to 4.67 V. Electrochemical testing was done at elevated
48
49 temperature 45 to 55°C instead of room temperature because of LFS’ poor intrinsic electronic
50
51 (6x10-14 S/cm)8 and ionic26 conductivities.
52
53 Results and Discussion
54
55 LFS crystallization sequence
56
57
58
59 5
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 6 of 25

1
2
3 Hydrothermal precipitation: We investigated the reactions occurring during the
4
5 hydrothermal step by analyzing the formed precipitate and final solution. It was found that
6
7 essentially all of Si (> 99.3%) and Fe (100%) had precipitated during hydrothermal reaction step
8
9
but only about 25% of the Li had precipitated. The remaining 75% of the Li staying in solution
10 (Table S1 in Supplementary Information) precipitated as salt during subsequent solution
11
12 evaporation/drying treatment. The solution pH remained essentially the same (pH 8-9) before
13
14 and after the hydrothermal treatment. Titration analysis of the acid-digested precipitate, done in
15
glovebox to ensure no re-oxidation of any Fe2+ present (refer to Supporting Information), showed
16
17 negligible signs of reduction from Fe3+ to Fe2+ (< 0.2%). This finding differs from earlier work25
18
19 involving the lithium iron phosphate system, where it was reported Fe3+ to have been reduced to
20
21 Fe2+ during hydrothermal treatment-the reduction been attributed to the action of ethylene glycol
22 (EG) but with no direct analysis been provided. It is quite likely this difference between the two
23
24 systems to reflect stronger Fe-O bond strength in silicate vs. phosphate anion. After the
25
26 evaporative drying step at 80°C, the solids were subjected to XRD analysis. The XRD pattern
27
(Figure S1a in Supplementary Information) shows a rather amorphous material with no sharp
28
29 diffraction peaks but only two broad peaks at around ~ 30° and ~ 63°, respectively. These broad
30
31 peaks seem to overlap with the ferric oxy-hydroxide known as 2-line ferrihydrite
32
33 (FeOOH·H2O);27 in this case though at all likelihood they correspond to amorphous hydrated
34 III
35 iron (III) silicate, Fe (Si O )(OH) ·2H O, known as hisingerite that also exhibits two similar
2 2 5 4 2
36
28
37 broad peaks. As for the 25% of Li that co-precipitated with the amorphous iron(III)-silica
38
39 precipitate, it seemed to be in an adsorbed state favored by the moderately alkaline pH
40
41
environment. Following the evaporation of the solution (drying at 80°C) the remaining soluble
42 fraction (75 %) of Li crystallized out as lithium acetate into the hydrous ferric silicate matrix
43
44 (hisingerite) that serves as LFS precursor during subsequent thermal reductive annealing. In
45
46 other words no direct crystallization of Li2FeSiO4 had occurred during the hydrothermal
47
processing step when ferric iron was used as the source of iron, in contrast to the case of
48
49 hydrothermal synthesis of ferrous iron-derived LFS.13, 17, 29 It was also found from FTIR analysis
50
51 (Figure S2 in Supplementary Information) that as indicated by the presence of characteristic C-
52
53 O, C-H, N-H and O-H bending/stretching modes, the organic additives EN and EG are integral
54 building components of the LFS intermediate via their known action as bidentate complexing
55
56 ligand30 and surface active agent31 respectively.
57
58
59 6
60 ACS Paragon Plus Environment
Page 7 of 25 ACS Sustainable Chemistry & Engineering

1
2
3
4
5 Reductive annealing-nucleation and transformation: After water evaporation, the organic
6
7 additive-assembled hydrous ferric silicate intermediate was subjected to reactive thermal
8
9
transformation/annealing in a 5% H2 reducing atmosphere at increasing temperature and time to
10 determine its crystallization pathway. First XRD characterization after 3 hours holding time at
11
12 200°C was done (Supplementary Information of Figure S1b). As it can be seen, the pattern
13
14 remained predominantly amorphous, but this time weak peaks seemed to start emerging in the
15
region 24° to 37°, where key peaks of crystalline LFS reside. This can be interpreted as evidence
16
17 of the early stage organization of Li-FeII-SiO4 constituents into nuclei of LFS by the reducing
18
19 action (FeIII
FeII) of H2. TEM analysis (image shown in Figure S3a in Supplementary
20
21 Information) revealed indeed the onset of LFS nucleation, as judged from some barely-seen
22 broken diffraction rings at the centre of the SAED pattern (shown in Figure S3b in
23
24 Supplementary Information).
25
26 Following this early annealing step at 200°C, the temperature was further increased to
27
400°C where it was held for different times, 1 min, 1 hr, 3 hrs and 10 hrs. The XRD patterns of
28
29 Li2FeSiO4 annealed at different times at 400°C were shown in Figure 1. By examining the XRD
30
31 pattern of the material annealed at 400°C for 1 minute, it can be seen that crystalline Li2FeSiO4
32
33 had already formed out of the nucleation cluster observed at 200°C (Supplementary Information
34 of Fig. S1b). Comparing the XRD patterns at different annealing times it is seen that the full
35
36 width half maximum (FWHM) decreases as the annealing time increases (Fig 1b), which is an
37
38 indication of improved crystallinity as a result of diffusion-dependent crystal growth and
39
40
refinement (crystal ordering). It is also noted that the main crystal phase did not change with
41 increasing annealing time. The monoclinic (112) peak was observed for all samples, as labeled in
42
43 Figure 1, although due to nanocrystalline nature some small peak features were not fully
44
45 resolved. By examining the TEM images and selected area diffraction patterns (Figure S4-S7 in
46
Supplementary Information), the predominant presence of monoclinic crystalline LFS phase was
47
48 observed in all samples annealed from 1 minute up to 10 hours at 400°C. As annealing time
49
50 increases the small nuclei grow and bond into each other to form polycrystalline particles. A thin
51
52 amorphous layer with about less than 3-5 nanometer thickness was observed in all samples
53 (Figure S4-S7 in Supplementary Information), which most likely represents the in situ formed
54
55 carbon layer from the decomposition of the organics as further discussed in the next section.
56
57
58
59 7
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 8 of 25

1
2
3 From this first XRD analysis the co-presence of the LFS orthorhombic phase cannot be ruled out.
4
5 It is also worth noticing that small impurity peaks (possibly Fe3O4 and Li2SiO3, labelled with
6
7 stars) were present at shorter annealing time that disappeared as the annealing time increased to
8
9
10 hours, indicating the formation of impurity-free LFS nanocrystals.
10
11 (a) (b)
12
13
(112)

(112)
14
15
16
17 *
18
19 * *
20
*
21
22
23
24
25 Fig. 1. (a) XRD patterns (synchrotron source, λ=0.6886 Ǻ) for Li2FeSiO4 samples annealed at 400°C for 1 min, 1
26
hour, 3 hours and 10 hours. (b) A zoomed in view of the XRD patterns in (a) focusing on the range from 10-16°. *
27
28 indicates presence of minor impurities.
29
30
31
32 It must be underlined that during annealing not only a physical transformation of the
33
34 amorphous precipitate into LFS crystals occurs but also co-current chemical reduction of ferric
35
iron to ferrous iron by hydrogen. To monitor this reaction, LFS samples at different annealing
36
37 times were digested and analyzed by titration to determine the percentage content of ferric to
38
39 ferrous reduction. Thus it was found after the 1 min annealing point at 400°C only 60.2% of Fe
40
41 was reduced to ferrous state. After 3 hrs, the degree of reduction was 69%, after 6 hrs was 97%,
42 with no apparent further increase after 10 hrs. Therefore, 6 hours was selected as the standard
43
44 annealing time.
45
46
Crystalline properties of LFS400 and LFS700 particles
47
48
49 Phase composition: LFS has been reported to exhibit different crystal phase polymorphs
50
depending on the temperature of synthesis such as low-temperature orthorhombic (Pmn21 or βII)
51
52 obtained at 200°C, high-temperature monoclinic (P21/n or γs) obtained at 600-800°C, and high-
53
54 temperature orthorhombic (Pmnb or γII) obtained at 900°C.6, 13 Previous studies have mostly
55
56 focused on the electrochemistry of the monoclinic (P21/n) LFS synthesized by annealing of
57
58
59 8
60 ACS Paragon Plus Environment
Page 9 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 ferrous-derived precursor phases at around 700°C.14, 16, 32-33 Here we prepared LFS samples at
4
5 both 400°C and 700°C to evaluate the impact of the ferric-derived precursor and annealing
6
7 temperature on crystal phase properties and ultimately electrochemical response. The XRD
8
9
spectra (Figure 2) of both 400°C and 700°C (LFS400 and LFS700) samples (annealed for 6
10 hours) were fitted using Rietveld refinement to quantify the LFS phases present. The
11
12 predominant phase in both annealed samples was determined to be the high temperature
13
14 monoclinic (P21/n) phase, but to co-exist with a small fraction of orthorhombic LFS of different
15
space group. Thus the LFS400 material comprised 10% low temperature orthorhombic (Pmn21),
16
17 while the LFS700 one comprised 25% high temperature orthorhombic (Pmnb) with the bulk in
18
19 both cases being the monoclinic (P21/n, 75%) phase. These results show a gradual phase
20
21 transition with increasing annealing temperature from Pmn21
P21/n
Pmnb. The majority of
22 previously reported studies on high temperature monoclinic LFS make no mention of the
23
24 presence of secondary orthorhombic phase except perhaps the work of Bini et al.34 despite
25
26 theoretical predictions of phase co-existence.10, 35
Interestingly in an annealing study by
27
Armstrong et al.36 involving as starting material hydrothermally precipitated LFS from Fe(II)-
28
29 containing solution the obtained phase at 400 °C was Pmn21 and not P21/n that was produced
30
31 here. It becomes evident therefore that the ethylenediamine-Fe(III)-silicate intermediate of the
32
33 present work makes possible the formation of P21/n monoclinic LFS phase at a lower annealing
34 temperature by opening a lower activation energy crystallization pathway.
35
36 Upon electrochemical cycling, the high temperature (~700°C) monoclinic LFS phase (P21/n
37
38 or γs) is well established to undergo a phase transition to low temperature orthorhombic one
39
40
(Pmn21 or βII).7, 14-15, 33, 37 However, due to the favorable energetic phase-coexisting behavior10, 35
41 and the experimental findings of this work, the presence and type of the secondary orthorhombic
42
43 phase should not be ignored, as it may play an important role in triggering the monoclinic to
44
45 orthorhombic phase transformation during electrochemical cycling.15 In other words, the
46
secondary phase (Pmn21) nanograins may alter the phase transformation reaction (P21/n

47
48 Pmn21) by acting as nucleation (seed) sites and in doing so affecting the long-term cycling
49
50 performance of LFS cathodes.
51
52
53
54
55
56
57
58
59 9
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 10 of 25

1
2
3
4
5 (a) (b)
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Figure 2. The Rietveld refined XRD spectra (Co Kα, λ ~ 1.78901 Å) of LFS400 (a) and LFS700 (b) materials. The
20
21 black lines indicate the experimentally collected spectra, the red bubbles indicate the Rietveld fitted patterns
22 assuming the coexistence of LFS monoclinic and orthorhombic phases, the blue lines indicate the discrepancies
23
24 between the experimental and calculated results. The XRD pattern of the LFS400 sample was fitted as a 10:90
25 mixture of Pmn21 and P21/n phases and that of the LFS700 sample as a 25:75 mixture of Pmnb and P21/n phases
26
27 (fitting error within 7%).
28
29
30 Morphology: TEM coupled with selected area electron diffraction (SAED) analysis
31
32 confirmed the Rietveld Refined XRD phase compositions of the LFS400 (Figure 3a, b, c) and
33 LFS700 (Figure 3d,e,f) samples. Both monoclinic and orthorhombic regions were visualized
34
35 distinctly in LFS400 and LFS700 nano/submicron particles. In the LFS400 sample, (102) and
36
37 (31 1 ) planes were detected with interplanar d-spacing of 0.35 nm and 0.24 nm, which
38
39 correspond to a monoclinic region of the sample; (001) and (210) planes were detected with
40
interplanar d-spacing of 0.52 nm and 0.26 nm, which correspond to an orthorhombic region of
41
42 the sample. Similarly, in the LFS700 sample, the (010) and (102) planes were detected with
43
44 interplanar d-spacing of 0.52 nm and 0.38 nm, corresponding to a monoclinic region of the
45
46 sample; the (100) and (010) planes were detected with interplanar d-spacing of 0.31 nm and 0.52
47 nm, corresponding to an orthorhombic region of the sample. From the TEM images (Figures 3a,
48
49 d), the crystallite size in LFS400 sample was deduced to be around 50 nm while for the LFS700
50
51 sample it was close to 200 nm due to apparent grain coarsening effect. It can be further noticed
52
that the crystallite clusters were covered with carbon and had some degree of porosity (Figure
53
54 S8, S9 in Supplementary Information), which are desirable features for a cathode material as
55
56 they allow for electrolyte infiltration and enhanced conductivity.7
57
58
59 10
60 ACS Paragon Plus Environment
Page 11 of 25 ACS Sustainable Chemistry & Engineering

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33 Figure 3. TEM characterization of LFS400 (a, b, c) and LFS700 (d, e, f) samples. (a, d) Typical crystal morphology
34
35 and size. (b) A monoclinic region of the LFS400 sample, where the diffraction spots of the (102) and (311) planes
36 are observed with interplanar d-spacing of 0.35 nm and 0.24 nm and (c) an orthorhombic region, where the
37
diffraction spots of the (001) and (210) planes are seen with interplanar d-spacing of 0.52 nm and 0.26 nm. The
38
39 incident electron beam was parallel to the [120] zone axis. (e) A monoclinic region of the LFS700 sample, where the
40
diffraction spots of the (010) and (102) planes are observed with interplanar d-spacing of 0.52 nm and 0.38 nm,
41
42 indexed to the zone axis [201] and (f) an orthorhombic region of the same sample, where the diffraction spots of the
43 (100) and (010) planes are seen with interplanar d-spacing of 0.31 nm and 0.52 nm. The incident electron beam was
44
45 parallel to the [001] zone axis.
46
47
48 The size and morphology of LFS400 and LFS700 nanostructured particles were also
49
50
examined using scanning electron microscopy (SEM) (Fig. 4). The LFS400 image (Fig 4a)
51 shows that the material was in the form of secondary submicron sized porous agglomerates of
52
53 unitary primary nanoparticles (typically around 40-60 nm). In contrary, the LFS700 image (Fig.
54
55 4b) shows that the agglomerates hierarchically organized into a 3D framework of rings of unitary
56
57
58
59 11
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 12 of 25

1
2
3 crystallites. The primary crystallites this time are seen to have grown to bigger size (around 100-
4
5 200 nm) as a result of the coarsening effect of the elevated annealing temperature (700°C). The
6
7 rings form openly connected porous cages of submicron size. Such porous structure can be
8
9
advantageous for the LFS cathode as it would allow for unhindered electrolyte infiltration and
10 thereby enhanced lithium ion transport.7 Additional EDS analysis coupled with SEM imaging
11
12 (Figure S10 and S11 in Supplementary Information) revealed the key constituent elements (C,
13
14 Si, Fe, O) to be uniformly dispersed in the selected region with no obvious signs of impurities
15
(i.e. SiO2 where Fe was deficient or iron oxides where Si was deficient) in both LFS400 and
16
17 LFS700 samples. Furthermore, the BJH pore size distribution derived from the BET adsorption
18
19 measurement showed that the LFS400 material had an average pore width around 50 nm, which
20
21 corresponds to Type I mesoporous material. (Supplementary Information Fig. S12) For LFS700
22 pristine sample it was mainly Type II macroporous however after ball milling it became Type I
23
24 mesoporous with pore size around 50nm as well (Supplementary Information Fig. S13). This
25
26 confirmed their microscopically observed porous structure.
27
28
29
30
31
32
33
34
35
36
37
38
39
40 Figure 4. SEM secondary electron images of LFS400 (a) and LFS700 (b) samples.
41
42 Carbon coating and iron speciation: In addition to the abovementioned phase and
43
44 morphology characteristics of the two annealed LFS samples, the in situ coated carbon layer
45
46
(evident in the TEM images given in Figure S8, S9 in Supplementary Information) and the
47 chemical states of the surface Fe species, are also important material attributes for LFS in Li-ion
48
49 battery cathode application. The LFS particles were coated with carbon as a result of the
50
51 decomposition of the surface-anchored organic additives (EN and EG) during annealing. The
52
carbon content was quantified using thermal gravimetric analysis (TGA) to be about 8% for
53
54 LFS400 (Figure S14 in Supplementary Information.) and slightly less than 5% for LFS700.
55
56 Furthermore, XPS spectra confirmed the presence of carbon on the surface as shown in Figure 5.
57
58
59 12
60 ACS Paragon Plus Environment
Page 13 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 Various carbon bonds were detected, including C-C, C-O, and C=O, located at 284.8 eV, 286.6
4
5 eV, and 289.9 eV respectively38-39 in both LFS400 and LFS700 samples. Furthermore, it is
6
7 interesting to notice that besides carbon, nitrogen signals were also detected (as -C-N-
8
9
(quaternary), -C=C-NH (pyrrolic) and -C-N=C- (pyridinic) bonds-as shown in Figure S15 in
10 Supplementary Information), strongly suggesting the decomposition of ethylenediamine
11
12 rendering the carbon layer N-doped. The presence of sp2-typed unsaturated carbon bonds
13
14 together with the N bonding made the carbon layer potentially more conductive due to the
15
additional electrons donated to the conduction band,40-41 an important aspect in enhancing the
16
17 functionality of LFS that has very poor electronic conductivity (10-12 – 10-16 S • cm-1 at room
18
19 temperature).42
20
21 XPS Fe spectra analysis revealed the LFS400 nanoparticles to have some Fe3+ species at the
22 outmost surface but not the LFS700 particles (Figure 5 b,d). The Fe 2p3/2 bonding at 711.4 eV of
23
24 the LFS400 spectrum corresponds to the ferric oxidation state while the Fe 2p3/2 binding at 710.4
25
26 eV of the LFS700 spectrum to the Fe2+ oxidation state.25 The presence of the ferric species in
27
LFS400 was ascribed in part to the inevitable air oxidation of the nanoparticles (smaller size than
28
29 that of LFS700-refer to Figure 3 a,d). The presence of ferric species in the bulk of the samples
30
31 was quantified by chemical titration following digestion. The corresponding quantities in
32
33 LFS400 and LFS700 was 4% and 1% ferric respectively. In addition to surface oxidation
34 incomplete reduction of ferric to ferrous during reductive annealing is held responsible for the
35
36 higher ferric fraction in LFS400 in agreement with the analysis presented in the previous section.
37
38 Other than quantifying the ferric amount, the chemical analysis of the digested solids yielded
39
40
deficient in Li content, namely the Li: Fe ratio is 1.69 and 1.72 in LFS400 and LFS700 samples.
41 (Table S2 in Supplementary Information) Such ratio corresponds to 15% of the iron as ferric in
42
43 pure LFS. Considering the fact that very little ferric (<4%) is actually present, this may imply
44
45 from a charge balance point of view the possible presence of about 10% of fayalite (Fe2SiO4) as
46 43
an impurity, which as isostructural olivine type mineral with the LFS is not XRD
47
48 distinguishable.
49
50
51
52
53
54
55
56
57
58
59 13
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 14 of 25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28 Figure 5. XPS spectra of LFS400 (a, b) and LFS700 (c, d) samples. The carbon spectra (a, c) were fitted to various
29
C-C and C-O bonds. The Fe spectrum for LFS400 sample (b) shows a shift of binding energy towards 711.4 eV
30
31 suggesting the presence of minor ferric impurity at the surface. The Fe spectrum of LFS700 sample (d) however
32 shows a peak binding energy at 710.4 eV corresponding solely to ferrous signal.
33
34
Mechanism of LFS particle formation- The enabling role of ethylenediamine
35
36
37 As mentioned earlier in addition to the three constituent components of LFS, introduced as
38
39
LiAc, Fe(NO3)3 and SiO2, ethylenediamine (EN) and ethylene glycol (EG) were also employed
40 as organic additives. By performing parallel tests without the two organic additives or using one
41
42 (EN or EG) organic additive, it was determined that only when EN is present, the ferric iron
43
44 silicate co-precipitate gel formed during the hydrothermal step led upon reductive annealing to
45
LFS crystallization (refer to XRD patterns of the respective annealed products in Figure S18 of
46
47 the Supplementary Information). The enabling role of ethylenediamine (EN) may be linked to its
48
49 dual function as bidentate ligand (NH2-CH2-CH2-NH2)44 and alkaline reagent that permits via
50
51 complexation (Fe3+ + zEN
Fe(EN)z3+) the controlled reaction of ferric ions30 with silicate
52 anions (Reaction 1 in Scheme 1) generated in situ by dissolution of silica as consequence of the
53
54 released alkalinity (pH ~9). Without EN the spontaneous hydrolysis of free (uncomplexed) ferric
55
56 ions45 leads to intermediates that are not amenable to transformation upon annealing to LFS.
57
58
59 14
60 ACS Paragon Plus Environment
Page 15 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 Meanwhile, ethylene glycol (EG) that is known for its surface active properties31plays a
4
5 supporting role to EN by stabilizing the hydrothermally formed precursor particles into a gel-like
6
7 assembly that upon subsequent annealing leads to mesoporous LFS crystal formation. It is
8
9
interesting that EG alone is not effective in rendering the Fe(III)-derived precipitate amenable to
10 transformation by reductive annealing to LFS even if the solution pH was the same (8-9) as when
11
12 EN was used (see XRD patterns in Figure S18). By contrast when hydrothermal precipitation
13
14 was done from Fe(II) solution, EG as per previous work46 was effective as crystallization
15
additive to obtain LFS. We conclude therefore that it is the bidentate complexing of ferric ions
16
17 by ethylenediamine the key in controlling the formation of the Li-FeIII-SiO4 precursor that has
18
19 provided favorable crystallization pathway to obtain LFS by reductive annealing (Reaction 2 in
20
21 Scheme 1).
22 The present LFS hydrothermal synthesis system that features ferric as the source of iron
23
24 differs from the hydrothermal system reported by Sirisopanaporn et al.13 where the LFS crystals
25
26 formed directly by hydrothermal precipitation. In their system, similar temperature/time
27
conditions were used except for the substitution of ferrous chloride for ferric nitrate. It appears
28
29 therefore that with the initial presence of Fe3+ ions the direct formation of crystalline Li2FeIISiO4
30
31 or the delithiated Li1FeIIISiO4 crystalline structure is blocked. Instead an amorphous precipitate
32 III
33 of Li-FeIII-SiO4 (resembling hisingerite: Fe (Si O ) (OH) ·2H O) was formed via the mediation
34 2 2 5 4 2
35 action of EN. 30, 47
The EN and EG organics of the resultant hydrothermal intermediate acted as a
36
37 source of carbonization during annealing forming an in situ carbon coating increasing the
38
39 material’s conductivity - an essential property in Li-ion electrode materials. The two-step
40
41
(hydrothermal precipitation and reductive annealing) reaction is graphically summarized in
42 Scheme 1.
43
44
45 2  + 2  + 4   +  +   →  −      reaction 1; hydrothermal reaction)
46
47  −      +  + 3  → 2 )

 *
48 +,-./0- -01,2.345
49 651 73 3524-84-6.345
9::::::::::::::;     * <=>?@A 2; <BC>?D ?<=AEF@<G=?@A
50
51
52
53
54
55
56
57
58
59 15
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 16 of 25

1
2
3
4
5 Fe2+ Carbon matrix
6
7
8 200°C (Ferric reduction and
9 lithium incorporation)
10 Organic Fe3+
11 molecules
12
13 Scheme 1. Reactions and schematic depicting the organic-stabilized amorphous precursor and its reactive
14 transformation during annealing under hydrogen into carbon-coated mesoporous LFS crystals.
15
16
17
18
19 Electrochemistry
20
21 The two LFS materials were subjected to a head-to-head electrochemical comparison by
22
23 measuring their initial galvanostatic charging/discharging behavior at a rate of C/10 at 55°C. As
24
described above the LFS400 material had a nanoparticle size ~50 nm while the LFS700 material
25
26 was coarser about 150-250 nm size. To ensure crystal size played no role in the electrochemical
27
28 comparison we had the LFS700 sample ball milled down to similar size (~50 nm) as the LFS400
29
30 (Figure 3a and Figure S17 in Supplementary Information). The first three charge/discharge
31 cycles for the two LFS materials are shown in Figure 6. Considering the charging profiles, we
32
33 observe that in both samples significant polarization has occurred during the first charge, despite
34
35 their nanosizing, as originally reported by Nyten et al.16 Thereafter (cycles 2 and 3) the intrinsic
36
37
LFS polarization was significantly removed and the charging profiles were stabilized with
38 practically equivalent corresponding charging capacities 180 and 170 mAh/g for LFS400 and
39
40 LFS700 respectively. These capacities correspond to slightly over 1 Li extraction as per charging
41
42 reaction shown below:
43
44 Li2FeIISiO4
LiFeIIISiO4 + Li+ + e- (1)
45
46
47
Shifting our attention to the discharging profiles we observe the two LFS materials to
48 exhibit significant differences despite that both were predominantly monoclinic and having
49
50 similar nanoparticle size. Thus, while LFS400 exhibited almost full reversibility after the third
51
52 cycle (170 mAh/g discharge capacity vs. 180 mAh/g charge capacity) the LFS700 exhibited only
53
125 mAh/g discharge capacity vs. 170 mAh/g charge capacity. This means that re-lithiation of
54
55 the charged LFS700 (reversal of reaction (1)) was severely hampered by some material defect.
56
57
58
59 16
60 ACS Paragon Plus Environment
Page 17 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 This capacity differential (charge vs. discharge capacity that amounts to 26%) for the LFS700
4
5 material on a first observation seems to correspond to the presence of 25% high-temperature
6
7 orthorhombic phase (Pmnb), which was found in an earlier study to suffer from severe
8
9
polarization15. CV analysis (curves presented in Figure S19b) seems to support this postulation.
10 Thus, in the formation cycle of the LFS700 cathode three anodic peaks were evident, namely at
11
12 2.86, 3.15 and 3.40 V. As per Sirisopanaporn et al.13, Pmnb has the lowest anodic potential (~2.9
13
14 V vs. Li+/Li) among the different LFS polymorphs reflecting a more covalent crystal structure
15
(shorter Fe-O bond) than the monoclinic P21/n, hence its irreversible behavior resulting in lower
16
17 discharge capacity. Alternatively the higher charge capacity of LFS700 vis-à-vis its discharge
18
19 capacity might indicate a contribution from a side reaction16, for example reaction of the pristine
20
21 LFS electrode with the electrolyte48.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Figure 6. The first three galvanostatic charging/discharging cycles of the LFS400 and LFS700 (ball milled) samples
39
40 at charging rate of C/10 (1C=166 mAh/g) at 55°C.
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 17
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 18 of 25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22 Figure 7. Comparison of cyclic voltammograms of LFS400 and LFS700 (ball milled) samples for second and third
23 cycles at 0.04 mV/s (corresponding to C/20) over the voltage range from 1.5 V to 4.67 V and 45°C.
24
25
26
27 Meanwhile from the shape of the discharging curves in Figure 6, we can deduce differences
28
29 in the Li-ion storage mechanism between the two LFS materials. Thus, the LFS700 (ball milled)
30
31 sample that exhibited a quasi-plateau in the 2.8-2.5 V discharge range has the characteristics of
32 two-phase reaction storage mechanism as also proposed in earlier studies involving monoclinic
33
34 LFS synthesized at 600-800°C.14, 16 In contrast the discharge curves of LFS400 were sloped
35
36 indicative of continuous solid solution mechanism: Li2-xFe1-xIIFexIIISiO4. These observations
37
38
were further corroborated by the shape of the CV curves for the second and third cycles
39 presented in Figure 7. Thus in the case of LFS700 the oxidation/reduction peaks (2.9 and 2.6 V)
40
41 were relatively narrow and more defined than the corresponding ones (3.1 and 2.6 V) of LFS400,
42
43 indicating the Li extraction/reinsertion reaction in LFS700 is characteristic of a two-phase
44
reaction mechanism16 and LFS400 is more of solid-solution type. This distinct Li-ion storage
45
46 behavior of LFS400 as discussed elsewhere is owed to its nanograined structure (abundance of
47
48 nanoscale grain boundaries).49 It is the combined effects of ethylenediamine/ethylene glycol and
49
50 medium range annealing temperature (400°C) as presented in this work that endowed the LFS
51 with mesoporosity and nanograined structure. Post-synthesis high energy ball milling of the
52
53 macroporous coarser LFS700 particles although successful in rendering the material mesoporous
54
55 and nanosized did not have the same effect on Li-ion storage properties as the LFS400. This
56
57
58
59 18
60 ACS Paragon Plus Environment
Page 19 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 clearly exemplifies the importance of achieving nanostructure control during the actual
4
5 crystallization process rather than through mechanical means.
6
7 Finally, it is of interest to compare these preliminary electrochemical measurements obtained
8
9
with the Fe(III)-derived LFS700 material of the present work to the performance of the most
10 common Fe(II)-derived LFS materials. For comparison we chose the monoclinic LFS (P21/n)
11
12 reported by Xu et al.46 as it had similar porous structure (compare Fig. 3c in ref no. 46 with
13
14 Figure 4b of this work) obtained also via a combination of hydrothermal precipitation (assisted
15
with ethylene glycol) and annealing/carbonization (for carbon coating) at 650ºC. Thus, in the
16
17 latter case the initial discharge capacity at C/10 (the same rate as applied here) was 72 mAh/g,
18
19 increasing and stabilizing at 110 mAh/g after 30 cycles. This increase in capacity attributed by
20
21 the authors to time required for the electrolyte to completely contact the cathode porous material
22 during cycling. By contrast our LFS700 exhibited initially 105 mAh/g discharge capacity that
23
24 increased to 125 mAh/g after only the third cycle (Figure 6b). This higher and faster capacity
25
26 attainment other than being due to the higher cycling temperature must at least in part reflected a
27
superior porous cathode structure. But other than superior cathode material characteristics, the
28
29 new synthesis route developed in this work provides attractive sustainability features when
30
31 compared to the typical synthesis approach as exemplified by the method of Xu et al.46. The
32
33 advantages include shorter hydrothermal processing time (12 hours vs. 72 hours), near
34 stoichiometric use of lithium (10% excess vs. 100%), higher yield with the use of concentrated
35
36 solutions (1 mol/L Fe vs. 0.16 mol/L Fe), lower cost abundant precursor chemicals (Fe(III) vs.
37
38 Fe(II) salts and SiO2 vs. Si(OC2H5)4 (TEOS)), and ease of handling (use of Fe(III) vs. air-
39
40 sensitive Fe(II) precursors). In addition, the H2/Ar (3-5vol% H2) gas mixture used is within
41 safety range and is widely employed in the metal treatment and semiconductor manufacturing
42
43 industries for annealing purposes.
44
45
46
47
48
49
50 Conclusion
51
52 In this work, nanostructured Li2FeSiO4 mesoporous particles have been successfully
53
54 synthesized using a novel sustainable two-step process that show highly promising
55
characteristics for fabricationg Li-ion battery cathodes. The first step involves ethylenediamine-
56
57
58
59 19
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 20 of 25

1
2
3 regulated hydrothermal precipitation from an aqueous solution at 180°C containing iron as ferric
4
5 salt rather than ferrous salt that previous studies have employed, plus colloidal silica and lithium
6
7 acetate. This step leads to the formation of a hydrous ferric silicate gel that is further stabilized
8
9
by the co-presence of ethylene glycol. Upon drying, the hydrothermal intermediate is subjected
10 to annealing in the presence of diluted H2 gas within its safety range (≤5vol% H2/Ar) that
11
12 triggers reduction of FeIII to FeII and simultaneous formation of crystalline Li2FeSiO4. We
13
14 demonstrate that unless ethylenediamine is used as complexing agent during hydrothermal
15
precipitation no LFS production is possible. Ethylenediamine is shown to assist in the formation
16
17 of the hydrated iron (III) silicate complex that provides a favorable crystallization pathway for
18
19 LFS. After reductive annealing at 200°C for 3 hours a nucleation cluster has been found to form
20
21 that subsequently grows into mainly monoclinic phase (P21/n) crystallites that assemble together
22 into porous cage-like aggregates as temperature is raised to 400°C and 700°C. The effect of
23
24 annealing time was investigated up to 10 hours at 400°C and it was found to promote
25
26 crystallinity and modest growth of nanocrystallites to ~40-50 nm size with 97% reduction of
27
Fe(III) to Fe(II) after 6 hours. The LFS400 material is endowed with nanograined structure and
28
29 be mesoporous. By contrast the LFS700 material is made of coarser grains (150-250 nm) that
30
31 upon mechanical milling become mesoporous as well. In both cases, the particles are found to be
32
33 coated with a N-doped carbon layer as result of in situ decomposition of the organic additives.
34 Head-to-head electrochemical comparison of the two LFS materials revealed differences in terms
35
36 of mode of Li-ion storage correlated to their nanostructure features and type of co-existing
37
38 orthorhombic phase (Pmn2 vs. Pmnb). Thus, the LFS400 material enriched with abundant
39
40
nanoscale grain boundaries exhibits near one-Li discharge capacity and solid solution storage
41 mechanism while the LFS700 material exhibits two-phase storage mechanism and somewhat
42
43 lower capacity. Overall the mesoporous C-coated nanostructured LFS materials derived from
44
45 Fe(III) salt compare favorably to those obtained more commonly from Fe(II) salt and as such are
46
viable cathode candidates. Finally, ethylenediamine-regulated hydrothermal processing
47
48 combined with reductive annealing provides sustainable crystallization options for producing
49
50 other air-sensitive energy storage materials with controlled nanostructure.
51
52
53
54 Supplementary Information
55
56
57
58
59 20
60 ACS Paragon Plus Environment
Page 21 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 The Supporting Information is available free of charge on the ACS Publications website at
4
5 DOI:xxxxxxxxxx
6
7
Detailed experimental procedure in preparing mesoporous LFS crystals; ICP analysis of the
8
9 solution after hydrothermal precipitation; XRD, FTIR, TEM characterization of the
10
11 hydrothermal intermediate, TEM images of the LFS400 samples obtained at different
12
13 annealing time; additional TEM, SEM/EDS, gas adsorption and pore size analysis, TGA,
14 XPS, Raman spectra and acid digestion analysis of the LFS400 and LFS700 samples; XRD
15
16 patterns of annealed products for purpose of screening of the organic additives; cyclic
17
18 voltammograms of the initial three cycles of the LFS400 and LFS700 samples.
19
20 Acknowledgements
21
22
23 This work is funded by Hydro-Quebec and NSERC (Natural Sciences and Engineering Research
24 Council of Canada) through a CRD grant. Prof. Xia Lu thanks the funding supports for State Key
25
26 Laboratory of Organic-Inorganic Composites (oic-201701011) and the Fundamental Research
27
28 Funds for the Central Universities (ZY1720).
29
30
31
32
33
References
34
35 1. Blomgren, G. E., The development and future of lithium ion batteries. J. Electrochem. Soc. 2017,
36 164 (1), A5019-A5025.
37 2. Larcher, D.; Tarascon, J. M., Towards greener and more sustainable batteries for electrical
38 energy storage. Nat Chem 2015, 7 (1), 19-29.
39 3. Goodenough, J. B.; Park, K.-S., The Li-Ion Rechargeable Battery: A Perspective. J. Am. Chem. Soc.
40
2013, 135 (4), 1167-1176.
41
42 4. Andre, D.; Kim, S.-J.; Lamp, P.; Lux, S. F.; Maglia, F.; Paschos, O.; Stiaszny, B., Future generations
43 of cathode materials: an automotive industry perspective. J. Mater. Chem. A 2015, 3 (13), 6709-6732.
44 5. Zaghib, K.; Guerfi, A.; Hovington, P.; Vijh, A.; Trudeau, M.; Mauger, A.; Goodenough, J. B.; Julien,
45 C. M., Review and analysis of nanostructured olivine-based lithium recheargeable batteries: Status and
46 trends. J. Power Sources 2013, 232, 357-369.
47 6. Islam, M. S.; Dominko, R.; Masquelier, C.; Sirisopanaporn, C.; Armstrong, A. R.; Bruce, P. G.,
48 Silicate cathodes for lithium batteries: alternatives to phosphates? J. Mater. Chem. 2011, 21 (27), 9811-
49
9818.
50
51
7. Ni, J.; Jiang, Y.; Bi, X.; Li, L.; Lu, J., Lithium Iron Orthosilicate Cathode: Progress and Perspectives.
52 ACS Energy Lett. 2017, 2 (8), 1771-1781.
53 8. Gong, Z.; Yang, Y., Recent advances in the research of polyanion-type cathode materials for Li-
54 ion batteries. Energy Environ. Sci. 2011, 4 (9), 3223-3242.
55
56
57
58
59 21
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 22 of 25

1
2
3 9. Zhang, P.; Hu, C. H.; Wu, S. Q.; Zhu, Z. Z.; Yang, Y., Structural properties and energetics of
4
Li2FeSiO4 polymorphs and their delithiated products from first-principles. Phys. Chem. Chem. Phys. 2012,
5
14 (20), 7346-7351.
6
7 10. Lu, X.; Chiu, H.-C.; Bevan, K. H.; Jiang, D.-T.; Zaghib, K.; Demopoulos, G. P., Density functional
8 theory insights into the structural stability and Li diffusion properties of monoclinic and orthorhombic Li
9 2 FeSiO 4 cathodes. J. Power Sources 2016, 318, 136-145.
10 11. Boulineau, A.; Sirisopanaporn, C.; Dominko, R.; Armstrong, A. R.; Bruce, P. G.; Masquelier, C.,
11 Polymorphism and structural defects in Li2FeSiO4. Dalton Trans. 2010, 39 (27), 6310-6316.
12 12. Armstrong, A. R.; Kuganathan, N.; Islam, M. S.; Bruce, P. G., Structure and lithium transport
13 pathways in Li2FeSiO4 cathodes for lithium batteries. J. Am. Chem. Soc. 2011, 133 (33), 13031-13035.
14
13. Sirisopanaporn, C.; Masquelier, C.; Bruce, P. G.; Armstrong, A. R.; Dominko, R., Dependence of
15
16 Li2FeSiO4 Electrochemistry on Structure. J. Am. Chem. Soc. 2011, 133 (5), 1263-1265.
17 14. Masese, T.; Orikasa, Y.; Tassel, C.; Kim, J.; Minato, T.; Arai, H.; Mori, T.; Yamamoto, K.; Kobayashi,
18 Y.; Kageyama, H.; Ogumi, Z.; Uchimoto, Y., Relationship between Phase Transition Involving Cationic
19 Exchange and Charge–Discharge Rate in Li2FeSiO4. Chem. Mater. 2014, 26 (3), 1380-1384.
20 15. Lu, X.; Wei, H.; Chiu, H.-C.; Gauvin, R.; Hovington, P.; Guerfi, A.; Zaghib, K.; Demopoulos, G. P.,
21 Rate-dependent phase transitions in Li2FeSiO4 cathode nanocrystals. Sci. Rep. 2015, 5, 8599.
22 16. Nytén, A.; Abouimrane, A.; Armand, M.; Gustafsson, T.; Thomas, J. O., Electrochemical
23
performance of Li2FeSiO4 as a new Li-battery cathode material. Electrochem. Commun. 2005, 7 (2),
24
25
156-160.
26 17. Dominko, R.; Conte, D.; Hanzel, D.; Gaberscek, M.; Jamnik, J., Impact of synthesis conditions on
27 the structure and performance of Li 2 FeSiO 4. J. Power Sources 2008, 178 (2), 842-847.
28 18. Muraliganth, T.; Stroukoff, K. R.; Manthiram, A., Microwave-Solvothermal Synthesis of
29 Nanostructured Li2MSiO4/C (M = Mn and Fe) Cathodes for Lithium-Ion Batteries. Chem. Mater. 2010,
30 22 (20), 5754-5761.
31 19. Rangappa, D.; Murukanahally, K. D.; Tomai, T.; Unemoto, A.; Honma, I., Ultrathin Nanosheets of
32 Li2MSiO4 (M = Fe, Mn) as High-Capacity Li-Ion Battery Electrode. Nano Lett. 2012, 12 (3), 1146-1151.
33
20. Chen, J., A review of nanostructured lithium ion battery materials via low temperature synthesis.
34
35 Recent Pat. Nanotechnol. 2013, 7 (1), 2-12.
36 21. Brochu, F.; Guerfi, A.; Trottier, J.; Kopeć, M.; Mauger, A.; Groult, H.; Julien, C. M.; Zaghib, K.,
37 Structure and electrochemistry of scaling nano C–LiFePO4 synthesized by hydrothermal route:
38 Complexing agent effect. J. Power Sources 2012, 214 (Supplement C), 1-6.
39 22. Chen, Z.; Qiu, S.; Cao, Y.; Qian, J.; Ai, X.; Xie, K.; Hong, X.; Yang, H., Hierarchical porous Li 2 FeSiO
40 4/C composite with 2 Li storage capacity and long cycle stability for advanced Li-ion batteries. J. Mater.
41 Chem. A 2013, 1 (16), 4988-4992.
42
23. Aravindan, V.; Karthikeyan, K.; Lee, J.; Madhavi, S.; Lee, Y., Synthesis and improved
43
44
electrochemical properties of Li2MnSiO4 cathodes. J. Phys. D: Appl. Phys. 2011, 44 (15), 152001.
45 24. Li, D.; Zhang, W.; Sun, R.; Chen, G.; Fan, X.; Gou, L.; Mao, Y.; Zhao, K.; Tian, M., Soft-template
46 construction of three-dimensionally ordered inverse opal structure from Li 2 FeSiO 4/C composite
47 nanofibers for high-rate lithium-ion batteries. Nanoscale 2016, 8 (24), 12202-12214.
48 25. Sun, C.; Rajasekhara, S.; Goodenough, J. B.; Zhou, F., Monodisperse Porous LiFePO4
49 Microspheres for a High Power Li-Ion Battery Cathode. J. Am. Chem. Soc. 2011, 133 (7), 2132-2135.
50 26. Billaud, J.; Eames, C.; Tapia-Ruiz, N.; Roberts, M. R.; Naylor, A. J.; Armstrong, A. R.; Islam, M. S.;
51
Bruce, P. G., Evidence of Enhanced Ion Transport in Li-Rich Silicate Intercalation Materials. Adv. Energy
52
Mater. 2017, 7 (11), 1601043-n/a.
53
54 27. Jia, Y.; Xu, L.; Fang, Z.; Demopoulos, G. P., Observation of Surface Precipitation of Arsenate on
55 Ferrihydrite. Environ. Sci. Technol. 2006, 40 (10), 3248-3253.
56
57
58
59 22
60 ACS Paragon Plus Environment
Page 23 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 28. Manceau, A.; Ildefonse, P.; Hazemann, J. L.; Flank, A. M.; Gallup, D., Crystal chemistry of hydrous
4
iron silicate scale deposits at the Salton Sea geothermal field. Clays Clay Miner. 1995, 43 (3), 304-317.
5
29. Kageyama, H.; Hashimoto, Y.; Oaki, Y.; Imai, H., Six-armed twin crystals composed of lithium iron
6
7 silicate nanoplates and their electrochemical properties. CrystEngComm 2015, 17 (44), 8486-8491.
8 30. Zhang, M.; Chen, K.; Chen, X.; Peng, X.; Sun, X.; Xue, D., Ethylenediamine-assisted crystallization
9 of Fe 2 O 3 microspindles with controllable size and their pseudocapacitance performance.
10 CrystEngComm 2015, 17 (7), 1521-1525.
11 31. Kim, D.-H.; Kim, J., Synthesis of LiFePO4 nanoparticles in polyol medium and their
12 electrochemical properties. Electrochem. Solid-State Lett. 2006, 9 (9), A439-A442.
13 32. Ferrari, S.; Capsoni, D.; Casino, S.; Destro, M.; Gerbaldi, C.; Bini, M., Electrochemistry of
14
orthosilicate-based lithium battery cathodes: a perspective. Phys. Chem. Chem. Phys. 2014, 16 (22),
15
16 10353-10366.
17 33. Brownrigg, A. W.; Mountjoy, G.; Chadwick, A. V.; Alfredsson, M.; Bras, W.; Billaud, J.; Armstrong,
18 A. R.; Bruce, P. G.; Dominko, R.; Kelder, E. M., In situ Fe K-edge X-ray absorption spectroscopy study
19 during cycling of Li 2 FeSiO 4 and Li 2.2 Fe 0.9 SiO 4 Li ion battery materials. J. Mater. Chem. A 2015, 3
20 (14), 7314-7322.
21 34. Bini, M.; Ferrari, S.; Ferrara, C.; Mozzati, M. C.; Capsoni, D.; Pell, A. J.; Pintacuda, G.; Canton, P.;
22 Mustarelli, P., Polymorphism and magnetic properties of Li2MSiO4 (M = Fe, Mn) cathode materials. Sci.
23
Rep. 2013, 3, 3452.
24
25
35. Saracibar, A.; Van der Ven, A.; Arroyo-de Dompablo, M., Crystal structure, energetics, and
26 electrochemistry of Li2FeSiO4 polymorphs from first principles calculations. Chem. Mater. 2012, 24 (3),
27 495-503.
28 36. Armstrong, A. R.; Sirisopanaporn, C.; Adamson, P.; Billaud, J.; Dominko, R.; Masquelier, C.; Bruce,
29 P. G., Polymorphism in Li2MSiO4 (M = Fe, Mn): A Variable Temperature Diffraction Study. Angew. Chem.,
30 Int. Ed. 2014, 640 (6), 1043-1049.
31 37. Masese, T.; Tassel, C. d.; Orikasa, Y.; Koyama, Y.; Arai, H.; Hayashi, N.; Kim, J.; Mori, T.;
32 Yamamoto, K.; Kobayashi, Y., Crystal Structural Changes and Charge Compensation Mechanism during
33
Two Lithium Extraction/Insertion between Li2FeSiO4 and FeSiO4. J. Phys. Chem. C 2015, 119 (19),
34
35 10206-10211.
36 38. Gelius, U.; Heden, P.; Hedman, J.; Lindberg, B.; Manne, R.; Nordberg, R.; Nordling, C.; Siegbahn,
37 K., Molecular spectroscopy by means of ESCA III. Carbon compounds. Phys. Scr. 1970, 2 (1-2), 70.
38 39. Pertsin, A.; Gorelova, M.; Levin, V. Y.; Makarova, L., An XPS study of the surface–bulk
39 compositional differences in siloxane‐containing block copolymers and polymer blends. J. Appl. Polym.
40 Sci. 1992, 45 (7), 1195-1202.
41
40. Liu, H.; Zhang, Y.; Li, R.; Sun, X.; Désilets, S.; Abou-Rachid, H.; Jaidann, M.; Lussier, L.-S.,
42
43 Structural and morphological control of aligned nitrogen-doped carbon nanotubes. Carbon 2010, 48 (5),
44 1498-1507.
45 41. Wang, H. g.; Wu, Z.; Meng, F. l.; Ma, D. l.; Huang, X. l.; Wang, L. m.; Zhang, X. b., Nitrogen‐
46 Doped Porous Carbon Nanosheets as Low‐Cost, High‐Performance Anode Material for Sodium‐Ion
47 Batteries. ChemSusChem 2013, 6 (1), 56-60.
48 42. Ni, J.; Zhang, L.; Fu, S.; Savilov, S.; Aldoshin, S.; Lu, L., A review on integrating nano-carbons into
49
polyanion phosphates and silicates for rechargeable lithium batteries. Carbon 2015, 92, 15-25.
50
51
43. Zhang, Q.; Ge, S.; Xue, H.; Wang, X.; Sun, H.; Li, A., Fabrication of a fayalite@C nanocomposite
52 with superior lithium storage for lithium ion battery anodes. RSC Adv. 2014, 4 (102), 58260-58264.
53 44. Paoletti, P., Formation of metal complexes with ethylenediamine: a critical survey of equilibrium
54 constants, enthalpy and entropy values. Pure Appl. Chem. 1984, 56 (4), 491-522.
55
56
57
58
59 23
60 ACS Paragon Plus Environment
 ACS Sustainable Chemistry & Engineering Page 24 of 25

1
2
3 45. Cheng, T. C.; Demopoulos, G. P., Hydrolysis of Ferric Sulfate in the Presence of Zinc Sulfate at
4
200 °C:  Precipitation Kinetics and Product Characterization. Ind. Eng. Chem. Res. 2004, 43 (20), 6299-
5
6308.
6
7 46. Xu, Y.; Shen, W.; Zhang, A.; Liu, H.; Ma, Z., Template-free hydrothermal synthesis of Li2FeSiO4
8 hollow spheres as cathode materials for lithium-ion batteries. J. Mater. Chem. A 2014, 2 (32), 12982-
9 12990.
10 47. Acharya, S.; Maheshwari, N.; Tatikondewar, L.; Kshirsagar, A.; Kulkarni, S., Ethylenediamine-
11 mediated wurtzite phase formation in ZnS. Cryst. Growth Des. 2013, 13 (4), 1369-1376.
12 48. Arthur, Z.; Chiu, H.-C.; Lu, X.; Chen, N.; Emond, V.; Zaghib, K.; Jiang, D.-T.; Demopoulos, G. P.,
13 Spontaneous reaction between an uncharged lithium iron silicate cathode and a LiPF 6-based electrolyte.
14
Chem. Commun. 2016, 52 (1), 190-193.
15
16 49. Lu, X.; Chiu, H.-C.; Arthur, Z.; Zhou, J.; Wang, J.; Chen, N.; Jiang, D.-T.; Zaghib, K.; Demopoulos, G.
17 P., Li-ion storage dynamics in metastable nanostructured Li2FeSiO4 cathode: Antisite-induced phase
18 transition and lattice oxygen participation. J. Power Sources 2016, 329 (Supplement C), 355-363.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 24
60 ACS Paragon Plus Environment
Page 25 of 25 ACS Sustainable Chemistry & Engineering

1
2
3 For Table of Contents Use Only
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32 Description of TOC:
33
34 Mesoporous Li2FeSiO4 (LFS) particulates produced sustainably in high yield from ferric salt
35
36 solution with help of ethylenediamine; annealing tunable Li-ion storage properties
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 25
60 ACS Paragon Plus Environment