13 MINERAL PROCESSING AND METALLURGICAL TESTING

Mineral processing testing was performed to evaluate the potential of spodumene

concentrate production and lithium hydroxide monohydrate (LHM) production separately. The
spodumene concentrate production test work is presented in Section 13.1. The
electrochemical production of LHM is presented in Section 13.2.

13.1 Spodumene Concentration

Preliminary metallurgical investigation of the Whabouchi deposit was first carried out in 2010
and 2011 by SGS at Lakefield, Ontario.

Some of the bench scale and pilot plant test work results were reported in previous Technical
Reports in 2012, 2013, 2016, and 2018.

This Section provides a summary on the most relevant work that was performed over the years
to develop the actual process flow sheet used to concentrate the spodumene. This involves
ore sorting, hydro-classification, dense media separation (“DMS”) and flotation methods. It
also includes summaries from screening, settling, filtration, freezing, drying, and magnetic
separation tests performed by various laboratories and suppliers.

13.1.1 ORE SORTING TESTING

13.1.1.1 TOMRA Test Work Program

Nemaska provided TOMRA with 1.8 tonnes of feed ore ranging in size from 40 mm to 9.5 mm
to conduct testing program to remove black rock (amphibolite) from white rocks (pegmatite).
The material was sent to the TOMRA test center in Wedel, Germany.

For the first tests of the Nemaska ore, several TOMRA sorting technologies and several
approaches were considered to select the ones that responded best. The data collected
according to the characteristics of the rocks present in the ores made it possible to show the
potential for sorting of each sensor type for a specific task. Two (2) technologies were targeted
for sorting trials: surface color differentiation sorting technology and X-ray transmission
(“XRT”). XRT uses atomic density difference separation technology.

Below the best results of the two (2) size fractions are discussed.

a. Ore Sorting on - 40 mm + 20 mm, Amphibole Removal

Using XRT, the accepted material white rocks recovery was 95.8%, with white rock at 98.7%
concentration. The photo in Figure 13.1 illustrates the visual results. There is very little
displaced material in the accepted stream.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 110

August 2019
DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.1 – TOMRA Ore Sorting Test #1

Source: TOMRA test center in Wedel, Germany March 2017

b. Ore Sorting on - 20 mm + 9.5 mm, Amphibole Removal

Using XRT, the accepted material white rocks recovery was 96.8%, with white rock at 99.2%
concentration.

13.1.1.2 Steinart Test Work Program

Steinert US was approached to test Nemaska Whabouchi ore with sensor-based sorting
techniques to beneficiate/upgrade the ore by removing waste rock consisting mainly of
amphibolite. The test sample provided consisted of pegmatite and dark amphibolite waste
rocks.

The objective was test sorting efficiency of Steinert ore sorting equipment. The sorting sensor
selected was X-Ray Transmission or XRT. X-ray sensitive line-scan sensor provides high
resolution X-ray absorption images. An X-ray scintillation crystal sensor can capture up to
2,500 lines per second.

Nemaska had prepared two (2) tonnes of sample. The sample was screened at 10 mm to
remove the fines which is not suitable for this sorting application. The first set of three tests
were performed on - 50mm + 15mm material. This size range is good ore sorter feed material,
about ratio 1:3 sorting size. The next three (3) tests were done on - 15mm + 10mm material.
Finally, the last three (3) tests were done on - 50 + 10mm size range, outside the ideal size
ratio.

a. Ore Sorting on - 50 mm + 15 mm

The coarse sample - 50 mm + 15 mm yielded excellent results. Test #1 yielded the best result
at 99.4% separation efficiency of the white ore and was separated from the dark amphibolite.
The photos in Figure 13.2, illustrate the visual results. There is very little misplaced material in
the accepted and rejected streams.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 111

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.2 – Steinert Ore Sorting Test #1

Source: Steinert US Test Laboratory April 2017

b. Ore Sorting on - 15 mm + 10 mm

The finer sample - 15 mm + 10 mm was processed through the same equipment and also
produced very good results, but noticeably less than the coarse separation. The best
separation efficiency obtained was 96.7%.

c. Ore Sorting on - 50 mm + 10 mm

For the combined sample - 50 mm + 10 mm, the best separation efficiency obtained was
94.5%.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 112

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

13.1.2 ERIEZ HYDRAULIC SEPARATION TEST WORK

Hydraulic separation was performed at Eriez Central Test Laboratory in Erie, PA. Eriez received
spodumene ore from Nemaska for testing. Hydraulic separation test work was done with two
(2) different size fractions (- 8 mm + 0.85 mm) and (-0.85 mm). The tests were aimed to
remove mica from the ore.

13.1.2.1 Hydraulic Separation on - 8 mm + 0.85 mm, Mica Removal

The coarser sample - 8 mm + 0.85 mm was processed using a CrossFlow separation equipment,
following the flow sheet shown in Figure 13.3. The material was processed in 9 × 16 in. Eriez
CrossFlow Separator. The teeter water up flow as 1.83 cm/s.

Figure 13.3 – Flow Sheet CrossFlow Test #6

Source: Eriez Flotation Division Test Laboratory 2017.

The photos shown in Figure 13.4 illustrate the effectiveness of the CrossFlow separator on
coarse muscovite particles.

To a certain extent, the K2O concentration can be related to the muscovite concentration in
the head samples. However, it is not the only potassium bearing mineral. In the case of
Whabouchi ore, the presence of K-Feldspar does not allow to evaluate the muscovite recovery
based on the K₂O analyses. The CrossFlow separator overflow, which contains mainly coarse
muscovite flakes and fines particles entrained with the upward current was screened at 2.0
mm. This allowed the recovery of mainly muscovite and the return of entrained particles into
the flotation circuit. The analytical results are presented in Table 13.1.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 113

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Pure muscovite is about 11.8% K2O. It can be seen that the screen oversize contains a very
high concentration of muscovite as the other potassium bearing minerals are not retained by
the screen because of their size and shape.

Figure 13.4 – Photographs Classified CrossFlow Test #6

Source: Eriez Flotation Division Test Laboratory 2017.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 114

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.1 – CrossFlow Separation Test #6

ID Stream

Weight (%)

Li2O (%)

Dist. Li2O (%)

K2O (%)
Dist. K2O (%)

1 Feed 100.00 1.70 100.00 2.97 100.00

2 XF6 Overflow 16.88 0.64 6.32 5.28 30.03

3 XF6 Underflow 83.12 1.91 93.68 2.50 69.97

4 Screen Overflow 0.69 0.44 0.18 10.02 2.32

5 Screen Underflow 16.19 0.64 6.14 5.07 27.71

6 DMS Feed 99.31 1.71 99.82 2.92 97.67 Source: Eriez Flotation Division Test Laboratory 2017.

In the table 13.1, the Li2O concentration in the screen overflow corresponds to the natural
content in the muscovite. The loss in spodumene is negligible.

13.1.2.2 Hydraulic Separation on - 0.85 mm Mica Removal

The finer sample - 0.85 mm was processed using CrossFlow separation equipment, following
the flow sheet shown in Figure 13.5. Two (2) CrossFlow separators in series with screens were
used. The screen oversize is considered mica waste and the remainder will be flotation feed.

Figure 13.5 – Flow Sheet CrossFlow Test #1 and #2

Source: Eriez Flotation Division Test Laboratory 2017.

Both CrossFlow overflows were screened at 212 microns. The analytical results are presented
in Table 13.2.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 115

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

As can be seen, the weight rejection at the second CrossFlow, which used much more water,
was very high. This is considered too high and cannot be accepted by the process as a high loss
of lithium was observed. The CrossFlow operation will have to be adjusted to make a slightly
finer cut which will be sufficient to reject entrained muscovite since its shape factor allows it to
be rejected at lower water flows.

Table 13.2 – CrossFlow Separation Test #1 and #2

ID Stream
Weight (%)

Li₂O (%)

Dist. Li₂O (%)

K₂O (%)

Dist. K₂O (%)

1 Feed 100.00 1.39 100.00 2.32 100.00

2 XF 1 Overflow 34.87 0.85 21.24 2.55 38.35

3 XF 1 Underflow 65.13 1.68 78.76 2.20 61.65

4 Screen 1 Overflow 2.10 0.34 0.52 5.72 5.19

5 Screen 1 Underflow 32.76 0.88 20.73 2.35 33.15

6 XF 2 Overflow 18.19 1.06 13.86 2.68 20.98

7 XF 2 Underflow 46.94 1.92 64.90 2.01 40.67

8 Screen 2 Overflow 12.70 0.58 5.27 3.18 17.43

9 Screen 2 Underflow 5.49 2.18 8.59 1.50 3.55

10 Mica Rejects 14.81 0.54 5.78 3.55 22.63

11 Combined Screen U/F 38.25 1.07 29.32 2.23 36.70 Source: Eriez Flotation Division Test
Laboratory 2017.

13.1.3 DMS TEST WORK

13.1.3.1 DMS Test Work at SGS – 2011

Four (4) composites were prepared as pilot plant feed. Two (2) composites were prepared for
the flotation pilot plant and were labelled as “Outcrop Sample for flotation” and “Mine
Representative Sample for Flotation”. Two (2) more composites were prepared for DMS pilot
plant test work and were labelled as “DMS Outcrop Composite” and “DMS-Mine
Representative Composite”. The mine representative sample was composed of outcrop
material and drill core rejects. A detailed description of these composite is given in the SGS
report titled: “A pilot plant investigation into The Recovery of Spodumene from the
Whabouchi Property, Project 12486-003 – Final Report, April 2, 2012”. The chemical analyses
of these composites are presented in Table 13.3.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 116
August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.3 – Chemical Analysis of the Flotation Pilot Plant Composites

Composite Unit Outcrop

Mine Representative

DMS Outcrop Middlings

DMS Mine Rep. Middlings

Ele mental

Analysis

Li % 0.76 0.72 0.68 0.86

Li2O % 1.64 1.55 1.47 1.85

BE ppm 178 156 --- ---

Beryl % 0.27 0.24 --- ---

Wh ole Rock Anal ysis

SiO2 % 74.4 74.4 76.1 76

Al2O3 % 15.9 15.9 14.9 15.7

Fe2O3 % 0.78 0.9 0.85 0.88

MgO % 0.09 0.18 0.09 0.08

CaO % 0.26 0.43 0.3 0.3

Na2O % 3.36 3.34 3.42 3.05

K2O % 2.45 2.63 2.08 2.08

TiO2 % 0.01 0.03 0.01 0.01

P2O5 % 0.1 0.12 0.1 0.1

MnO % 0.08 0.1 0.1 0.08

Cr2O3 % 0.03 0.03 0.03 0.03

V2O5 % < 0.01 < 0.01 < 0.01 < 0.01

LOI % 0.72 0.84 0.57 0.84

Sum % < 98.5 98.9 98.6 99.1

Source: SGS Mineral Services; Project 12486-003 – Final Report, April 2, 2012

Semi quantitative XRD analyses of composites are provided in Table 13.4. The results indicate
that the outcrop and DMS mine representative samples are very similar in make-up. The Mine
Representative sample has less spodumene than the two (2) others. The mineral spodumene
content is slightly over 20%.

The Mine Representative sample was crushed and screened, and the liberation of Li-minerals
was assessed at 91.4%.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 117

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.4 – Semi –Quantitative XRD Analyses of the Composite

Mineral

Outcrop Sample (Weight%)

DMS Mine Rep Sample (Weight%)

Mine Rep Sample (Weight%)

Quartz 26.1 26.9 33.4

Spodumene (Monoclinic) 22.6 22.3 20.8

Albite 31.1 31.2 29.3

Microcline 14.9 17.7 11.5

Magnesiohornblende - - 1.0

Muscovite 5.3 1.9 4.0

Total 100.0 100.0 100.0

Source: SGS Mineral Services; Project 12486-003 – Final Report, April 2, 2012

The DMS pilot plant test work was carried out on two (2) samples, the blasted sample and the
mine representative sample, from the Whabouchi deposit by SGS at Lakefield in 2011. The
flow sheet used during the pilot plant, consisting of several unit operations, includes crushing,
scrubbing, screening, several dense media separation stages, magnetic separation and
dewatering. The DMS test plant was equipped with a 150-mm dense media cyclone. Up to
eight (8) DMS stages were used for the blasted sample, whereas only four (4) stages were used
in test work on the mine representative sample. Magnetic separation was conducted by using
a high intensity rare earth roll magnetic separator to upgrade DMS sinks on previously dried
feed.

The results of the DMS pilot plant were reported in an addendum by SGS (12486-004) entitled
“An Investigation into DMS Plant Testing on Material from the Whabouchi Lithium Deposit”
issued November 11th, 2011 and incorporated in the November 16, 2012 NI 43-101 Technical
Report Preliminary Economic Assessment.

The main findings from these results were that for mine representative sample, the 4-stage
DMS flow sheet rejects 40% of the feed mass as tailings at a top size of 9.5 mm, at a loss of
10% Li. At a top size of 9.5 mm, approximately 11% of the feed mass is recovered as
spodumene concentrate grading 6.4% Li2O and 45% Li distribution. The combined middlings,
49% of the feed mass, represent the remaining 45% Li distribution.

13.1.3.2 Met-Solve Laboratories Test Program – 2013-2014

In 2013, Nemaska contracted Met-Solve Laboratories to carry out DMS pilot plant test work.

DMS testing was investigated as the costs of this form of processing will be considerately lower
as compared to flotation.

DMS pilot plant testing was carried out using a single 250 mm separator to simulate multi-
stage separations. The primary objective of this test program was to determine the overall
grade and recovery of lithium using a DMS system.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 118

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

The multi-stage separator has been shown to offer better performance compared to single
stage, two (2) products and other conventional dense medium cyclone separators. The
following section is a summary of the report issued by Met-Solve titled: “Nemaska Lithium Inc.,
Dense Media Separation, MS 1467 issued August 13, 2013”.

Four (4) drums of pre-crushed sample (- 9.5 mm + 0.5 mm) weighing approximately 900 kg
were sent from SGS at Lakefield to Met-Solve laboratories in Langley, BC. Only half of the
sample (approximately 450 kg) was used for these DMS tests.

A total of seven (7) DMS tests (FT201 to FT207) were carried out on these pre-crushed
samples. Approximately 65 kg of representative sample was used for each test.
The initial five (5) DMS tests (FT201 to FT205) were aimed to simulate a 4-stage DMS
concentrator with a re-circulated test on the crushed middlings (Figure 13.6). Each run
consisted of passing the floats through the single media separator twice to simulate a 2-stage
concentrator. The total sinks (combined Sinks 1, Sinks 3 and Sinks 4) can be considered final
concentrate, while Floats 2 is final tailings. Floats 4 and the –0.5 mm rejects could be sent to
the flotation circuit to improve the lithium recovery.

Figure 13.6 – General Process Flow Sheet for Initial DMS Test Work (Tests FT201-FT205)

Source: MS1467; Nemaska Lithium Inc., Dense Media Separation Final Report – August 13th,
2013

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 119

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.5 lists the results of the five (5) tests. In each test, propensity of the spodumene ore
to dense media separation was confirmed.

Table 13.5– Process Flow Sheet for Initial DMS Test Work Results

Test

Mass (%) Li2O Distribution (%) Li2O Assay (%)

Total Sinks

Floats 2

Floats 4

Total Sinks

Floats 2

Floats 4

Calc. Head

Total Sinks

Floats 2

Floats 4

FT201 17.9 78.9 1.6 59.9 35.6 1.6 1.61 5.39 0.73 1.61
FT202 16.2 79.1 2.4 56.7 35.3 3.6 1.62 5.67 0.72 2.39

FT203 14.9 81.7 1.7 53.3 40.6 2.7 1.59 5.67 0.79 2.52

FT204 17.5 78.2 2.4 59.0 34.2 3.3 1.68 5.56 0.72 2.32

FT205 17.6 77.9 2.0 59.2 34.3 2.1 1.68 5.66 0.74 1.73 Source: MS1467; Nemaska Lithium Inc.,
Dense Media Separation Final Report – August 13th, 2013

Two (2) additional tests, FT206 without and FT207 with crushing middlings, were carried out to
simulate a dynamic 3-stage DMS circuit, in order to assess the effectiveness of adding more
separation stages. Test FT206 yielded a higher recovery of 66.8%, but with a lower concentrate
grade of 5.17% Li2O. The tailings grade was 0.74% Li2O, which is comparable to the initial
tests. The results of test FT206 are listed in Table 13.6.

Table 13.6 – FT206 3-Stage DMS Test

Description: FT206 (3-Stage DMS)

1st Stage 2nd Stage 3rd Stage

*D50 (Specific Gravity): 3.00 3.00 2.87

Specific Gravity of Dense Media: 2.84 2.84 2.44

Medium Inlet Pressure (psi): 28 28 28 *Based on tracer tests Feed Particle Size: (- 9.5 mm
+ 0.5 mm)

Description

Weight Assay Distribution

kg % Li2O % Si % Al % Li2O % Si % Al %

Sinks 1 4.27 6.8 6.16 29.6 11.60 24.2 6.1 10.3

Sinks 2 1.67 2.7 5.89 31.0 11.55 9.0 2.5 4.0

Sinks 3 8.07 12.9 4.50 32.8 9.86 33.5 12.7 16.5

Total Sinks 14.01 22.4 5.17 31.6 10.59 66.8 21.3 30.8

Floats 48.48 77.6 0.74 33.7 6.89 33.2 78.7 69.2

Calc Head 62.48 100.0 1.74 33.2 7.72 100.0 100.0 100.0

Calc Head (SFA) 1.65 33.7 7.81

Source: MS1467; Nemaska Lithium Inc., Dense Media Separation Final Report – August 13,
2013

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 120
August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

The flow sheet of test F207 with the crushing step is shown in Figure 13.7. Test FT207 yielded
the best results, based on recovery and final concentrate grade. This indicates that lower
density settings with less coarse feed may yield even improved results.

Figure 13.7 – General Process Flow Sheet for Initial DMS Test Work (Tests FT207)

Source: MS1467; Nemaska Lithium Inc., Dense Media Separation Final Report – August 13th,
2013

The results of test FT207 indicate that using the 3-stage DMS the Floats 1 lithium content can
be reduced with a lower dense media pulp specific gravity. The sinks grade can be improved by
using some sinks as middlings to be crushed and re-circulated. In conclusion, test FT207 had a
low loss

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 121

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

of lithium while producing a relatively large quantity of DMS concentrate. The low tailings part
is important as the lower grade sinks can be reprocessed as middlings in the flotation circuit.
The results from FT207 are listed in Table 13.7.

Table 13.7 – Results of Test FT207

Description

Weight Assay Distribution Li2O Si Al Li2O Si Al

(kg) (%) (%) (%) (%) (%) (%) (%)

Sink 1 7.40 10.9 5.87 29.6 11.55 37.6 9.4 15.7

Sink2 (–0.5 mm) 2.16 3.2 3.52 32.1 9.28 6.6 3.0 3.7

Sink3 (–3.4 mm) 4.76 7.0 1.85 36.1 8.03 7.6 7.4 7.0
Sink3 (–0.5 mm) 2.74 4.1 1.30 32.6 7.27 3.1 3.8 3.7

Sink 4 1.69 2.5 6.32 30.9 11.95 9.3 2.2 3.7

Sink 5 3.37 5.0 5.02 32.7 10.55 14.7 4.7 6.5

Sink 6 4.30 6.3 2.38 34.8 8.26 8.9 6.4 6.5

Total Sinks 26.42 39.0 3.83 32.6 9.65 87.7 36.9 46.7

Floats 1 37.36 55.2 0.31 35.7 7.08 10.0 57.1 48.4

Floats2 3.91 5.8 0.69 35.5 6.77 2.3 5.9 4.8

Calc. Head 67.69 100.0 1.71 34.4 8.06 100.0 100.0 100.0 Calc. Head (SFA) 1.65 33.7 7.81

Source: MS1467; Nemaska Lithium Inc., Dense Media Separation Final Report – August 13th,
2013

In 2014, test FT700 was carried-out to re-visit the FT207 test, without crushing and re-
circulation of the lower grade Sinks. FT700 indicates that when using the 3-stage DMS, the
Floats lithium content can be reduced with a low dense media pulp specific gravity, similar to
FT207. Test FT700 had the lowest loss of lithium for the standard test; the latter is of prime
importance as the lower grade sinks can be reprocessed as middlings in the flotation circuit.
The results from FT700 are listed in Table 13.8.

13.1.3.3 COREM Heavy Liquid Separation Test Work Program

To confirm the cut points for DMS heavy liquid separation tests were conducted on Whabouchi
feed, with a grade 1.77% Li2O. The tested samples were crushed into three size fraction
ranges, ‒12.5 mm +0.85 mm, ‒9.5 mm +0.85 mm and ‒6.3 mm +0.85 mm. The separation cut
points were 2.96 and 2.70. The results were presented in a Power Point presentation
“Résultats préliminaires – Caractérisation minéralogique des roches concassées du dépôt de
Whabouchi” or (Preliminary Results - Mineralogical Characterization of Crushed Rocks from
the Whabouchi Deposit). The results are presented in Table 13.9.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 122

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.8 – Results of Test FT700

Description: FT700 (3-Stage DMS) 1st Stage 2nd Stage 3rd Stage

*D50 (Specific Gravity): 2.940 2.855 2.700

Specific Gravity of Dense Media: 2.670 2.370 2.700

Medium Inlet Pressure (psi): 25 23 24

Back Pressure (mm): 200

Description

Weight Assay Distribution Li2O Si Al Li2O Si Al (kg) (%) (%) (%) (%) (%) (%) (%)

Sinks 1 33.23 9.67 5.88 33.0 11.80 33.5 8.7 14.2

Sinks 2 46.10 13.42 4.30 35.9 10.16 34.1 13.1 16.9

Sinks 3 68.37 19.91 1.85 37.6 8.03 21.8 20.3 19.8

Total Sinks 147.69 43.00 3.52 36.0 9.54 89.3 42.1 50.9

Floats 195.75 57.00 0.32 37.4 6.95 10.7 57.9 49.1

Calc Head 343.44 100.0 1.70 36.8 8.06 100.0 100.0 100.0

Calc Head (SFA) 1.65 33.7 7.81

Source: MS1467; Nemaska Lithium Inc., DMS Update – April 4, 2014

Table 13.9 – HLS Mineralogical Study

Size Fractions

Sinks — 2.96 Sinks — 2.70 Floats — 2.70 ‒0.85 mm

Weight (%)

Li2O (%)

Weight (%)

Li2O (%)

Weight (%)

Li2O (%)

Weight (%)

‒12.5 +0.85 mm 7.5 6.63 35.2 2.83 43.0 0.33 14.3

‒9.5 +0.85 mm 8.6 6.43 26.6 2.84 45.0 0.24 19.8

‒6.3 +0.85 mm 12.6 6.47 22.3 2.78 41.1 0.17 24.0

Source: COREM, Project T2081; Prelim results, Caractérisation minéralogique des roches
concassées du dépôt de Whabouchi, 02 Nov. 2017.
The fraction - 0.85 mm will go to fine ore processing. HLS Floats are final tailings and the - 12.5
mm has the highest tailings grade. The finest fraction (- 6.3 mm) has the lowest tailings
indicating that Spodumene was better liberated in the finer fractions. Since Nemaska elected
to use ‒9.5 mm, these results have been listed in Table 13.10.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 123

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.10 – Size Fraction ‒9.5 mm HLS

Stream ‒9.5 mm

Weight (%)

Li2O (%)

Li Rec. (%)

Sinks — 2.96 10.7 6.43 38.8

Sinks — 2.70 33.2 2.84 53.6

Float — 2.70 56.1 0.24 7.6

Feed (Calc.) 100.0 1.77 100.0

Source: COREM, Project T2081; Preliminary results, Caractérisation minéralogique des roches
concassées du dépôt de Whabouchi, November 2. 2017.

13.1.4 DERRICK TEST WORK PROGRAM – FINE SCREENING TESTING

Derrick received samples produced by SGS Minerals during the Pilot plant operation (2017).
200 kg drums of CrossFlow overflow and flotation feed were sent to Derrick's Buffalo facility.
Most of the tests were done on the flotation feed to split the material between coarse and fine
flotation.

13.1.4.1 Fine Screening at 212 microns

Tests #1 was performed on overflow from the Fine Muscovite Removal CrossFlow separator
and the screening yielded good results with near 94% efficiency.

Flotation feed screening using screen openings of 0.21 mm was done in Test #2, Test #3 and
Test #9. There is a significant quantity of fines in the oversize which may interfere with the
Hydro-Float separation. The results are listed in Table 13.11.
Table 13.11 – Fine Screening Tests at 212 Microns

Test Number

Test Screen Opening (mm)

Feed Rate (t/h)

Feed Solids (%)

Wash Water, (m3/h)

Cumulative Percentage at 0.212 mm Screening Efficiency Feed Oversize Undersize

1 0.21 34.0 8.75 0.0 5.58 47.0 1.64 93.9

2 0.21 67.3 35.4 0.0 39.9 69.1 6.94 80.4

3 0.21 67.3 35.4 28.4 39.9 74.1 7.45 83.6

9 0.21 68.0 25.3 28.4 40.3 82.9 9.63 87.2

Source: Derrick Corporation, Buffalo, NY. September 2017

13.1.4.2 Fine Screening at 250 Microns

The flotation feed fine screening tests using 0.25 mm screen openings were done in Test #4,
Test #5 and Test #10. Again, there is a significant quantity of fines in the oversize which may
interfere with the Hydro-Float separation. The results are listed in Table 13.12.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 124

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.12 – Fine Screening Tests at 250 Microns

Test Number

Test Screen Opening (mm)

Feed Rate (t/h)

Feed Solids (%)

Wash Water, (m3/h)

Cumulative Percentage at 0.212 mm Screening Efficiency Feed Oversize Undersize

4 0.25 59.5 35.4 0.0 39.9 70.8 8.68 81.0

5 0.25 59.5 35.4 28.4 39.9 78.2 9.85 84.9

10 0.25 68.0 25.3 28.4 40.3 85.1 11.94 87.3

Source: Derrick Corporation, Buffalo, NY. September 2017

13.1.4.3 Fine Screening at 300 Microns

The flotation feed fine screening tests using 0.30 mm screen openings were done in Test #6,
Test #7 and Test #8. Even with the larger screen openings, too many fines are present in the
oversize and may interfere with the Hydro-Float separation. The results are listed in Table
13.13.

Table 13.13 – Fine Screening Tests at 300 Microns

Test Number

Test Screen Opening (mm)

Feed Rate (t/h)

Feed Solids (%)

Wash Water, (m3/h)

Cumulative Percentage at 0.212 mm Screening Efficiency Feed Oversize Undersize

6 0.30 67.3 35.4 0.0 39.9 72.6 10.9 81.4

7 0.30 67.3 35.4 28.4 39.9 79.0 12.6 83.9

8 0.30 68.0 25.3 28.4 40.3 86.1 14.9 85.5

Source: Derrick Corporation, Buffalo, NY. September 2017

13.1.5 FLOTATION TEST WORK

13.1.5.1 SGS Flotation Pilot Plant Test Work – 2011

In January 2011, Nemaska contracted SGS to carry out a pilot plant and bench scale testing
program as part of a second phase of the Whabouchi lithium Project. The objectives of the
second phase were:

• To produce two (2) tonnes of spodumene concentrate with a grade of six percent (6%) or
higher for electrochemical test work;

• To confirm and optimize previous bench test work;

• To generate engineering data for concentrator design.

Findings from the pilot plant program were reported by SGS in the report “Project 12486-003”,
April2, 2012, and are summarized in the following Sections.
Pilot plant flotation tests were performed on four (4) composites, the flotation outcrop
(blasted composite), flotation mine representative composite, combined DMS middlings and
undersize

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 125

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

fractions from outcrop and finally combined DMS middlings and undersize fractions from mine
representative composite. Mine representative composite was composited of outcrop material
and drill core assay reject samples. The detailed description of these composites is given in the
previously mentioned SGS report.

Material from the four (4) composites were processed through the flotation pilot plant in a
sequential manner, 21 pilot plant tests were performed.

Various flow sheet configurations were tested in the pilot plant campaigns. The objectives
were to find the best and optimal operating conditions for spodumene separation and
recovery.

In total, more than 40 tonnes of material were processed through this pilot plant. The final
flow sheet of the last run of the pilot plant PP21 is shown in Figure 13.8. Pilot plant test PP21,
yielded the most efficient processing results.

The final flow sheet consists of the following circuits:

• Grinding and screening;

• Primary de-sliming and mica flotation;

• Dewatering, scrubbing and secondary desliming;

• Spodumene flotation;

• Magnetic separation circuit.

The crushed composite feed was ground using a rod mill and screened at 300 µm. The screen
oversize (+300 µm), was sent back to rod mill, while the screen undersize (-300 µm) was
subjected to primary de-sliming in a cyclone where the overflow exited as slimes. The primary
de-sliming cyclone underflow was conditioned with AERO 3030C or Armac collector prior to
subjecting to mica rougher flotation stage followed by scavenger stage.

The rougher and scavenger concentrates were combined and transferred to a cleaner stage.
Mica cleaner concentrate was collected in 200 L plastic drums, while mica cleaner tailings were
dewatered in 2-stage cycloning and then underwent a scrubbing stage where dispersant D618
and NaOH were added. The scrubbed slurry was then subjected to a secondary de-sliming
stage before pH adjustment and high-density conditioning with spodumene collector (LR19).
The spodumene rougher flotation concentrate was subjected to 2-stage cleaning while
spodumene rougher tailings were sent to final tailings.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 126

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.8 – Final Flotation Pilot Plant Flow Sheet

Source: SGS Mineral Services; Project 12486-003 – Final Report, April 2,, 2012

The first and second cleaner tailings were recycled back and combined with mica flotation
tailings. The final spodumene concentrate was conditioned with acid before magnetic
separation to remove iron impurities. Pilot Plant test PP21 yielded the most significant results.
These metallurgical results for test PP21 are summarized and presented in Table 13.14.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 127

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.14 – Optimal Pilot Plant Performance for PP21 Test

Composite PP No Products

Mass Assay (Adj) Distribution

(%) (% Li) (% Li2O) (% Li)

DM S

Min e Representative
Compos ite

PP21

Feed 100 0.88 1.85 100

Cyclone O/F 2.65 0.66 1.42 2.03

Mica 1st Clnr Conc 9.2 0.34 0.73 3.63

2nd Slimes 2.89 0.9 1.94 3

Spod Rghr Tail 63. 0.19 0.4 13.7

Spod Mag Conc 0.18 0.23 0.5 0.05

Spod Final (Non-Mag) Conc 22.1 3.02 6.5 77.5

Source: SGS Mineral Services; Project 12486-003 – Final Report, April 2, 2012

The conclusions from these pilot plant results were that using flotation a spodumene final
concentrate grade of 6.0% Li2O or higher with more than 77% lithium recovery representing
22.1% weight could be obtained consistently.

These results show also that the lithium losses depend on the nature of the feed composite.
For the DMS-Mine representative middling composite (PP21), the majority of lithium losses
occurs in the spodumene tailings (13.7%), followed by slime removal (5.0%), mica concentrate
(3.6%) and spodumene magnetics (0.05%) for a total of 22.4%. According to the results, the
majority of the losses in the rougher tailings occur in the 100 meshes fraction. Coarse grain
spodumene can be difficult to float thus the grinding size should be controlled to keep the K80
of the flotation feed to about 200 µm.

These results highlight the importance of mica flotation circuit ahead of spodumene flotation;
by eliminating the mica flotation step, significant increase in muscovite grade was observed.
The lithium oxide concentration in spodumene concentrate increased from 5.6% in PP19 as
compared to 6.5% in PP21, which confirmed that removing mica ahead of spodumene
flotation helped to increase final spodumene concentrate grade.

About three (3) tonnes of concentrate grading 6.0% Li2O was generated by combining
concentrates from pilot plant campaigns PP12 to PP21.

Low magnetic intensity separation (about 800 Gauss) was used to separate iron contaminant
particles from the flotation concentrate. The iron grade of the lithium concentrate was about
2.11% Fe2O3.

13.1.5.2 SGS Minerals Test Work Program – 2017

SGS Minerals Lakefield received an estimated 500 tonnes of pre-screened fines (< 850 µm)
from the Nemaska Whabouchi mini-DMS operation. The aim of this test work program was to
produce
Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 128

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

flotation concentrate for the electrochemical demonstration plant (P1P) at Shawinigan. Half of
this material was processed using the flow sheet as shown in Figure 13.9.

Muscovite (Mica) was removed using Hydraulic separation (CrossFlow) and was screened. The
screen oversize was considered mica waste and the undersize was reground to < 430 µm,
deslimed and processed in wet magnetic separation. Before flotation, an attrition scrubbing
stage was done on the material before a final desliming step to prepare the particles for the
conditioning stage. The flotation was split in coarse hydroflotation (for + 0.18 – 0.43 mm) and
fine conventional flotation (on – 0.18 mm) with separate conditioning stages for each.

a. Pilot Plant Test #1 to Test #17

This period was mainly commissioning of the equipment and process optimisation. After a few
extended continuous runs, it became clear that:

CrossFlow separator feed had to be introduced with less energy; Magnetic separation was
effective in removing residual iron rich particles; Flotation conditioning is critical; To
achieve no fines in the coarse after fine screening was not possible; HydroFloat separation
was going to be very sensitive to feed variations; Vacuum filtration of the concentrate
produced a dry product by touch.

CrossFlow separation required a high percent solids and laminar flow into the separator. This
was accomplished with the introduction of cyclones prior to CrossFlow separators. It became
clear that the second separator was superfluous.

The magnetic separator mass pull is very low at 2.2%. However, the important aspect that
needs to be monitored in the magnetic separation stage is to maximize rejection while
preventing spodumene loss. Operating conditions can be adjusted to meet this criterion.

Hydro-Float separation was not successful at SGS, probably due to screening fines in the
screen oversize. The finer material soaks up the reagents disproportionally and most finer
material are unselectively removed during Hydro-Float separation resulting in poor grade and
poor recovery.

The lithium losses in slimes are about 3.5%.

b. Pilot Plant Test #18 to Test #23

The Hydro-Float was replaced by mechanical cells and later a flotation column for cleaning.
Very high grades up to 7.1% Li2O were achieved using the flotation columns, however, the
recovery grades were still low.
Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 129

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.9 – Pilot Plant Flow Sheet Test #1 to Test #17

Source: SGS Minerals Canada Inc.; Test Laboratory, Lakefield, ON. December 2017.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 130

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

c. Pilot Plant Test #24 and Test #25

A hydraulic separator was introduced to remove the fines from the rougher flotation tailings of
0.86% Li2O. The hydraulic separator underflow (1.16% Li2O) was re-floated using the
HydroFloat separator after reconditioning. The Hydro-Float concentrate grade was only 3.55%
Li2O. The upgrading ratio of the Hydro-Float unit is limited. To produce an acceptable
concentrate, it must be provided with a feed of sufficient grade. This test provided an
upgrading ratio on 3 to 1.

d. Pilot Plant Test #26 and Test #30

These tests involve the successful upgrading of earlier subpar concentrate grade to above 6%
Li₂O. The old concentrate was re-conditioned and re-floated using mechanical cells.

e. Pilot Plant Test #31 and Test #40

These tests involve the use of mica flotation between grinding and magnetic separation and
spodumene flotation using mechanical cells. A greater than 6% Li₂O concentrate was produced
with a recovery above 80%. For these tests, all the ore was ground to less than 300 microns.
The Lithium losses due to de-sliming was 6.8% or about double the amount compared when
ground to 500 microns.
13.1.5.3 COREM Flotation Test Work – 2017

a. Bench Scale Flotation Test Work

COREM performed bench scale flotation test with the aim to determine the most influential
flotation parameter. Four main parameters are collector dosage rate, flotation alkalinity,
conditioning pulp density, flotation time. The results are presented in Table 13.15.

The results indicated that a minimum five (5) minutes of conditioning time is required. The
other parameters are not statistically significant due to interactions. However, Test 22
delivered the best results and these parameters are, therefore, assumed to be superior.

b. Pilot Plant Flotation Test Work

The pilot scale circuit was based on the proposed flow sheet and included: magnetic
separation, attrition scrubbing, de-sliming, conditioning and flotation. The flotation steps
consist of rougher, scavenger and cleaner flotation using 3-inch flotation columns. The
flotation feed sample was pre-screened to less than 212 microns.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 131

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.15 – Main Flotation Parameters Test

Test

Collector (kg/t)

pH

Cond. Time (min)

Cond. Solids (%)

Weight Rec.1

Li2O (%)

Li2O Rec.2

18 3.6 9 5.0 50 13.3 6.01 72.8

19 2.4 9 5.0 50 19.7 5.34 92.4

20 1.2 9 5.0 50 9.4 5.77 50.7

21 2.4 10 5.0 50 17.3 5.45 86.9

22 2.4 8 5.0 50 16.9 5.96 91.8

23 2.4 9 5.0 70 18.3 5.42 90.4

24 2.4 9 5.0 60 19.3 5.34 92.2

25 2.4 9 5.0 40 17.2 5.60 87.1

26 2.4 9 2.5 50 13.4 5.70 70.7

27 2.4 9 1.0 50 8.6 5.62 43.5

28 2.4 9 2.5 70 11.5 5.81 56.6

29 2.4 9 1.0 70 9.9 5.92 55.8

Source: COREM, Preliminary Report, Project T2181; 3 inches column flotation optimization for
Whabouchi spodumene ore, August 2,2017. 1 Weight recovery = concentrate weight /
flotation feed weight × 100%. 2. Li2O recovery = weight of Li2O in concentrate / weight of Li2O
in flotation feed × 100%.

Medium Intensity Magnetic Separation (MIMS) was performed on dry ore and was used to
remove the residual ferrosilicon found in the provided sample which comes from the dense
media separator circuit (mini-DMS operation 2017). A final magnetic separation was
performed on the final concentrate.

In total three (3) pilot plant flotation tests were conducted using the flow sheet shown in
Figure 13.10.

Table 13.16 is the pilot plant results. COREM endeavored to simulate the proposed fine
flotation part of the flow sheet. The aim was to confirm the column flotation performance in
the flow sheet. The rougher flotation time estimate and recoveries are listed in Table 13.16.
The performance of the test done at COREM did not meet expectations. However, the
flotation feed to these tests was less then 1% Li2O which is an important factor in the flotation
performance. Another important observation is that the reagent scheme differed significantly
from the optimized parameters that were recently developed at SGS Lakefield and used by
Eriez in the next Section.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 132

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.10 – General Process Flow Sheet for COREM Column Flotation Pilot Plant

Source: COREM, Preliminary Report, Project T2181; 3 inches column flotation optimization for
Whabouchi spodumene ore, 02 August 2017.

Table 13.16 – Pilot Plant Flotation Test

Pilot Plant Test

Collector (kg/t)

pH

Cond. Time (min)

Rougher Float. Time (min)

Rougher Weight Rec.

Cleaner Li2O (%)

Cleaner Li2O Rec.

1 4.8 8 25 17.5 10.7 5.33 59.3

2 4.8 8 25 17.1 8.5 5.16 47.1

3 4.8 8 12 7.5 8.3 4.77 27.9

Source: COREM, Preliminary Report, Project T2181; 3 inches column flotation optimization for
Whabouchi spodumene ore, August 2, 2017.

13.1.5.4 Eriez Flotation Test Work – 2017

Eriez Flotation Division (EFD) was provided with one (1) bulk bag of minus 850 microns (“μm”)
low grade ore from the SGS test work program 2017, and one (1) bulk bag of DMS circuit
middlings (D80 = 682 μm) from Nemaska to study the laboratory‐scale split feed flotation
response. This test work was a continuation of recent pilot testing efforts performed at SGS-
Lakefield. The patented Hydro-Float technology and flotation columns, inclusive of proprietary
Eriez Cavitation‐Tube

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 133

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

sparging technology, were used to treat 497×212 μm and 212×27 μm size fractions,
respectively. This split feed flotation approach provides the maximum separation efficiency.

a. Hydro-Float Tests

Coarse flotation tests were conducted on both an independent low-grade ore, as-received
from SGS, and a blend of DMS circuit middlings and fresh flotation feed. During treatment of
the coarse size fraction using Eriez Hydro-Float fluidized bed flotation, optimal upgrade ratios
of approximately 1.90-2.0 were achieved at Li₂O recoveries of 92-95%. A 5.8-6.0% Li₂O product
was yielded at Li2O recoveries of nearly 95-97% during treatment of a 3.07% Li₂O feed.

The Hydro-Float feed was nearly 40% passing 300 μm. Size-by-size assays of the HydroFloat
overflow indicate Li₂O grades of the plus 300 μm particle size fractions are greater than 6.3%.
The lower grade concentrates are within the minus 300 μm size fractions. Although a coarse
6% Li₂O Hydro-Float concentrate is achievable without classification of the overflow, it is
recommended that the circuit be designed such that the Hydro-Float concentrate can be
scalped to remove fines floated unselectively and re-process them in a fine flotation circuit to
improve global spodumene recovery. This is especially important if the feed grade decreases
below 2.4% Li₂O, as demonstrated in preliminary coarse flotation testing.

b. Column Flotation Tests

Sixteen (16) rougher column flotation tests were conducted on a de-slimed 212 × 27 μm blend
of the SGS Lakefield and DMS samples at varying operating conditions. Optimal upgrade ratios
of approximately 2.4 were achieved at Li₂O recoveries of 88.5-92.5%, as a concentrate grade of
over 6.1% was achieved in rougher column flotation. A concentrate grade of 6.6% was realized
at 88.2% Li2O recovery and 34.4% concentrate mass yield in a rougher-cleaner open circuit.

Such results were ascertained following a 10-minute scrubbing period using 104 g/t NaOH pH
modifier and 250 g/t soltisperse dispersant at 65% solids, by weight. In addition to scrubbing,
the rougher and cleaner flotation feed were conditioned for 15 and 2 minutes, respectively,
using a cumulative 146 g/t H₂SO₄ (return slurry to neutral pH), 32 g/t Na₂SiO₃, 350 g/t FA-2,
and 150 g/t TP-100. Throughout rougher-cleaner column flotation testing, maximum rougher
and cleaner carrying capacities of approximately 4.1 tph/m² and 3.9 tph/m², respectively, were
achieved at a total circuit retention time of nearly 14.7 minutes (6.3 minutes in rougher and
8.4 minutes in cleaner).

The use of wash water in column flotation significantly improved the spodumene concentrate
grade and flotation upgrade ratios. This is because wash water efficiently rids the froth of
entrained gangue minerals such as quartz, mica and feldspar etc., pushing those minerals back
into the pulp phase. The use of a minimal sodium silicate dosage (32 g/t) also increased
flotation selectivity throughout spodumene flotation testing. However, when added in

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 134
August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

excessive quantities, sodium silicate can render the froth brittle. As a result, careful monitoring
of its addition is necessary to optimize the flotation performance.

13.1.6 SETTLING, FILTRATION AND FREEZING TESTS

Various design tests have been done to size dewatering equipment and evaluate freezing of
the final concentrate. Settling, filtration and freezing test work was done, which can be used
for the sizing of concentrate and tailings thickeners and filters.

13.1.6.1 Settling Test Work

SGS performed static settling test to determine the optimal feed densities and type of
flocculant. Thereafter they performed dynamic thickening testing. The results for thickening
tests are listed in Table 13.17. The most important tests involved spodumene concentrate and
final tailings. The thickening hydraulic test data was used by suppliers for thickener sizing. The
dynamic thickening obtained results with a spodumene concentrate thickener underflow
between 61 and 63 weight percent solids. This was as expected. The final tailings underflow
density was 61%. Both test samples had low turbidity overflows with total suspended solids
(“TSS”) of ten (10) ppm.

Table 13.17 – Dynamic Thickening Test Work Results

Stream Description

Flocculant Type

Flocculant addition rate (kg/t)

Feed Pulp

(% w/w)

Underflow Pulp (% w/w)

Dynamic Settling Rate (t/m2/h)

Spodumene Concentrate

Ciba Magnafloc 10

0.010 19 61 to 63 0.62

Combined Final Tailings

Ciba Magnafloc 10

0.028 17.8 61 0.5

Source: SGS Mineral Services; Project 12486-005A – Final Report, June 5, 2012.

13.1.6.2 Filtration Test Work

a. SGS Filtration Test Work on the Spodumene Concentrate and the Final Tailings

The vacuum filtration test had good results for the spodumene concentrate producing a cake
of 8.3% moisture by weight at minus 0.4 g bar vacuum, while the lower vacuum of minus 0.7 g
bar yielded a higher cake moisture content of 12.2%. The final tailings did not have similar
results and ended up with wet cakes of 13.9% and 12.8% at minus 0.4 g bar and minus 0.7 g
bar vacuum levels respectively.

The pressure filtration test produced good results for the spodumene concentrate with a cake
of 7.7% moisture by weight at a pressure of 4.1 bar, while the higher pressure 6.9 bar yielded a
higher cake moisture content of 9.1%. The final tailings had similar results and ended up with
dry cakes of 6.2% and 6.6% at 4.1 bar and 6.9 bar respectively as shown in Table 13.18.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 135

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.18 – Filtration Test Work Results

Stream Description

Vacuum Filtration Pressure Filtration

Throughput rate (kg/m2/h)

Vacuum Level (barg)

Cake Moisture (% w/w)

Throughput rate (kg/m2/h)

Pressure Level (bar)

Cake Moisture (% w/w)

Spodumene Concentrate

497 –0.4 8.3 561 4.1 7.7

463 –0.7 12.2 541 6.9 9.1

Combined Final Tailings

425 –0.4 13.9 486 4.1 6.2

486 –0.7 12.8 670 6.9 6.6

Source: SGS Mineral Services; Project 12486-005A – Final Report, June 5, 2012.xc

SGS reports that the spodumene concentrate filtration test results were as expected and were
considered normal behaviour for fast settling coarse material.

b. Bokela Tailings Filtration Testing

Bokela performed six (6) filtration tests. Three (3) different filtration rates and two (2) different
air flows were tested at two (2) differential pressures. The tailings were of fast sedimentation
and high filtration speed, a pan filter with a filter area of 25 m² was recommended.

The pan filter is able to handle throughput of 37.8 t/h.

In Table 13.19, two (2) layout cases are shown for a higher moisture content (7,200 m³/h
vacuum demand) and for a lower moisture content (13,000 m³/h).

Table 13.19 – Tailing Filtration Test at 200 Micron

Item

Filter Area (m2)

Filter Rate (dry t/h)

Diff. Pressure (bar)

Airflow

(m3/h)

Expected Moisture (%)

Guaranteed Moisture (%)

1 25 26.5 0.55 to 0.6. 13,000 13.5 to 14.0

2 25 37.8 0.6 13,000 14.0 15.0

3 25 47.2 0.60 to 0.65 13,000 14.0 to 14.5

4 25 26.5 0.35 to 0.40 7,200 14.5 to 15.2

5 25 37.8 0.35 to 0.45 7,200 15.2 16.2

6 25 47.2 0.40 to 0.45 7,200 15.2 to 15.7

Source: BOKELA Ingenieurgesellschaft fuer Mechanische Verfahrenstechnik mbH Karlruhe,

Germany. November 2017

13.1.6.3 Freeze Test Work

SGS performed two (2) freezing tests on spodumene fine concentrates containing 3% and 5%
moisture. SGS placed 750 grams samples on trays and exposed those -18°C temperature for 24
hours.
Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 136

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Neither of the two (2) samples indicated major freezing problems as none of the samples froze
as one solid block. The frozen 3% moisture sample contained many very small fragile lumps.
The frozen 5% moisture sample had larger lumps, but these were also fragile crumbling on
slight contact.

13.1.7 DMS CONCENTRATE DRYING TESTING

ThermoPower received Nemaska Whabouchi DMS concentrate to perform drying testing. The
aim was to find out the temperature at which the material would be free flowing.

13.1.7.1 Complete Drying of DMS Concentrate

To determine the residence time for completely drying a sample of DMS concentrate in an
indirectly heated rotary tube furnace. It took an average of 7 minutes and 2 seconds for the
product to reach a temperature of 110°C and thus to complete drying. 110°C is well above the
boiling temperature of water (≈ 94°C) of the test location.

13.1.7.2 Drying DMS Concentrate at 80°C

The average residual moisture was 3.4% after drying. Figure 13.11 shows product that was
cooled from 80°C to 60°C and then inspected for moisture and adhesion properties before
being dried in the Convection Oven. The material was mixed and inspected by hand in the
drying tray and can be seen to still stick to contact surfaces (Figure 13.11).

13.1.7.3 Drying DMS Concentrate at 85°C

The average residual moisture was 1.3% after drying at 85°C. This material is reasonably free
flowing.

13.1.8 MAGNETIC SEPARATION TEST

13.1.8.1 Eriez Test Work Program Magnetic Separation

Magnetic separation was performed at Eriez. Eriez Central Test Laboratory received two (2)
samples of spodumene concentrate containing hornblende from Nemaska for testing. Dry
magnetic separation on Sample #1 was less than 3 mm, and wet magnetic separation on
Sample #2 less than 0.85 mm. The objective of the test work was to remove iron impurities
from the Spodumene concentrate through both dry and wet magnetic separation processes.
Later, another sample was sent to Eriez for Dry Magnetic Separation test optimization.

a. Dry Magnetic Separation

The coarser sample 3 mm × 0 was processed using dry magnetic separation equipment, while
the other underwent wet magnetic separation.

The dry 3 mm × 0 feed was processed on a Ceramic Magnetic Drum Separator to remove any
residual ferromagnetic debris present. A feed rate of 2 tonnes per hour per foot (t/h•ft‒1) was
used. This produced a Reject #1 and non-magnetic product #1.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 137

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.11 – ThermoPower Drying Test #2, 80° Discharge Temperature

Source: ThermoPower Furnaces, Olifantsfontein, Gauteng, SA. June 2017

The non-magnetic #1 fraction was reprocessed on the Salient Pole-2 Rare Earth Magnetic
Drum Separator The drum surface speed was set to a feed rate of approximately 2 t/h•ft‒1,
producing Reject #2 and corresponding non-magnetic product.

Again, the non-magnetic fraction was reprocessed on a Rare Earth Magnetic Roll Separator.
This part of the sample was processed in three steps; reprocessing the non-magnetic fraction
each time. For the first step, the feed rate was set to 100 pounds per hour per inch of feed
width (lb/h•in), the other two (2) steps were processed at 75 lb/h•in. The sample fractions of
dry magnetic separation were weighed, and the percentages are presented in Table 13.20.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 138

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.20 –Magnetic Separation Test

Stream
Dry Magnetic weight (%)

Wet Magnetic 1.0 T- weight (%)

Wet Magnetic 1.3 T- weight (%)

Feed 100.0 100.0 100.0

Reject #1 0.8 7.0 7.0

Non-magnetic #1 99.2 93.0 93.0

Reject #2 13.0 0.3 0.3

Non-magnetic #2 86.2 92.7 92.7

Reject #3, #4 & #5 15.8 32.3 37.9

Final non-magnetic 70.4 60.4 54.8

Source: Eriez Magnetics Central Test Laboratory Spodumene & Hornblende – Report Project
MTR# 17-033; February 23, 2017.

b. Wet Magnetic Separation

The 0.85 mm × 0 sample was first processed on the Low-Intensity Magnetic Separator (LIMS)
to remove ferromagnetic debris. The sample was mixed with water to produce a slurry of
approximately 20 percent solids and processed through the LIMS with a nominal field intensity
of 950 Gauss.

The non-magnetic fraction from the LIMS was reprocessed through a Salient Pole Rare Earth
Magnetic Wet Drum Separator (SP WDS) to remove additional weakly-ferromagnetic particles.
The magnetic fraction was dried for mass yields.

Two (2) batches of the non-magnetic fraction from the SP WDS were reprocessed on the Wet
High Intensity Magnetic Separator, (WHIMS). One batch was processed at a background field
intensity of 1.0 Tesla and the other was processed at 1.3 Tesla. For both batches, a 2-mm rod
matrix was used to collect the magnetic fraction, and low intensity jigging was used to agitate
the slurry during processing.

c. Dry Magnetic Separation (second Eriez Test)

A new sample of DMS concentrate was provided to Eriez in early 2018. The objective was to
optimize the dry magnetic separation for the coarse DMS material. Four (4) tests were done.
The first three were open circuit test where a non-magnetic concentrate, a magnetic reject
and a middling product were recovered. For the last test, the middling material was
reprocessed and a non magnetic concentrate was recovered and added to the first stage non
magnetic product. The second stage magnetic material was rejected with the magnetic tailings
of the first stage.

After analyzing the results and discussing with Eriez, it was determined that the best
separation arrangement is to perform a first separation stage that produces a final non
Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 139

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

magnetic concentrate and an intermediate product that will be reprocessed in a second stage.
The second stage non magnetic material is combined with the first stage concentrate and the
magnetic product is the final tailings.

Table 13.21 –Dry Magnetic Separation Test (Second Eriez Test)

Test 1 Test 2 Test 3 Test 4

Wt%

Li₂0 (%)

Rec (%)

Wt%

Li₂0 (%)

Rec (%)

Wt%

Li₂0 (%)

Rec (%)

Wt%

Li₂0 (%)

Rec (%)

Feed 100.0 4.58 100 100.0 4.76 100 100.1 3.92 100 100.0 5.69 100

Concentrate 70.1 5.96 91.3 71.7 5.98 90.0 63.8 5.91 96.2 94.1 5.98 98.9

Middlings 8.0 4.11 7.2 7.5 5.42 8.5 9.3 1.05 2.5 1.6 3.20 0.9

Tailings (magne tics)

21.9 0.32 1.5 20.8 0.34 1.5 27.0 0.19 1.3 4.3 0.28 0.2

Stage 2 magnetics 0.78 0.82 0.1

Stage 2 concentrate 0.79 5.68 0.79 Combined concentrate 94.9 5.98 99.7

13.1.8.2 PhySep Dry Magnetic Separation Testing

Four tests were conducted during on August 16, 2017. An impact of the feed temperature, and
feed rate on the separator performance were evaluated. To prorate the results, the laboratory
vibrating feeder feed rate was then converted to equivalent feed rate for the industrial size
unit based on the lab unit feed rate.

Test #1 was conducted at the ambient temperature of 27°C (in the lab), and Test #2, Test #3,
and Test #4 modelled the actual temperatures of the dryer discharge around 100°C. For three
(3) latter tests the feed material was preheated to about 120-130°C in the oven, and then fed
to the feed hopper of the magnetic separator test unit.

The non-magnetics of the Tests # 2, 3, and 4 were collected and re-heated to about 100°C at
magnetic separator feed. After the third (and final) pass magnetic reject and non-magnetics
were cooled, weighed and packaged for assaying elsewhere.

The least rejects occurred in Test #3 with a feed rate of five (5) t/h and a weight recovery of
93.9%. There are visually no dark minerals in the non-magnetic stream (Figure 13.12).

The assay results showed that the higher temperature of material did not negatively impact
the separation performance. However, suppliers indicated it could deteriorate the magnets
properties and some cooling should be applied to operate under 85°C. The higher processing
rate resulted in a lower separation performance (less contaminant rejection), but the
separation does not need to be perfect as long as the majority of hornblende is removed to
prevent issues in the roasting process. Lithium losses must be minimized.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 140

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.12 –Hot Dry Magnetic Separation Test #3, 99°C Feed Temperature

Source: DRA pictures taken at PhySep Components and Service, Jacksonville, FL. August 2017

The technical results are listed in Table 13.22.

Table 13.22 – Dry Magnetic Separation Results

Stream Test #1 Test #2 Test #3 Test #4

Feed temperature, ⁰C 27 100 99 99

Feed rate, t/h 2.5 2.5 5.0 5.0

1st pass mag rejects 8.1% 6.6% 4.2% 6.9%

2nd pass mag rejects 1.4% 1.1% 1.0% 2.5%

3rd pass mag rejects 0.6% 0.7% 0.8% 2.0%

Total magnet rejects 10.1% 8.5% 6.1% 11.4%

Final non-magnetics 89.9% 91.5% 93.9% 88.6%

Source: PhySep Components and Service, Jacksonville, FL. August 2017

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 141

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

13.2 Electrochemical Process Test Work

This Section of the Report summarizes the process and electrochemical testing conducted on
the Nemaska Whabouchi spodumene concentrate. This summary is based on reports provided
by multiple independent testing laboratories, pilot facilities, research centers, and specialized
equipment manufacturers.

Test work was performed to fix or optimize the process flow sheet for the production of high-
quality battery-grade Lithium Hydroxide Monohydrate (“LHM”) and provide engineering data
for all the major equipment selection and sizing, including:

• Concentrate Crushing and Screening (now performed at Whabouchi concentrator);

• Calcination;

• Acid Bake;

• Leaching;

• Impurity Removal;

• Electromembrane Process;

• LHM Crystallization;

• LHM Drying;
• Acid Concentration.

The following Sections give an overview of the test work performed. They are organized by
plant area when possible.

13.2.1 CRUSHING AND SCREENING TEST WORK – 2018

Concentrate produced at Whabouchi is a mix of DMS (P100 9 mm) and Flotation (P100 850
microns) concentrates. Based on Calcination requirements, the DMS spodumene concentrate
must be crushed to less than 1 mm.

Spodumene crushing and screening test work was conducted by several suppliers in 2018 to
confirm technology selection and final sizing of equipment.

SGS Canada was first contracted to evaluate the grindability characteristics of the concentrate
(Bond, Rod Mill and Abrasion Indices), as well as laboratory scale investigation of HPGR
technology using a LABWAL unit.

Based on SGS results, HPGR was selected as the preferred technology for crushing the DMS
concentrate. More samples were sent to technology suppliers, namely Metso, Weir and TKIS
for pilot test work to determine the HPGR sizing parameters, as well as evaluating rolls wear
life.

All tests were conclusive and confirmed that HGPR is an appropriate technology for this
application.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 142

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Screening test work was also performed at several technology providers to confirm that
screening dry concentrate at 1 mm is feasible and to evaluate the impact of humidity on
screening performances.

Note that following these tests, the spodumene crushing sector was transferred to the
Whabouchi site.

13.2.2 CALCINATION AND ACID BAKE TEST WORK – FALL 2011

Spodumene calcination and acid bake test work was conducted by Feeco in the fall of 2011.

13.2.2.1 Sample Preparation

Fine and coarse concentrates from the original concentration flow sheet were shipped from
SGS to Feeco in Green Bay, Wisconsin. The coarse material was generated in the DMS process
while the fine material was produced through flotation.
Two (2) blends of concentrates were prepared using the as-received coarse material and the
dried fine material. Approximately 1.7 tonnes of a first blend were prepared, containing 75%
fines and 25% coarse. A second blend (1.5 tonnes) was prepared, containing 50% fines and
50% coarse.

13.2.2.2 Spodumene Conversion (Calcination)

Calcination converts the spodumene's crystalline structure from alpha to beta. The conversion
occurs at a temperature of about 1,025 ºC. The beta phase is reactive with sulfuric acid and
produces lithium sulfate, which is amenable to leaching while the alpha crystalline structure
remains unreactive.

The test work was carried out over three (3) days using a 760 mm (30") diameter by 6 m (20')
long parallel flow gas-fired rotary kiln. The feed rate was maintained at 50 kg/h and a
thermocouple was inserted in the discharge end hood to monitor the product temperature.

13.2.2.3 Acid Baking (Sulphation)

A paddle mixer was used to blend the beta-spodumene with 93% sulfuric acid. The acidified
material was returned to the kiln after the kiln had cooled to about 200ºC. The kiln feed rate
was between 135 and 180 kg/h. The retention time for the acid bake was 32 minutes.

13.2.2.4 Additional Information

The following observations were made by Nemaska while attending the test work program.
This information was not published by Feeco.

Coarse alpha-spodumene concentrate was easily converted to beta-spodumene but required a

longer residence time in the kiln.

The presence of iron in the concentrate affected the conversion temperature. A higher
temperature could be reached without causing the formation of glass beads when the iron was
removed from the concentrate before conversion.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 143

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Sulfuric acid was used with approximately 30% stoichiometric excess.

The acid bake kiln temperature of 200ºC was thought to be sufficient for the acid bake step as
it was expected that the exothermic reaction would raise the material temperature to 250ºC.

13.2.3 CALCINATION AND ACID BAKE TEST WORK – WINTER 2013

A second round of spodumene conversion and acid bake test work was performed by SGS in
the winter of 2013. The objectives of these tests were to produce a water leach residue from
the two (2) types of concentrate (DMS concentrate and flotation concentrate) and to perform
batch calcination optimization tests.

13.2.3.1 Sample Preparation

DMS (coarse) and flotation (fine) concentrates were used to perform the test work. The
different concentrates were tested separately to assess the impact of particle size on
conversion and acid bake.

13.2.3.2 Spodumene Conversion

Twenty-nine (29) samples varying in weight from one to two kilograms (1 to 2 kg) were tested
in a muffle furnace at 1,050°C. Average sample SG was 3.15. The first 14 samples were
flotation concentrate and the subsequent 15 samples were DMS concentrate. Retention times
were varied between 20 minutes and 165 minutes. Full conversion was deemed to have been
achieved when the sample specific gravity dropped below 2.50.

Ten (10) additional optimization tests were performed using a 15.2 cm diameter by 193.5 cm
long parallel flow rotary kiln with thermocouples inserted in the sample bed. Seven (7) tests
were performed on flotation concentrate and three (3) on the DMS concentrate. Sample size
for each test was three (3) kg. Retention times were varied between 30 minutes and 120
minutes and reaction temperatures were varied between 950°C and 1,050°C.

13.2.3.3 Acid Baking

The same muffle furnace was used to perform the acid baking step. Samples of one to two
kilograms (1 to 2 kg) were placed in the furnace along with 30% excess sulfuric acid for 30
minutes after the temperature reached 250°C. Temperature was monitored by a
thermocouple placed in the center of the feed.

Further, newly transformed beta-spodumene concentrate from ten (10) additional conversion
optimization tests were acid baked using 30% excess sulfuric acid for 30 minutes at a
temperature of 250°C.

13.2.3.4 Concentrate Leaching

Two (2) water leaching tests were performed. One (1) on the calcined DMS concentrate (a
total of 32.31 kg) and one (1) on the calcined flotation concentrate (a total of 21.50 kg). Assays
results of

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 144

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

the concentrate leach residues determined that a lithium extraction of 98.6% was achieved for
the calcined DMS concentrate and a lithium extraction of 95% was achieved for the calcined
flotation concentrate.

13.2.3.5 Additional Information

The following observations were made following the described test work:

• Using a rotary kiln at an operating temperature of 1,050°C with a retention time of 30

minutes was sufficient to achieve complete phase conversion.

• Extractions of 98.6% lithium were achieved on the DMS concentrate and 95.0% on the
flotation concentrate.

• A conversion temperature of 950°C is too low as only 69.4% lithium extraction was achieved.

• Both fine and coarse particles become very friable during conversion, with coarser particles
becoming much finer.

13.2.4 CALCINATION TEST WORK – 2017

Calcination tests were performed with two (2) reputable flash calciner system suppliers (FLS
and TKIS). Different types of spodumene concentrate were tested:

• DMS spodumene concentrate prepared from Whabouchi ore by Nemaska's pilot unit in
2017.

• Float spodumene concentrate prepared from Whabouchi ore by SGS in 2012

• Market concentrate, a concentrate from the market.

13.2.4.1 FLSmidth – Flash Calcination Test Work - May 2017

In May 2017, FLSmidth (“FLS”) performed a 4-day test program (approximately 15 tonnes)
using a pilot scale gas suspension flash calciner at their facility. The objectives were to confirm
that suitable conversion level could be achieved, to determine the potential for coating
formation in the flash calciner and to generate data required for the design of a commercial
flash calciner system.

Test work was performed with Whabouchi DMS or DMS/Float mix concentrate, as well as
purchased market concentrate.

Purchased market concentrate was used to debug the pilot unit, minimize the amount of
Whabouchi concentrate that needed to be used, and provide a comparison point. Due to the
restricted amount of flotation concentrate available, certain tests were only performed with
DMS feed.

The market concentrate and the DMS samples were crushed to less than one (1) mm using a
closedcircuit roll crusher. The float sample was processed in flash dryer and then blended with
crushed DMS using a 1:1 ratio in a V blender.

The conversion test to beta-spodumene was performed over a period of 82 hours. The
temperature was maintained around 1,100°C. The product particle specific gravity was used to
evaluate
Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 145

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

conversion performance. This was then later followed by controlled acid bake and leach tests
to determine the lithium and impurity level extraction rates.

The test work highlighted the importance of residence time and temperature. Good
spodumene conversion of over 95%, as determined by Nemaska’s standardized lithium
extraction test, could be achieved with both the Whabouchi DMS and the Whabouchi
DMS/float mix concentrate under the correct operating conditions.

Risks of accretion and wear remain present, and learnings from the test work were included in
the final equipment design.

In September/October 2017, a more substantial test campaign took place, involving

approximatively 55 tonnes of material (mostly DMS). This material was milled to < 1 mm
calciner feed material using a ball mill.

This campaign allowed for study of longer term accretion and wear potential, as well as testing
of auxiliary equipment and/or process conditions potentially present in the flow sheet.

Figure 13.13 – FLS Pilot Flash Calciner

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 146

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.23 – May 2017 Calcination Test Work Feed Composition

Element

Greenbushes (%)
Whabouchi DMS (%)

Whabouchi Float (%)

SiO2 67.9 68.2 63.4

Al2O3 24.7 24.1 24.7

Li2O 5.94 6.05 6.07

Fe2O3 0.76 1.6 2.22

MgO 0.15 0.23 0.23

CaO 0.42 0.39 0.5

Na2O 0.55 0.53 0.55

K2O 0.87 0.31 0.69

TiO2 0.04 0.04 0.04

P2O5 0.37 0.05 0.15

MnO 0.08 0.18 0.36

Cr2O3 0.03 0.03 0.03

V2O5 <0.01 <0.01 <0.01

LOI 0.65 0.65 1.14

Sum 102.46 102.36 100.08

Table 13.24 – May 2017 Calcination Test Work Product Quality

Sample

Specific |Gravity (g/cm3)

Lithium Extraction (%)

Condition 1 (Greenbushes) 2.57 73.1

Condition 2 (DMS) 2.53 95.4

Condition 3 (DMS) 2.6 70.8

Condition 4 (DMS/Float mix) 2.5 97.5

Condition 5 (DMS/Float mix) 2.61 91.9

Condition 6 (Greenbushes) 2.57 84.7

Condition 7 (Greenbushes) 2.58 86.0

Condition 8 (Greenbushes) 2.86 32.1

13.2.4.2 TKIS Flash Calcination Test Work - May 2017

In May 2017, a 3-day test was performed by a second reputable flash calciner supplier,
ThyssenKrupp Industrial Solutions (“TKIS”). The primary objective was to determine if the
material was suitable for thermal treatment in a flash calcination system to achieve high
conversion from

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 147

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

alpha to beta spodumene. The test also aimed to determine the operational parameters of a
commercial unit.

The samples received were dried. All the DMS and some of the Greenbushes was ground to
less than 500 microns in a vertical mill, with the balance of Greenbushes milled to 1,500
microns in a HPGR. The float concentrate was mixed with DMS concentrate using a 1:1 ratio.

In total, ten (10) trials were performed during the campaign. A minimal temperature of
1,050°C was necessary to achieve the mineral transformation process from alpha to beta
spodumene. The conversion was however limited by the short retention time (0.5 s) in the test
unit. The supplier noted that higher conversion rates are expected with longer retention time
in a commercial unit (2.5 s). This was demonstrated via re-feeding material into the system.
Particle densities, which are correlated to the level of conversion, were successfully reduced
after being processed multiple times in the flash calciner test unit.

Information related to accretion risk was gathered, allowing for input to engineering solutions.

13.2.5 TKIS FLASH CALCINATION TEST WORK – OCTOBER/NOVEMBER 2018

From October 28 to November 2, 2018, TKIS proceeded with a large-scale test at their semi-
industrial test flash calciner (“POLCAL1”) located at their research center in Beckum, Germany.

Figure 13.14 – TKIS POLCAL1 Building

The main objectives were as follows:

• Verify findings from small scale POLCAL2 tests (September 2017) on calcining temperature
and product quality;

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 148
August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

• Testing of material processing at longer production times and evaluation of the formation of
accretions;

• Optimization of process operation to reduce formation of material build-ups at critical points

inside the gas duct and testing of the ability of blasters to dislodge the build-up;

• Testing of product quality depending on longer residence time as in the previous small-scale
campaign, including testing of mechanical methods to increase the residence time;

• Definition of product quality after calcination.

As with the smaller unit tested in 2017, the flash calciner reflected the commercial
configuration of three (3) preheating cyclones and one (1) material capture cyclone. Flash tube
internal diameter is 320 mm. Calcined material was cooled after capture in two (2) sequential
cooling screws, while the hot gases reporting to the flash tube, generated from pneumatic
cooling in the commercial unit, were simulated by way of a hot gas generator. Energy addition
to the base of the flash tube was at two (2) locations, from a sub-stoichiometric burner (same
principle as the commercial unit), and feed introduction was done through a dispersion box
analogous to the commercial unit. Off-gas dust, captured in a baghouse, was continuously
returned to the process.

Total flash tube length was 20 m, of which 16 m could be maintained at operation

temperature, resulting in a retention time of between 1 and 1.13 seconds in the hot zone (as
compared to >2 seconds for the commercial unit). Material refeed was used to simulate
residence times similar to the commercial unit (note: a re-feed is actually equivalent to less
than twice the retention time, as some of that second-pass time is “lost” to heating of the
material to the conversion temperature).

Whabouchi DMS was milled to < 2 mm in a hammer mill and <1 mm in an HPGR. The final
product was mostly <500 µm material. Whabouchi Float material was dried in a flash drier. The
commercial target ratio (42%/58% DMS/Float) was achieved by continuous use of two
gravimetric dosing devices, followed by a pug mixer to ensure full blending. Both coarser and
finer DMS were tested separately, though the focus of the work was the <1 mm material.
Greenbushes material was used for commissioning of the unit, but only Whabouchi material
for the actual test work.

A feed rate of 200 kg/h was set over the complete trial. Some non-feeding periods occurred
due to switching of material and special events (blockages etc.). Fourteen (14) different
settings were tested.

The primary mean of spodumene conversion determination was XRD analysis. Linear
correlation with true density measurements (pycnometer) was validated.

Results:
• For the tested material mixtures gas temperatures of approx. 1070°C were needed within
the calciner to achieve conversion rates from α- to β-spodumene above 80% on the first pass;

• The first pass (single pass through the calciner) results in high conversion rates up to 95.3%
at 1,100°C average calciner temperature;

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 149

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

• The second pass (refeeding of calcined product) was done to simulate the longer residence
time of the industrial calciner and results in the highest conversion rate of 98.6% at 1,100°C
average calciner temperature. Worst recorded conversion with prolonged exposure to the
heat was 92.3%, at 1,070°C;

• In general, higher gas temperatures and longer residence times resulted in higher conversion
rates;

• Blaster effect could not be validated, as build up was minimal. Location of buildup was close
to the direct NG addition points (these addition points are only present in the test unit and not
present in the commercial unit);

• Effect of using a mechanical method for increasing the holding time on spodumene
conversion was generally good. The test work also showed that the material remained free
flowing after being maintained hot, motionless and deaerated for up to 60 seconds.

13.2.6 ACID BAKE AND LEACH TEST WORK – 2016-2017

In 2016, a lab test program was initiated with COREM in Quebec City. The purpose of this
program was to evaluate the impact of several parameters on the performance of the lithium
extraction: acid concentration and impurity levels, mixing temperature and time, acid bake
temperature and residence time, particle size, etc. A standardized test was developed that
allowed comparison of small changes in operating conditions on lithium and impurity
extraction rates. This activity allowed the selection of optimal operational parameters for the
commercial plant.

The tests used a representative beta-spodumene sample as feed and a synthetic acid based on
the expected recycled acid composition. The beta-spodumene sample elemental composition
is provided in the Table 13.25. This sample was generated from a flash calcination pre-test in
2014.

A standardized lab test method was developed that allowed comparison of small changes in
operating conditions on lithium and impurity extraction rates. This method includes the
following steps:
• Heating of the synthetic acid to the expected commercial plant operating temperature;

• Heating of the beta-spodumene feed sample to the expected commercial plant operating;

• Mixing of the acid and beta-spodumene for a duration predetermined in the test plan;

• Heating of the mixture at temperature and for the duration predetermined in the test plan;

• Water leach at 50% solids at ambient temperature;

• Filtration and washing.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 150

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.25 – Acid Bake Test Work Beta-Spodumene Feed Composition

Component

Composite #5 (% w/w)

Li2O 6.03

SiO2 66.70

Al2O3 23.10

Fe2O3 1.40

MgO 0.10

CaO 0.20

Na2O 0.80

K2O 0.40

TiO2 0.01

P2O5 0.10

MnO 0.20

Cr2O3 <0.01

V2O5 <0.01
ZrO2 <0.02

ZnO 0.03

LOI 0.14

The test results demonstrated that the lithium recovery levels were affected by the acid
strength and composition, excess acid, mixing time and temperature of acid and spodumene
prior to acid bake, time and temperature in acid bake, spodumene size distribution, etc. The
extraction rates from a wide variety of samples of different chemical or mineral compositions
were not investigated. Lithium recoveries between 96 and 97% were achieved for the optimal
operating parameters.

The major impurities extracted were aluminum and iron. The aluminum in the final solution
varied between 3,600 and 5,000 mg/L and the iron varied between 1,200 and 1,700 mg/L.

In addition to the actual operating parameters of the acid bake and leach, the degree of
conversion in the preceding calcination step was determined to be critical to achieving target
extraction rates.

13.2.7 LEACHING AND IMPURITY REMOVAL TEST WORK – PHASE 1 – 2011

The Phase 1 test program was carried out in November 2011 by SGS. Phase 1 consisted of
concentrate leaching, primary impurity removal, secondary impurity removal and final
purification by ion exchange.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 151

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

13.2.7.1 Sample Preparation

Approximately 2,600 kg of acid baked beta-spodumene (solid lithium sulfate) were processed
during the first phase of the pilot plant.

The two (2) blends tested were used to feed the pilot plant (75/25 and 50/50). Table 13.26
shows the chemical analysis of the samples.

Table 13.26– Phase 1 Pilot Plant Feed Analysis

Sample

Li (%)

Si (%)

Al (%)
Fe (%)

Na (%)

S (%)

Cr (g/t)

Zn (g/t)

Mn (g/t)

Mg (g/t)

Ca (g/t)

K (g/t)

75/25 2.24 25.0 10.5 1.04 0.39 6.09 167 134 1,962 1,186 3,431 3,653

50/50 2.29 24.4 10.4 0.96 0.36 6.06 163 103 1,755 905 2,311 3,376

Source of the table: SGS Report

13.2.7.2 Concentrate Leach and Primary Impurity Removal

The objectives of the concentrate leach and the Primary Impurity Removal (“PIR”) were to
dissolve the lithium sulfate from the acid baked beta-spodumene and remove the major
impurities. Approximately 1,200 kg of 75/25 and 1,400 kg of 50/50 were processed in 85
hours.

In the concentrate leach step, solids were mixed with site water at 50% solids by weight and
agitated. Lithium and contaminants (Fe, Al, Si, Mn and Mg) were leached and the final pH of
the slurry was around 1.7. The lithium sulfate contained in the acid baked beta-spodumene
was 100% leached.

In the PIR step, the pH of the slurry was elevated by adding hydrated lime. The major
impurities (Fe, Al and Si) were precipitated as insoluble metal hydroxides. Air was sparged into
the PIR tanks to maintain the oxidative potential of the slurry.

The slurry coming from PIR step was filtered on pan filters and the filtrate proceeded to
Secondary Impurity Removal (“SIR”).

Vacuum filtration testing was carried out on direct PIR discharge and thickened underflow. At
0.7 bar vacuum level the direct discharge sample filter cake moisture was 26.7%.

13.2.7.3 Secondary Impurity Removal

The objective of the secondary impurity removal was to precipitate Ca, Mg and Mn impurities
from the PIR filtrate.

The pH of the solution was increased by adding sodium hydroxide (NaOH) in the first two (2)
tanks. Sodium carbonate (Na2CO3) was added in the third tank to convert all the remaining
divalent impurities to insoluble carbonates.

Impurity levels at the discharge averaged one (1) mg/L Mn, 14 mg/L Mg and 241 mg/L Ca in
the pilot plant runs. Concentrations as low as two (2) mg/L Mg and 200 mg/L Ca were attained
by optimizing key parameters such as retention time and reagent addition. The overall lithium
recovery was 99.1%.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 152

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

13.2.7.4 Ion Exchange

The objective of the Ion Exchange (“IX”) circuit is to further reduce the calcium and magnesium
tenors from the SIR discharge to ten (10) mg/L each.

The IX circuit consisted of three (3) columns packed with a cationic resin, which is selective
towards divalent and trivalent ions.

The process consisted in a lead/lag/regeneration operation. At any time, two (2) columns
would be removing Ca and Mg while the third column would be in resin regeneration mode.

One (1) cycle consists of the following steps: loading, feed wash, acid strip, acid wash,
regeneration and regeneration wash.

The magnesium tenors varied between 0.07 and 0.2 mg/L and 99.0% removal efficiency was
achieved. The calcium tenors varied between 2.4 and 5.7 mg/L and 97.6% removal efficiency
was achieved. Lithium losses in the IX circuit were minimal at 2.7%.

13.2.8 ELECTROMEMBRANE AND LHM CRYSTALLIZATION TEST WORK – PHASE 2 – 2011

Phase 2 of the test program was carried out in December 2011 by SGS. Phase 2 consisted of
membrane electrolysis and crystallization test work.

13.2.8.1 LiOH Membrane Electrolysis

The objective of the electrolysis process is to produce a lithium hydroxide solution from a high
purity lithium sulfate solution. The pilot plant was carried out in a 3-compartment membrane
electrolysis cell.

The central compartment was separated from the cathodic compartment by a cationic
membrane and from the anodic compartment by an anionic membrane.

Under the influence of an electric field, lithium ions from the central compartment were
transported through the cationic membrane. In parallel, the sulfate ions were moved to the
anodic compartment. The anodic reaction generates protons and sulfuric acid is produced.

The pilot plant was run for two (2) 5-day campaigns and samples were taken every four to six
(4 to 6) hours. The 3-compartment cell successfully produced lithium hydroxide at a 14.6 g/L Li
concentration and a 20-30 g/L sulfuric acid solution. Overall, the pilot plant operated for a total
of 228 hours.

13.2.8.2 LiOH-H2O Crystallization

The objective of the crystallization process is to produce high quality solid lithium hydroxide
monohydrate from the lithium hydroxide solution generated through membrane electrolysis.
Atmospheric and vacuum evaporation were tested in three (3) steps.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 153

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

The water evaporation rate was between 6.0 and 6.8 mL/min during the atmospheric
evaporation tests and the product concentration ranged between 18.2 and 19.9% Li. The
crystals were filtered, washed with distilled water and dried in an inert gas environment.

Vacuum evaporation was tested at two (2) different conditions: at 78-82 kPa (vacuum) and 62-
70 ºC and also at 62-70 kPa (vacuum) and 79-81 ºC. The evaporation rate was approximately
7.7 mL/min for both conditions and the final purity of the product was similar to the one that
underwent atmospheric evaporation (17.9% Li).

13.2.9 ACID BAKE, LEACH AND IMPURITY REMOVAL TEST WORK – 2017

In June and July 2017, the acid bake, concentrate leach, and impurity removal circuits were
piloted at COREM’s laboratory facilities in Quebec City. The primary objective was to generate
test material representative of the future commercial plant operations for solid/liquid
separation testing. The test work also permitted the confirmation and optimization of the
commercial plant design parameters, particularly for the purification circuit. The purification
circuit was designed based on the results obtained from the pilot test work described below.

13.2.9.1 Acid Bake

In total 290 kg of beta-spodumene was sulfated: 95 kg of beta-spodumene was sourced from

flash calcination test work conducted in 2017 and 200 kg from a previous calcination campaign
in 2014. The material was homogenized and mixed with 93% sulfuric acid in a V blender with a
stoichiometric acid excess of 30%. Then, the temperature was maintained at 290°C for 45
minutes in a pilot scale rotary furnace operating in continuous mode.

13.2.9.2 Concentrate Leach

The concentrate leach was performed at 65-70°C. Water was added to achieve a solids pulp
concentration of 50% and good agitation was provided through the test duration of one (1)
hour. The resulting slurry was then filtered and the solution was used to test the downstream
impurity removal steps.
Lithium extraction was 94.2%, which is less than the 96% observed during lab scale test work
under similar conditions also performed at COREM. The reasons for this difference are subject
to ongoing investigation but are possibly due to leach temperature, differences in the feed
material, or differences in the acid composition. Laboratory extraction rates of 96% form the
basis for the plant design.

13.2.9.3 Primary Impurity Removal

The objective of Primary Impurity Removal (“PIR”) is to precipitate the aluminum and iron
impurities from the concentrate leach solution. Three (3) reactors provided a combined
reaction time of one (1) hour at a temperature of approximately 65 to 70°C. A recycled lithium
sulfate solution from a third

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 154

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

party provider was added (representing 6% of the total volume) to the concentrate leach
solution. Then lime was added to increase the pH step-wise to the desired setpoint.

Figure 13.15 – Pilot Acid Bake Kiln at COREM

Three (3) PIR tests were undertaken and the results of all three tests are provided to
demonstrate repeatability.

Table 13.27 – Pregnant Leach Solution vs. PIR Treated Pregnant Leach Solution

Element

PLS Feed (mg/L)

PIR 1 (mg/L)

PIR 2 (mg/L)

PIR 3 (mg/L)

Aluminum 4,350 <0.5 <0.5 <0.5

Iron 2,065 <0.3 <0.3 0.4

Magnesium 169.5 225 285 326

Calcium 437.5 455 519 536

Silica 200.5 12 30 33
Manganese 213 138 170 178

The results show that aluminum and iron were successfully removed in PIR.

13.2.9.4 Secondary Impurity Removal

The objective of Secondary Impurity Removal (“SIR”) is to precipitate the impurities

manganese, silica, and a fraction of the magnesium from the PIR treated solution.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 155

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Six (6) SIR tests were conducted batch-wise at approximately 65°C. A single, baffled, vigorously
agitated reactor was used for each test. The pH was raised in the first hour of the test, and
then was maintained for the balance of the test, which was an additional 13 hours. Retention
time was provided to determine the extent of impurity removal as a function of time.
Neutralization was achieved by the addition of a 7% w/w solution of sodium hydroxide,
saturated with lithium.

The results of all six (6) tests are provided to demonstrate repeatability. Although SIR tests 1
through 4 were conducted over a 14-hour duration, the data presented represents the sample
taken at five (5) hours as the reaction was considered to be essentially complete by this time.
The data presented for SIR tests 5 and 6, which were conducted in a second SIR campaign,
represent the sample taken at four (4) hours.

Figure 13.16 – Pilot PIR Circuit at COREM

Table 13.28 – PIR Treated Pregnant Leach Solution vs. SIR Treated Pregnant Leach Solution

Element

SIR Feed (mg/L)

SIR 1

(mg/L)

SIR 2

(mg/L)
SIR 3

(mg/L)

SIR 4

(mg/L)

SIR 5

(mg/L)

SIR 6

(mg/L)

Magnesium 225 180 247 161 166 282 293

Calcium 455 544 540 545 580 618 519

Silica 12 <5 <5 <5 <5 <5 <5

Manganese 138 <0.05 <0.05 <0.05 <0.05 0.13 <0.05

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 156

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

It is evident that treatment by SIR eliminates the manganese and silica impurities to below
their respective targets of one (1) mg/L and five (5) mg/L for feed to electrolysis within a five
(5) hour reaction time.

13.2.9.5 Tertiary Impurity Removal

The objective of Tertiary Impurity Removal (“TIR”) is to remove magnesium and calcium from
the pregnant leach solution prior to polishing with ion exchange.

Three (3) independent TIR tests were conducted batch-wise at 60°C. A single, baffled,
vigorously agitated reactor was used. The pH was raised by the addition of a 7% w/w solution
of sodium hydroxide, saturated with lithium hydroxide. Other reagents were added to aid in
the removal of impurities. The test conditions chosen for TIR replicate the planned commercial
plant operation and are representative of that operation.

The results of all three (3) TIR tests are presented below.

Table 13.29 – SIR Treated Pregnant Leach Solution vs. TIR Treated Pregnant Leach Solution

Test Element

TIR Feed (mg/L)

TIR Treated (mg/L)

TIR 1

Magnesium 282 162

Calcium 618 11

TIR 2

Magnesium 257 10.9

Calcium 584 18

TIR 3

Magnesium 49.9 2

Calcium 573 11

It is immediately evident that test TIR 1 did not achieve the magnesium impurity removal
performance of tests TIR 2 and 3 due to faulty pH measurement during the test. As a result, TIR
1 results were excluded from consideration.

The test results demonstrated that the TIR process is capable of eliminating the magnesium
and calcium impurities to below their respective targets of ten (10) mg/L and 30 mg/L for feed
to ion exchange, which is tasked with polishing the pregnant leach solution of the minor
remaining quantities of magnesium and calcium prior to electrolysis.

13.2.9.6 Solid-Liquid Separation Test Work

One of the key objectives of the piloting activity was to generate material representative of the
future commercial plant operation for solid-liquid separation testing. This activity was
undertaken by Pocock Industrial Inc. at the COREM facility. Solid-liquid separation testing
includes evaluating settling, rheology, and filtration characteristics. A total of 13 samples were
tested representing concentrate leach, PIR, SIR, and TIR. Complete test data, figures and
correlation for solid liquid separation

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 157
August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

equipment selection and sizing were generated by Pocock Industrial. The solid-liquid
separation design criteria generated was used as the basis for equipment sizing in the leach
and purification circuits.

13.2.10 PIR THICKENER OPTIMISATION TEST WORK – AUTUMN 2018

The PIR thickener is sized based on design criteria generated from dynamic thickening tests
undertaken by Pocock Industrial using a continuous bench scale unit equipped with a
motorized rake. A pump continuously provides representative feed slurry to the test thickener
while an underflow pump withdraws underflow solids to maintain a constant pulp level during
testing. Flows are varied to test the hydraulic limits of the material. The resulting criteria is
used to determine the thickener diameter.

When the solid/liquid separation test work was conducted by Pocock at the COREM site in July
2017, the basis for the PIR thickener was that it would not make use of flocculant. In an effort
to reduce the diameter of the thickener, it was later decided that flocculant would be used.
Hence, in September of 2018, COREM undertook another campaign to generate material
representative of primary impurity removal precipitate for further solid/liquid separation test
work by Pocock. The continuously operated pilot scale primary impurity removal circuit
operated by COREM was the same as the one used in 2017.

A full day was required to reach steady state operation for each test condition and then a
sample for solid/liquid separation test work was harvested. The optimum conditions were
found to be a slurry generated whilst operating with a 300% seed recycle that is flocculated
upstream of the thickener. As a result of this test work, the thickener diameter was
significantly reduced.

13.2.11 ION EXCHANGE TEST WORK – 2014 AND 2016

During 2014 and 2016, test work was performed by Electrosynthesis Company that defined the
removal of divalent impurities (calcium and magnesium) remaining in the lithium sulphate
solution already purified by PIR, SIR, and TIR. The goal was to ensure robust operation and
optimal longterm performance of the electromembrane process which is sensitive to impurity
levels. The optimal ion exchange resin was identified along with key variables such as loading
rate, elution rate, and eluent concentration. It was demonstrated that ion exchange is an
effective means by which to reduce both the calcium and magnesium concentrations below
the target of one (1) mg/L in advance of electrolysis. This test work formed the basis of design
for the commercial ion exchange system.

13.2.12 ION EXCHANGE TEST WORK – 2019

Ion exchange column test work was conducted at SGS in Lakefield, Ontario during the months
of January and February of 2019. The test work program confirmed the selected resin’s
effectiveness in removing calcium and magnesium, the superficial velocity to be used in the
design, and the loading capacity of the resin. The height of resin used was the same height as
will be used in the commercial plant, thereby providing the same number of theoretical stages.
Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 158

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Tests were performed using a single column of commercial height and in lead-lag operation
using two columns of commercial height. Representative brine was obtained from the Phase 1
Plant (P1P) for the tests; 200 bed volumes were processed in each of three tests using single
columns and 360 bed volumes were process in the lead-lag test. Calcium was successfully
removed to below a detection limit of 0.06 mg/L. Magnesium was successfully removed to
below a detection limit of 0.008 mg/L.

Test work is ongoing at SGS to further optimize the water balance around ion exchange as well
as the life of the resin.

13.2.13 MEMBRANE ELECTROLYSIS TEST WORK

Various phases of membrane electrolysis test work on different configurations were

performed by the Electrosynthesis Company to improve upon the results obtained during
original SGS test work.

The objectives were to determine optimal flow sheet and operating parameters
(concentration, current efficiency, current density, configuration etc.) and to estimate
membrane life cycle.

13.2.13.1 Membrane Electrolysis Test Work – Autumn 2012

An initial series of test work was performed in the autumn of 2012. The objectives of the tests
were to:

• Increase the concentration of the lithium hydroxide and sulfuric acid streams produced;

• Improve current efficiency;

• Test different cationic and anionic membranes to increase operating temperature, reducing
cooling water requirements.

The experiments were performed in an Electrocell MP (100 cm2) using three (3) distinct
compartments: one compartment for the salt stream (Li2SO4), one for the base recovery
(LiOH) and one for the acid recovery (H2SO4). When an electric field is applied, the cations
migrate from the salt, through the cationic membrane and into the base loop. Anions migrate
through the anionic membrane and into the acid loop. Water molecules are split at the
electrodes located in the base and acid compartments, allowing the formation of the acidic
and basic solutions.
A single cationic membrane and two (2) anionic membranes were tested during a total of eight
trials.

Based on the results, it was concluded that the 3-compartment membrane electrolysis
successfully produced lithium hydroxide at high temperature (60°C). The tests determined
optimal base and acid concentrations, current density, current efficiency, cell voltage, water
transport and feed operating pH.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 159

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

13.2.13.2 Membrane Electrolysis Test Work – Summer 2013

A subsequent series of test work was performed in the summer of 2013. The objectives of the
tests were to determine if a 2-compartment cell could be used to generate lithium hydroxide
and if so, determine operating parameters and subsequent impact on 3-compartment cell
operation.

The experiments on the 2-compartment cells were performed in an ICI FM-01 electrolysis cell
(64 cm2) using two (2) distinct compartments: 1-compartment for the salt stream (Li2SO4) and
one for the base recovery (LiOH). When an electric field is applied, the cations migrate from
the salt, through the cationic membrane and into the base loop. Anions remain in the salt
compartment, being acidified by protons being generated at the anode. Water molecules are
split at the electrodes located in the base and salt compartments.

A single cationic membrane was tested during multiple trials.

Based on the results, it was concluded that 2-compartment membrane electrolysis successfully
converted a significant portion of the lithium in the feed stream with a high current efficiency.
The tests determined optimal base concentration, current density, current efficiency, cell
voltage, water transport and feed operating pH and extend of conversion.

Additional test work was performed on the salt recovered from the 2-compartment test work
to determine if further lithium hydroxide production could be achieved by processing the
waste stream in a 3-compartment cell. The experiments were performed in the same
Electrocell MP (100 cm2) used for the initial trials (Autumn 2012) and successfully
demonstrated that complete conversion could be achieved.

13.2.13.3 Membrane Electrolysis Test Work – Winter 2014

In the winter of 2014, long term stability testing on the 2-compartment cells and 3-
compartment cells were undertaken at the Electrosynthesis laboratory to determine
membrane and electrode lifespan. Test work was conducted over a period of 1,000 hours and
results were used to determine expected replacement frequency for the membranes and
electrodes allowing for calculation of the operating costs.

13.2.13.4 Phase 1 Plant Electrolysis Operation through June 2019

The Phase 1 Plant demonstration scale electrolyzer was successfully commissioned in February
2017 with 2 cells. Based on smooth operation, the cell pack was expanded to a nominal
capacity of 250 TPY LiOH.H2O equivalent in April 2017. The system ran smoothly, in
combination with the LiOH evaporator/crystallizer. It produced high-quality LiOH product from
recycled lithium sulfate provided by a client. The operation of the electrolyzer suggested
possibilities for optimization of the circuit.

13.2.14 LIOH.H2O CRYSTALLIZATION TEST WORK – 2014 AND 2017

During summer 2014, Lithium Hydroxide Monohydrate (LHM) crystallization test work was
performed by the GEA Group in Duisburg Germany to produce about 2 kg of battery grade
LHM representative

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 160

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

samples from lithium hydroxide solution generated through electromembrane process at

Electrosynthesis Company and generate basic data for engineering of crystallization unit
operations. 63 kg of solution was sent to GEA, from which 2 kg of LHM was produced. Final
product quality respected the more stringent cathode materials manufacturer's requirements,
with for Na tenor <20 ppm and a total of chromium, zinc and iron content of less than 150 ppb.
The test work done at GEA confirmed the proposed crystallization process flow sheet and
equipment selection for this feasibility study.

Additional test work was performed with GEA in 2017 to determine the impact of certain
impurities should they be present. This test work resulted in an adjustment in the purge
requirements from the LHM crystallizer.

13.2.15 LHM DRYING TEST WORK – FEBRUARY & APRIL 2018, AND ONGOING AT P1P

13.2.15.1 Piloting

Test work was performed at Andritz’ facilities in Germany. A first set of tests were performed
in February 2018, with the aim of validating operating parameters.

LHM slurry feed material was sourced from Nemaska’s Phase 1 Plant (P1P) facility and shipped
in 200 L barrels.
A small type HZ 25/0.1 peeler centrifuge was used for dewatering (2.5 L basket volume),
inerted with Nitrogen. Resulting free moisture was below 5%, for g-forces ranging from 150 to
700. Purity requirements were met at all wash ratios (incl. none), but a ratio greater than 0.5
kg/kg started affecting average particle size. The primary aim of dewatering was not for
centrifuge sizing, but for fluid bed dryer feed production.

The drying tests were performed in a pilot-plant sized fluid bed reactor (185 mm ID), aimed at
validating a Mollier-type diagram (dry LHM stability vs. temperature and ambient moisture
levels). The dryer is fluidized using Nitrogen wetted with steam (controlled addition, measured
on-line), at a controlled incoming temperature.

The off-gas is monitored for humidity and dedusted in a baghouse (dust is contaminated and
not collected). Material is added manually to the reactor on a semi-batch basis and left inside
the reactor for the duration of a run, with samples extracted at set time intervals. The addition
rate sets the bed temperature. Analysis is based on loss-of-weight in an oven, i.e. relies on the
(reasonable) assumption that drying precedes dehydration. PSD is quantified using laser
diffraction, and qualitatively assessed by micrographs.

A set of five (5) conditions was trialed, validating operating conditions as well as transition
stages (start-up/shut-down). The material proved more resilient than initially feared, with only
extreme dry conditions (very low gas dew points) leading to dehydration. Under expected
operating conditions, the product was fully dried, while retaining its hydrate to within
specifications (57.5% typical equivalent LiOH).

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 161

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

The drier could be started up from empty, had a very good fluidization behavior, and the
product qualitatively had good flowability. Some degree of caking was observable after >1
week in storage.

13.2.15.2 Marketing Sample Production

A mandate for marketing sample production provided further opportunity for testing
operating variables. This took place in April 2018, processing the balance of the LHM slurry
material available at Andritz’ test facility. In particular, considering evaluation of the earlier
tests showed some flexibility, this allowed for testing of more aggressive process conditions.

Due to the larger volumes involved, Andritz’s larger (commercial-sized) pusher centrifuge
SZ400/2-h could be used (a pusher is used in the commercial plant), running at 580 kg/h. This
resulted in a very low residual free moisture of less than 2.0% by weight. A total of 126 kg
(wet) material were produced and forwarded on to the drying test center.
Fluid bed drying took place in the same base unit as tested previously, modified to allow for
continuous feed (screw feeder) and continuous discharge (bed overflow into a sealed/inerted
bottle).

About 87 kg of usable product were generated, of consistent and good quality. Losses to the
off-gas, at around 20% of the feed, were significant, but expected in such a small unit. Design
flexibility on the freeboard of the commercial unit will allow for much less elutriation.

The bulk of the production was done under the operating conditions previously validated, at a
rate of about 2.8 kg/h discharge. Challenging material handling of the dewatered feed material
were confirmed, with the screw feeder hopper plugging, leading to the production reverting
back to a manual feed introduction.

Further tests showed that idling was possible without quality degradation at temperatures up
to approx. 70°C as long as the dew point of the gas was carefully controlled, while any
temperatures in excess of that led to product quality loss from dehydration. Conversely, even
small amounts of moisture (e.g. from feed material) at low temperatures lead to lumping.

13.2.15.3 LHM Dryer at P1P (on-going)

In January 2019, Nemaska installed a LHM dryer in the P1P demonstration plant. This unit was
sourced from the same Supplier as the commercial unit (GEA) and has the same mechanical
design. The process conditions are drawn from GEA’s past experience in the field. Though it
operates under slightly different conditions than proposed in the Andritz test work (colder and
dryer), they are still within the stability envelop validated earlier.

The P1P dryer is being used to gain practical experience related to the dryer operation.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 162

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

13.2.16 ACID RECYCLE SOLUBILITY TEST WORK – 2016 AND 2017

Nemaska intends to recycle the sulfuric acid produced in the electrochemical process. This
requires an evaporative crystallizer to concentrate the sulfuric acid, followed by a Spent Acid
Concentrator (“SAC”) process to remove water from sulfuric acid before it is recycled.

From June 2016 to October 2017, NORAM organized a series of tests at BC Research (“BCRI”)
to identify optimum process conditions for the concentration of the sulfuric acid solution. The
tests were also aiming at identifying conditions for the final evaporation stage required to
regenerate acid (Figure 13.17).

Figure 13.17 – Material Testing Rig for Solubility Tests at BCRI

This test work provided valuable information on the behaviour of lithium sulphate / sulfuric
acid / water ternary systems under various temperatures and pressure conditions. Based on
the experimental testing performed during that test, a conceptual flow diagram was created
for the acid regeneration process.

Initial work in 2016 was followed by work in 2017. This provided more detailed solubility data
on the ternary system at elevated temperatures (between 35 and 160°C). This was required to
confirm the conceptual design of the regeneration process. Ternary phase solubility diagrams
for lithium sulphate, sulfuric acid and water system were produced at various temperatures,
allowing the system to be designed for the whole range of concentrations, from the diluted
acid feed with dissolved sulphate salts to the final acid concentration at less than 10% H2O.

13.2.17 ACID RECYCLE TEST WORK – 2017

During the period between May and September 2017, laboratory and pilot scale test work was
performed with multiple reputable technology suppliers to validate both acid concentration
stages. These tests simulated the range of conditions (nominal and upset) envisioned for the
commercial acid recycle circuit. The objective was to confirm the equipment configuration,
allow design robustness and allow the suppliers to accurately quote the equipment (Figure
13.18).

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 163

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Figure 13.18 – Pilot SAC Unit

(Courtesy of De Dietrich)

13.2.18 OTHER TEST WORK – 2017-2019

Multiple other tests were performed in order to gather data for proper engineering of the
facility. This included:

• Material handling properties testing for proper sizing of silos, piles, chutes and conveyors.

• River water quality testing for sizing of water treatment systems.

• Materials of construction test work for selection of appropriate materials of construction

within the highly abrasive and corrosive sulphation area (pug mixer, acid bake kiln, etc.).

13.2.19 ELECTROCHEMICAL DEMONSTRATION PLANT – PHASE 1 PLANT –2017-2019

Phase 1 Plant is a ~1/65 scale demonstration plant with full scale electrolysis cells installed by
Nemaska representing an investment of approximately $42M. It is designed for continuous
operation, with complete instrumentation and DCS, allowing automated and safe operation. It
can produce high

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 164

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

quality lithium hydroxide monohydrate (LHM) from spodumene and recycled lithium sulfate
salts. It has a nameplate capacity of 500 t/y of LHM crystal production from recycled lithium
sulfate salts and 100 t/y of LHM from spodumene concentrate. The Phase 1 Plant design is
based on comparable industries (namely chlor-alkali), the traditional spodumene processing
flow sheet, internally developed processes, and know-how from reputable technology
suppliers. Its design basis and technologies were established through extensive laboratory and
pilot scale testing described above and realized under the supervision of Nemaska technical
team and/or designated engineering firms at independent testing and suppliers’ facilities since
2011. Figure 13.19 below shows part of the Phase 1 Plant.

Figure 13.19 – P1P Overview Picture - Purification and Crystallization Unit Operation

The objectives of Nemaska in building and operating the Phase 1 demonstration plant in
advance of starting commercial scale operation were multiple:

• To demonstrate its ability to repeatedly produce lithium hydroxide according to quality

specifications as defined by customers including battery customers;

• To qualify its products with customers and sign off-take agreements before starting
operation of the commercial plant;

• For the development of staff skills and internal processes and to provide strong foundations
for the integration of new staff in the commercial plant;

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 165

August 2019
DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

• Process improvements made during the life of the demonstration plant and operational
lessons learned can be integrated in the engineering of the commercial plant;

• To demonstrate the versatility of the process by also converting lithium sulfate salts (that are
produced by some customers in their industrial processes) into lithium hydroxide.

The timeline for the Phase 1 Plant is as follows:

• May 2016 – P1P financing completed. The total budget to build and operate the Phase 1
Plant for two years is $38 M;

• June 2016 – Beginning of construction at Shawinigan site;

• December 2016 – Phase 1 Plant operation team training;

• February 2017 – Start Electrolysis continuous operation on synthetic lithium sulfate solution
from LiOH and H2SO4;

• March 2017 – First tonne equivalent of LiOH-H2O solution produced from synthetic lithium
sulfate solution;

• April 2017 – Start Purification and Electrolysis on recycled lithium sulfate salts from a client;

• June 2017 – First tonnes equivalent of LiOH-H2O solution produced from recycled lithium
sulfate salts meeting Johnson Matthey Battery Materials battery grade specification;

• October 2017 – Start processing spodumene concentrate feed while over 20 tonnes of LHM
meeting our client specification were produced from lithium sulphate salts;

• December 2017 – First tonnes equivalent of LiOH-H2O solution produced from spodumene
concentrate meeting battery grade specifications;

• February 2019 – First tonnes of LiOH-H2O crystals and samples to potentials clients within 60
days of installing a fluid bed dryer and packaging equipment;

• May 2019 – Over 240 tonnes of Whabouchi spodumene concentrate processed and over 100
tonnes equivalent of LiOH-H2O produced from spodumene concentrate and/or recycled
lithium sulfate salts. Phase 1 Plant continues to deliver High purity lithium hydroxide
monohydrate commercial samples to potential clients internationally.

The Phase 1 Plant includes the following key unit operations:

• Acid bake pug mixer, kiln, and cooler;

• Leaching reactor and filter;

• PIR reactors and filter;

• SIR reactors and filter;

• TIR reactors and filter;

• IX columns;

• Electromembrane cells within electrolyzer;

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 166

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

• Crude and pure LHM crystallizer;

• LHM dryer;

• Services.

When running on recycled lithium sulfate salts, the acid bake and leaching steps aren’t used,
and impurity removal is adapted for the specific feedstock.

Since the acid regeneration circuit is not present, the spent anolyte from electrolysis is
recycled to PIR where it is neutralized.

The spodumene concentrate was sourced from the Whabouchi mine. Spodumene calcination
was performed offsite by third-party suppliers.

Key operational learnings by process sectors:

a. Lithium extraction circuit

Between Q4-2017 and Q1-2019, approximately 240 tonnes of Whabouchi spodumene

concentrate were processed. Operational learnings influenced commercial plant materials
selection to prevent premature corrosion and erosion of equipment.

b. Purification circuit

Conditions to produce gypsum with low lithium losses and removal of almost all metallic
impurities.

Conditions to produce high purity feed to electrolysis.

c. Electrolysis circuit

Operation of commercial scale electrolysis cell at various current densities. Confirmation of key
design parameters.

Conditions that affect membrane and anode life, and current efficiency. More than 5,000
hours of continuous operation to understand cell performance over time.

d. Lithium Hydroxide Monohydrate circuit

Confirmation of two step crystallization process configuration to obtain product quality. In Q1

2019 quality battery grade was confirmed for at least two potentials clients
Since the beginning of its operation in Q1 2017 to Q1 2019, many of those objectives and
milestones were achieved:

• Phase 1 Plant was operated in several campaigns and produced over 100 tonnes equivalent
of LiOH-H2O produced from spodumene concentrate and/or recycled lithium sulfate salts. The
plant was deliberately run at lower than nameplate capacity to ramp up and stabilize
operation and adapt operation to match the availability of feedstock. This material qualified as
battery grade as per typical market specifications.

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 167

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

Table 13.30 – Nemaska Lithium LiOH-H2O Produced at P1P from Recycled Lithium Sulfate
Solution

Element Unit

Market LiOH-H2O Specs* Span of Max Values

Nemaska LiOH-H2O Product measures

LiOH % w/w 54.8 - 56.5 Min 56.5

Ca mg/kg 10 - 100 < 4

Na mg/kg 20 - 500 < 4

K mg/kg 10 - 250 < 4

Mg mg/kg 10 < 4

Fe mg/kg 5 - 21 < 4

Al mg/kg 10 < 4

CO2 % w/w 0.035 - 0.35 < 0.1

Cl mg/kg 15 - 100 < 10

SO4 mg/kg 50 - 300 < 120

Cr mg/kg 5 - 100 < 4

Cu mg/kg 1 - 5 < 4

Ni mg/kg 1 - 10 < 4

Si mg/kg 20 - 30 < 4
Zn mg/kg 10 < 4

Sol. Acid mg/kg 40 - 1,000 < 50

* Data from publicly available company product list

• Since the end of 2016, engineers, process specialists, support teams, key management staff,
and 20 technical operators have been hired and trained to operate Phase 1 Plant. These highly
skilled technicians have educational backgrounds in mechanical, chemical, and electrical
disciplines as well as a range of technical work experience including chemical plant start-up.
Both process safety reviews during engineering and skilled personnel proved to be efficient at
preventing accidents during commissioning, start-up, and operation of the plant. This phase
allowed the development of important internal processes and know-how required for the
commercial phase.

• Learnings from Phase 1 Plant operation were transferred and integrated to the process and
engineering design of the commercial plant. This mitigates many of the technical risks
associated with new process development.

In 2019 the Phase 1 Plant will continue to produce LHM crystal from recycled lithium sulfate
salts and/or spodumene concentrate from the Whabouchi mine, to further demonstrate and
optimize the process, qualify product with clients, and to develop the know-how of the
workers in advance of the start-up of the full-scale commercial process. In addition,
optimisation tests will be done to increase current efficiency, increase lithium recovery in acid
bake and leaching, reduce impurities

Nemaska Lithium Inc. NI 43-101 Technical Report Report on the Estimate to Complete for the
Whabouchi Lithium Mine and Shawinigan Electrochemical Plant

/ Page 168

August 2019

DRA/Met-Chem Ref.: C3785- Final Report C3785-Final_Rep_20190809

concentration in electrolysis feed, reduce purge from the crude LHM crystalliser, and to extend
life of resins of IX columns before regeneration, amongst others.

The Phase 1 Plant has significantly contributed to de-risking the project. Nevertheless, it is
important to understand the limits of what could be tested and learnt. The following points
should be noted:

• The Phase 1 Plant does not have calcination, acid regeneration and purge circuits, therefore
these systems have not been tested except offsite or at supplier facilities;

• The Phase 1 Plant does not have either of the two-stage acid recirculation loop. As a result,
neither stage has been tested in P1P, and some operating conditions are somewhat different
(for example the pug mill uses fresh acid instead of recycled acid, which does not have the
same composition). In addition, the buildup of minor impurities through the acid recycle loop
cannot be observed;

• Due to design constraints, the purification circuit is not designed or operated exactly like the
commercial facility is expected to operate. The purification circuit is therefore able to test the
chemistry and certain aspects of the design but can only contribute qualitatively to other
aspects of the design (e.g. filter sizing);

• Due to design constraints, certain flows in P1P are of a different concentration, or of a

different composition, than what is planned for the commercial facility. This affects densities,
viscosities and some efficiencies. This means that direct comparison of the behavior within
equipment cannot always be made, although qualitative comparison can be made;

• Due to limited sulphation capacity, P1P generally operates on a blend of recycled lithium
sulfate salts and spodumene concentrate. The lithium sulfate salts are generally significantly
purer than the concentrate which aids the operation of the purification circuit. Long campaigns
with only spodumene concentrate as feed have not been performed, and therefore should any
impurities or problems accumulate over time, these will not have been observed.