Post on 07-Apr-2019
1
PRODUCTION OF MAGNESIUM
AND ALUMINUM-MAGNESIUM
ALLOYS FROM RECYCLED
SECONDARY ALUMINUM SCRAP
MELTS
A.J. Gesing1, S.K. Das1, R. Lutfe2
1. Phinix L.L.C., 2. MER Corp.
2
Outline
• Need for Mg recovery for recycling
• Mg recovery process concept
• Theoretical considerations and experimental
program
• Experimental results
• Markets and economics
• Proof-of-concept
3
Need for recovery and recycling of Mg scrap
• Mg is formed by HP diecasting
• In-house diecasting scrap is recycled
• Post-consumer Mg scrap is not recycled
• Mg scrap shred is usually left in Al fraction
• Most Mg is recycled as alloying element of Al
• Excess Mg is chlorinated out of Al alloy melt
>30,000 tonnes of Mg/year in NA• This chlorination is equivalent to burning
US$60 million/year in NA
4
Expensive prime dilution• 300,000 tonnes/y of energy intensive prime Al is used in
NA to dilute ~0.5wt% excess Mg in the UBC melt
destined for 3X04 can body alloy at a premium of
~US$950/tonne.
• 100,000 tonnes/y of expensive and energy intensive
prime Al is used in NA to dilute ~1 wt% excess Mg in the
mixed 3754-6111 new stamping plant scrap melt
destined for 6111 auto closure sheet at a premium of
~US$600/tonne
• Electrorefining of Mg for recycling from these sheet alloy
melts in NA would:
• recover 7,000 tonnes/y of Mg valued at:
$14 million/y
• avoid 400,000 tonnes/y of prime Al, with premiums of:
US$345 million/y
5
Mg recycling process concept
Replace chlorination
system with an
electrorefining cell in Al
remelt facility
+
Al melter
Electrolyte
Anode + Al-Si-Cu-Mn-Fe melt
Al-Mg-Si-Cu-Mn-Fe scrap
Prime purityMg-Al melt
Specification purityAl-Si-Cu-Mn-Fe
-
Cathode – Mg-Al
Al-Si-Cu-Mn-Fe melt
6
Mg recycling process concept
Replace chlorination
system with an
electrorefining cell in Al
remelt facility
7
Mg electrorefining cell
Mg-Al (r=1.6)
Electrolyte(r=2.0)
Al-Mg(r=2.3)
Mg2+ + 2 e- Mg
Mg Mg2+ + 2 e-
?
DE = 0DE = -RT/2F × ln( )
[Mg]Mg
[Mg]Al
Open circuit potential
8
Preferred Fluoride Electrolyte
Only one viable composition range for fluorides for 2.05 density
system:
84+/-3 LiF + 16+/-3 MgF2
Advantages:
• Very high ionic electrical conductivity low power consumption
• Very low electronic conductivity high current efficiency
• Low viscosity good circulation
• No complexing between Li+ and Mg2+ high mobility of Mg2+ in
the electrolyte.
• Not very hygroscopic
• Superheat in the right range for liquid containment in electrolyte
freeze
Disadvantages:
• Experimentally 0.2 wt% Li in MgAl cathode metal
9
Electrorefining experiments
Anode +
Cathode -
Mg collection
Al impeller
Ar purge
Electrolyte
Al
C felt filter
Mg feed
Mg out
Vacuum
Electrolyte impeller
Lid
Alumina insulator
30 cm
10
Laboratory cell parameters
Electrolyte 84 LiF + 16 MgF2
Inter-electrode distance (cm) 8
Nominal cathode electrode area (cm2) 50–300
Cathode material graphite or steel
Anode crucible material graphite
High temperature electrical insulator >99.5% alpha alumina
Current density (A/cm2) 0.5–3.5
Cell current (A) 50–300
Cell voltage (V) 0.5–3
11
Electrorefining Results
Cell voltage and energy consumption
• Low cell voltage: 1 V @ 0.9 A/cm2, 2.1 V @ 3.5 A/cm2
• Stable operation: +/- 0.1 V up to CD of 3.5 A/cm2
• Energy consumption: 2.5 kWh/kg of Mg @ 0.9 A/cm2
0
0.5
1
1.5
2
2.5
0 1 2 3 4
Ce
ll vo
ltag
e (
V)
Current density (A/cm2)
12
Mg extraction from the anode alloy
• Mg added to anode at intervals to keep Mg in Al melt at ~1.5 wt%
• Al cathode sampled at intervals and analyzed for Mg content. Measured
concentration compared with calculations based on assumed current efficiency.
• Efficiency of Mg extraction from Al anode: ~ 100% (best fit of prediction to data)
10
15
20
25
30
35
40
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0 1 2 3 4 5 Mg
AZ6
1 a
dd
itio
n t
o A
l me
lt (
g)
Mg
in A
l an
od
e a
lloy
(wt%
)
Electrorefining time (hours)
predicted
average
Mg AZ61 addition
13
Electrolyte stability
• Electrolyte sampled at intervals and analyzed by EDX-XRF.
• Stable XRF intensity ratio IMg / IF indicates no change in the LiF:
MgF2 ratio during electrolysis
0.1
0.2
0.3
0.4
0.5
0 1 2 3 4 5 6 7
EDX
X-r
ay in
ten
sity
rat
io I
Mg/I
F
Elapsed time (h)
Electrorefining at current density of 0.88 A/cm2
14
Mg recovery at the
cathode
Mg weight
recovered > 62 g
Mg content of
cathode>99%
Mg recovery by
weight>87%
Mg recovery by
acid dissolution
- H2 evolution97%
14
Mg
Mg
electrolyte
Steel
Steel tube
+ Al
-
15
Cathode Mg product elemental composition Sample
Comments#1 wt% #2 wt%
Mg 98.60…. 99.66…. High selectivity for Mg
Al 0.86…. 0.0630 Variable between Mg droplets:
different current density
different Mg% in Al
Li 0.5400 0.2200
Na 0.0055 0.0075
Fe 0.0106 0.0021 (Si + Fe) 0.02~0.04% << 0.3% target
(Si+Fe+Cu+Zn+Mn+Co) 0.025~0.046%
NO electrolytic transfer to cathode
PRIME PURITY product
Si 0.0077 0.0395
Mn <0.0010 0.0011
Zn 0.0015 0.0016
Co 0.0035 0.0015
Cu <0.0010
• Cathode product used for ICP was recovered from the cathode surface during cell post-mortem
• Contaminants are diluted by a factor of 20-100 when Mg product is used as alloying additive to Al
• Na + Li can be refined out of the molten Al alloy
• Mg product composition is suitable for alloying prime Al alloys
16
Anode Al product ICP elemental composition
• Anode product was sampled during electrorefining run with power on.• Mg and alkali metals are extracted from the anode pool during electrorefining• Transition metals remain in the anode and are not transferred to the
cathode product
• Mg in the Al product can be reduced to 0.14 wt%, OK for foundry alloys
Start EndComment
wt% wt.%
Al 95.560
0
98.800
0
Al concentrates as Mg moves to cathode
Mg 3.3500 0.1400 20X reduction, 95% Mg removed
Li 0.1020 0.0090 << 0.05%, 90% Li removed
10X reduction during electrorefining
Na 0.0019 0.0014 No change
Fe 0.2220 0.2920
No change
No undesirable electrolytic transfer to the
cathode product
Si 0.0620 0.0550
Mn 0.5470 0.5430
Zn 0.1500 0.1530
Cu 0.3450
Co 0.0070 0.0060
17
Selected Markets• 38x and 319 foundry alloy: ( US market 1.0 ×106 t/y)
– 380.X scrap melter feed with 1-2 wt% Mg; product 0.1% Mg
– Drivers: eliminate chlorination, recover Mg, improve Al recovery
• 3004 can body sheet: (US market 1.5 × 106 t/y)
– 100% UBC melter feed (83% 3004 + 17% 5182, ~1.5 wt% Mg melt)
– product with ~1% Mg
– Drivers: reduce feed costs (no prime Al), recover Mg
• 6111 auto closure sheet: (US market 0.3 × 106 t/y)
– New, mixed 5754 (3% Mg) and 6111 (1% Mg) stamping plant sheet scrap
melter feed with 1.5%~2% Mg; product with ~0.75% Mg
– Drivers: reduce feed costs (no prime Al or scrap sorting ), recover Mg
• Mg hardener for alloying prime Al alloys
– Prime quality Mg electrorefined from each of the three markets above.
– Drivers: Low cost prime purity Mg (same production cost and
environmental impact as secondary Al alloy products <<< prime Mg
production cost and environmental impact)
18
Conclusions• Experimental results exceeded almost all project targets.
– Low energy consumption
– Low GHG emissions
– High current efficiency
– High Mg recovery
– Product compositions suitable for selected markets
• Techno-economic results
– Production costs competitive with Al secondary alloy production for
the selected markets
• Conceptual process and equipment design combined with positive
experimental results and profitable tecno-economic analysis give a
proof-of-concept for the RE12 electrorefining process of Mg
recovery from Al alloy melts.
19
AcknowledgementsFinancial and technical assistance:
US DOE ARPA-e, Contract Number DE-AR0000413.
James Klausner (Program Director), Bahman Abbasi,
Thomas Bucher and Daniel Matuszak.
Industrial and commercialization:
Ray Peterson of Real Alloy
Experiments at MER Corporation
Raouf Loutfy, Kevin Loutfy, David Thweatt, Y. Kim,
Jay DeSilva, Charles Ibrahim and Mr. Robert Hoffman
Mark Gesing of Gesing Consultants Inc.
Al alloy scrap samples
Real Alloy and Alcoa
FactSage training and support
Prof. Arthur Pelton, Christian Robelin, Aimen Gheribi, Patrice Chartrand