9th MMM-Intermag, Jan 5-9, 2004.
Anaheim, California
Improved Synthesis of FePt (CoPt) Nanoparticles Min Chen, Hao Zeng, Shouheng Sun IBM T. J. Watson Research Center, Yorktown Heights, New York 10598, USA In collaboration with C. Murray, H. Hamann, G. Held (IBM Watson) S. Raoux, M. Toney, J. Baglin, T. Thomson, B. Terris (IBM Almaden) J. Li, Z. L. Wang (Georgia Institute of Technology) J. P. Liu (Univ. of Texas, Arlington) S. Wang, R. White (Stanford Univ.) Support in part by DARPA/ARO 19-03-1-0038 and DARPA/ONR N00014-01-1-0885
¾Motivation ¾Current synthesis. ¾Improved synthesis. ¾Application potential.
FePt: Interesting bi-component materials FePt structure is composition dependent, Fe3Pt(fcc) FePt (fct) FePt3(fcc). Face centered tetragonal (fct) structure, also called L10 structure
Fct structured FePt has large anisotropy constant, K, and is magnetically hard, meaning smaller FePt particles can still be ferromagnetic with large coercivity.
∆E ~KV/kT 2 kB T tP f O 3 ⋅ ln H C = H O ⋅ 1 − ln 2 V K U - Sharrock
FePt: High anisotropy constant
CoSm FePt
100
anisotropy (106erg/cc)
CoPt
10
1 0
2
4
6 2
8
10
12
CoCr15Pt12. CoCrxPt12 CoCr15Ptx CoCr20Pt10B6 CoCrxTa4 FePt CoPt FePd Co5Sm Co3Pt Charap
∆E ~KV/kT
Relaxation time
τ = τ0eKV/2kT
14
5
MS (10 erg/cc) D. Weller, A. Moser, IEEE Trans. Mag. 2000, 35, 4423.
Ku V >> kT
KuV < kT
FePt: Interesting chemistry
Fe Pt Pt Fe
The binding of Fe with Pt makes Fe much more stable against deep oxidation. FePt is chemically more stable than other well-known hard magnetic materials, e.g. Sm-Co & Nd-Fe-B.
Fe Pt Pt Fe Pt Fe
Surface Pt can bind to S strongly to form Pt-S bond, facilitating site specific bonding of the particles to biomolecules for highly sensitive bioseparation and detection. Gu et al, Chem. Commun. 2003, 1966.
Potential applications (3 examples) ¾Data storage media (large Hc > 3000 Oe)
¾Permanent magnet (Large Hc + high moment) H
N
H
H
¾Bio-separation & detection. Mr
H Hc
S H
H
Ms
FePt synthesis via decomposition/reduction CO OC
CO
Fe
[Fe]
CO CO
CO
CH3 O O
CH3
O
2+
Pt
O CH3
CH3
[Fe-Pt] reduction
[Pt]
acac
Fe(CO)5 R-COOH
Fe R-NH2 Pt FePt
Sun et al, Science 2000, 287, 1889.
Pt(acac)2 R-NH2 Reduction
FePt composition control Fe
Pt Pt
Pt
Fe Pt
Fe Pt
Fe
80
Fe
70
Pt Fe
x in FexPt(100-x)
Fe
60 50 40 30 0.5
1.0
1.5
2.0
2.5
3.0
mmoles of Fe(CO)5
Fe(CO)5+ Pt(acac)2 + Oleic acid + Oleylamine ? mmol
0.5 mmol
0.5 mmol
0.5 mmol
Oleic acid: CH3(CH2)7CH=CH(CH2)7-COOH Oleyl amine: CH3(CH2)7CH=CH(CH2)7-CH2-NH2 Sun et al, IEEE Trans. Magn., 37, 1239 (2001).
Surfactant bonding to FePt surface –FTIR studies N. Shukla, et al, J. Magn. Magn. Mater. 2003, 266, 178
O
Monodendate form
O
Fe
H2N
Pt
NH2 bonds to Pt via electron donation
O
Fe
C O
Pt Fe
H2N
H
Bidendate form
Pt
N R
H
FePt Nanoparticle Structure As-syn particle thin film Fe/Pt
Annealing
C C CC C C C CC C C C C C CC C C C CC CCC C C C C C C C C CC C C C C C C C CC CC CC C CC C CCC CCC CCC CC C CC CC CC C C CC C C CC C C C C C C C C C C C C C CC C C C CC CC C C CCC C C C C C C C C C
Disordered FCC structure
Fe
Pt
N2, 580C, 30 min
Nanocrystalline thin film Ordered FCT intermetallic structure
Formation of fct structure depends on annealing conditions and Fe/Pt composition.
Intensity (normalized)
Fe70Pt30
Sun et al, Science 2000, 287, 1889; IEEE Trans. Magn. 2001, 37, 1239. Klemmer et al, Appl. Phys. Lett. 2002, 81, 2220.
Fe56Pt44
Fe48Pt42 Fe38Pt62 20
30
40
50
2θ
60
70
80
90
10000
Hc of FePt Nanoparticles
Annealed FePt N2, 580C 30 min 300 K Hc
8000
5000
4000
As-syn FePt 5 K Hc
Hc (Oe)
3500 3000
6000
Hc (Oe)
4500
4000 2000
2500
0
2000 1500
0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75
1000
X in FexPt(1-x)
500 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75
IEEE Trans. Magn., 37, 1239 (2001).
X in Fe x Pt(1-x)
500C annealed
550C annealed Gong, et al, JAP, 84, 4403 (1998). Weller, et al, IEEE Trans. Magn., 36, 10 (2000). 580C annealed
Coercivity of FePt NPs depends on Fe/Pt composition and annealing conditions. Fe-rich Fe55Pt45 films show the largest HC in evaporated, sputtered and chemically synthesized films.
Current concerns It is difficult to make FePt nanoparticles larger than 4 nm, preventing one from studying size-dependent structure transformation (from fcc to fct) and magnetic properties. Seed-mediated growth can be used to make bigger particles (up to 10 nm), but FePt compositions from one size range to another varies, making direct comparison of sizedependent properties impossible. To make fully ordered fct FePt needs high temperature annealing, but this high temperature annealing also results in particle sintering.
A: Fe, Pt order parameter B: (111) coherence length
He, 725C
J. Phys. Chem. B 2003, 107, 5419
===========================================================
Alternative efforts in making isolated ferromagnetic FePt particle arrays: 1) Making MFePt with M = Cu, Ag, Au to lower L10 phase formation temperature. e.g. Kang et al, Nano Lett. 2002, 2, 1033; JAP 2003, 93, 7178. Sun et al, JAP 2003, 93, 7337. 2) By doping more organic surfactant, the array can stand up to 800C, Momose et al, Jpn. J. Appl. Phys. 2003, 42, L1252.
=========================================================== There is no known method to align magnetic easy axis of the particles in an array.
Size/Temperature effect on Hc – Sharrock’s law 2 kB T tP f O 3 ⋅ ln H C = H O ⋅ 1 − ln 2 KU V
For a magnetically isolated and random oriented particle system
¾ A larger particle is thermally more stable than the smaller one (KuV ~ kT).
¾ For the same value of Hc at
room temperature, larger particle can have smaller Ku.
¾ Smaller Ku means lower
annealing temperature, thus less aggregation problems.
Improved synthesis Co-reduction of metal salts Composition control
= FeCl2 + Pt(acac)2 + LiBEt3H Sun et al, J. Phys. Chem. B 2003, 107, 5419. (~4 nm FePt)
= Fe(acac)3 + Pt(acac)2 + 1,2-C16H32(OH)2 Elkins et al, Nano Lett. 2003, 3, 1647. (~2 nm FePt)
= Fe(acac)3 + Pt(acac)2 + ethylene glycol
FeCl2 + Pt(acac)2
LiBEt3H
0.5mmol 0.5mmol 0.6mmol 0.4mmol
FexPt(1-x) Fe50Pt50 Fe60Pt40
Sun et al, J. Phys. Chem. B 2003, 107, 5419.
Jeyadevan et al, J. Appl. Phys. 2003, 93, 7547. Direct synthesis of partially ordered fct-FePt, but needs improvements in dispersion stability.
From core/shell Pt/Fe2O3 to FePt Pt(acac)2
Pt
Coating with Fe2O3
Pt
Reductive annealing
FePt
Teng et al, J. Am. Chem. Soc. 2003, 125, 14559. Potentially a good method to prepare large FePt nanoparticles, but it needs to solve the sintering problem under high temp (>550C) reductive annealing condition.
Improved synthesis - One step synthesis of larger FePt nanoparticles CO OC
Fe
CO CO
CO CH 3 O
O O
Pt
RCOO H/R-NH 2 CH 3
2+
O
CH 3
CH 3
•
No polyol reduction (to slow down FePt nuclei formation rate).
•
Size control – By adjusting concentration of oleic acid and Pt(acac)2. – By adjusting the reaction time at 140oC.
•
Shape control – Cubic FePt nanoparticles were obtained by adding oleic acid at room temperature and adding oleylamine immediately before the increase of temperature from 140 to 220 oC
•
Composition control – Molar ratio of Fe(CO)5/Pt(acac)2 – Molar ratio of oleic acid/oleylamine (or acidity of the solution).
Spherical FePt nanoparticles
6 nm spherical nanoparticles
8 nm spherical nanoparticles
Cubic-like 7 nm FePt nanoparticles
Discreet fct ordered 8 nm FePt nanoparticles
8 nm FePt, 560C 30 min, N2
Thermally annealed 8 nm FePt assemblies 4500 4000
N2, 600C 1h
(111)
4000
3000
Almost no sintering
Scherrer’s equation 2500 Lhkl = Kλ/βcosθ 2000 to get (111) coherence 1500 length. Intensity
Intensity
d ~ 8.5 nm
Lorentz fit to get β
3000
(111)
3500
2000
1000
1000 500
0 30
40
50
60
44
2θ
46
48
50
52
2θ
16000
N2, 700C 1h
(111)
12000
d ~ 32 nm sintering
Lorentz fit10000 to get β
8000
Intensity
Intensity
12000
14000
(111)
8000
Scherrer’s equation 6000 θ Lhkl = Kλ/βcos to get (111) coherence 4000 length. 2000
4000
0
0 30
40
50
2θ
60
46
47
48
2θ
49
50
Improved synthesis - Core/shell nanoparticles Fe(CO)5, air
Pt(acac)2 + Fe(CO)5 ------ FePt ------------- FePt/Fe2O3
6 nm FePt core with 1.2 nm Fe2O3 shell
7 nm FePt core with 1.2 nm Fe2O3 shell
Anti-sintering effect of the Fe2O3 shell FePt/Fe2O3 (7nm/1.2nm) w1/2 peak: 1.01883 Size: 10 nm
2000
10000
1000
0 25
30
35
40
45
50
55
60
2000
original Lorentz fit
48.1
6000 4000
55.4
38.3
27.9
2000
Co Kα 2 θ (°)
57.7
0 10000
48
25
30
1000
35
40
45
50
55
60
Original Lorentz fit
8000
Intensity
Intensity
w1/2 peak: 0.77231 Size: 13.0 nm
8000
Intensity
Intensity
FePt (7nm)
6000 4000 2000
0
0
45
46
47
48
49
50
51
46
47
48 kα 2θ (°) 49 Co
50
Co Kα 2 θ (°)
X Axis Title
Annealing under N2 at 650 oC for 1 hour. Fct FePt grain size is about 10 nm, no obvious sintering.
Annealing under N2 at 600 oC for 1 hour. Fct FePt grain size is increased to ~13 nm.
Hc of annealed FePt/Fe2O3 nanoparticles (Moment is not normalized)
Magnetic Moment (emu)
0.040
Annealed under forming gas at 650C for 1h, Hc ~11 kOe.
0.020
0.000
Annealed under N2 at 650C for 1h, Hc ~16 kOe.
-0.020
-0.040 -40000
-20000
0
20000
40000
Magnetic Field (Oe)
Important conclusion: Nanoparticles annealed under N2 show less degree of aggregation and large coercivity. They can be used to study single particle magnetism.
CoPt nanoparticles Co(acac)3 + Pt(acac)2 + Polyol/Hydrazine
4 nm CoPt nanoparticles
5 nm CoPt nanoparticles
Effect of composition and annealing temperature on coercivity Hc (Fe at.%=45.7)
20000
20000
15000
15000 Coercivity (Oe)
Coercivity (Oe)
Hc (at 675 oC)
10000
5000
10000
5000
0 20
30
40
50
60
70
Fe at. % in FePt
Optimum composition is Co:Pt=1:1
0 580
600
620
640
660
680
700
720
Annealing Tem perature (o)
Optimum annealing temperature (under N2 gas) is 675 oC
Hysteresis loop of CoPt nanoparticles
Magnetic Moment (emu)
0.05
0.03
0.00
-0.03
-0.05 -60000
-30000
0
30000
60000
Magnetic Field (Oe)
CoPt (Co:Pt=49.7%:50.3%) nanoparticle assembly annealed under N2 at 675 oC for 1 hr. Hc=16950 Oe
Next step: Control magnetic easy axis direction One method: Directly synthesizing fct-FePt nanoparticle dispersion Assembling these particles under a magnetic field. Add H field H-Loop of Co Particle Assembly at 300K The particles were deposited on SiO2/Si under in-plane field of 0.5T
-5
8.0x10
-5
1.5x10
out-of-plane loop
in-plane loop -5
6.0x10
-5
M (emu)
M (emu)
1.0x10 -5
4.0x10
-6
5.0x10 -5
2.0x10
0.0 -150
0.0 -100
-50
0
H-Field (Oe)
50
100
150
-150
-100
-50
0
H-Field (Oe)
50
100
150
Solvent evaporation
Next step: Control magnetic easy axis direction Alternative methods: 1) 2)
Magnetic annealing. Shape controlled assembly Cobalt nanoparticles
annealing under H
(??)
BaCrO4 nanorods
Kim et al, J. Am. Chem. Soc. 2001, 123, 4360.
Shape induced alignment seems to be more promising.
FePt NP array for ultra-high density data storage
SA film, ~ 120 nm thick 4
500 fc/mm
2
Key challenges:
0
MR signal [mV]
-2
¾To have an exchange-decoupled array.
-4 -6 -8
¾To have a uniform 2D assembly.
-10 -12
¾To control magnetic easy axis direction.
-14 -16 -18 -20
5000 fc/mm 0
5
10
15
20
x [µm]
Sun et al, Science 2000, 287, 1989.
25
Weller & Moser, IEEE Trans. Magn. 2000, 35, 4423.
Exchange-spring nanocomposites for high density magnetic energy storage Soft
Hard
Soft
¾An exchange-spring composite contains two modulated phases that are in intimate contact with one being magnetically hard and another magnetically soft.
H
¾The system can store high density magnetic energy because it has both large coercivity and high magnetic moment.
Illustration of a modulated hardsoft exchange-coupled system
¾The key for high performance composites: The size of the soft phase is at ~<10nm; Magnetic easy axis is aligned. H
N
H Kneller, E. F., Hawig, R. IEEE Trans. Magn., 27, 3588 (1991). Skomski, R., Coey, J. M. D.
H Magnetic hysteresis loops for a nonexchange coupled system (top) and an exchange-coupled system (bottom).
H Phys. Rev. B, 48, 15812 (1993).
H
S
H
Schrefl, T., Kronmüller, H., Fidler, J. J. Magn. Mag. Mater., 127, L273 (1993).
Isotropic nanocomposites
10
(A) Fe58Pt42 (B) FePt:Fe3Pt
8
B (kG)
6
m (arb. unit)
0.4 4 nm:4 nm 0.2
4 2
0.0
0
-0.2 -0.4
-2
A -60 -40 -20 0 20 40 60 H (kOe)
4 nm:8 nm
10
0.1 8
0.0 -0.1 -0.2
-8
-6
(BH)max~ 14.7 MGOe (117 kJ/m3)
-4 H (kOe)
-2
0
12
B (kG)
m (arb.unit)
0.2
(BH)max ~ 20.1 MGOe (160 kJ/m3)
6
Ms = 1050 emu/cc Hc = 2.4 T (BH)max ~24 MGOe (191 kJ/m3)
4
B -60 -40 -20 0 20 40 60 H (kOe)
H-loops for the FePt-Fe3Pt composites made from self assembly and reductive annealing of 4nm-4nm and 4nm-8nm FePt-Fe3O4 binary nanoparticles. Zeng et al, Nature 2002, 420, 395.
2 0 -10
-8
-6
-4
-2
0
H (kOe)
(BH)max, energy product, reflects the ability for a composite to store magnetic energy, the larger the better.
Conclusion remarks 1) There are a lot rooms for improvements in FePt (CoPt) synthesis. 2) Large FePt (CoPt) nanoparticles can be made by one-step chemical synthesis. 3) Core/shell structure with hard inorganic shell may be necessary for anti-sintering during high temperature annealing (an alternative anti-sintering route is via doping more organic surfactant into the assembly), and rodshaped particles may induce much needed magnetic easy axis alignment. 4) The particle assemblies with controlled interparticle interactions and magnetic easy axis direction should have great potential for future ultra-high density information storage applications. (For bio-magnetic applications, see FB-03 & FB-07.)
60 ⋅ k B ⋅ T Minimal Stable Grain Diameter DP = K 3
alloy system
Co-alloys
L10 phases
RE-TM
material
K: anisotropy in 1e7 erg/cc
Ms: saturation magnetizati on in emu/cc
Hk: anisotropy field in kOe
Dp: minimal stable diameter in nm
CoCrPtX 0.20 Co 0.45 hex Co3Pt 2.0
200-300 1400 1100
15-20 6.4 36
8-10 8 4.8
FePd FePt
1.8 6.6-10
1100 1140
33 116
5.0 2.8-3.3
CoPt MnAl
4.9 1.7
800 560
123 69
3.6 5.1
1270 910
73 240-400
3.7 2.2-2.7
Nd2Fe14B 4.6 Co5Sm 11-20
See e.g. D. Weller and A. Moser, IEEE Trans. Magn.35, 4423(1999)
Nanocomposite High Ku Materials SmCo//Cr PrCo//Cr am. CoSm
Lambeth (1991);Liu et al.(1995) Malhotra et al.(1996) Kubota, Marinero, Toney (1999)
FePt:ZrOx CoPt:SiO2 FePt:B CoPt:M (Ag, C) CoPt:C Fe/Pt/Al2O3
Coffey et al. (1995) Ichihara et al. (1998) Li, Lairson, Kwon (1999) Stavroyiannis et al.(1998) Yu, Liu, Sellmyer (1999) Bian, Laughlin... (1999)
FePt:C
Sun, Murray, Weller, Moser, Folks (2000)
Formation of ordered L1o-FePt phase HRTEM image of an individual Fe52Pt48 nanocrystal after annealing at 530 °C for 1 hour
Fourier transform of the corresponding HRTEM image
Enlarged HRTEM image of the chemically ordered L1o-FePt structure
Corresponding simulated HRTEM image
Simulated electron diffraction pattern
Dai, Sun, Wang, NanoLett., 1, 443 (2001).
1.0
Remanence curves of self-assembled Fe58Pt42 nanoparticles annealed at different conditions.
650 C 5 min N2 700 C 60 min FG
0.0
-0.5
1.0 700 C 60 min FG 600 C 60 min FG 550 C 60 min FG
-1.0 -70
-60
-50
-40
-30
-20
-10
0
H (kOe)
δM plots for self-assembled Fe58Pt42 nanoparticles annealed at different conditions
0.5
δM
m (a.u.)
0.5
0.0
0
H. Zeng et al, APL, 80, 2583(2002)
5
10 15 20 25 30 35 40 45 50
H (kOe)
Polar Kerr angle (deg)
Hc=4.6 kOe
-20
-10
0.05 0.04 0.03 0.02 0.01 0 -0.01 0 -0.02 -0.03 -0.04 -0.05
Thin film magnetics 10
20
300 K 10 K
H (kOe)
Kerr, 3 layers, 4nm particles 580C, 30min, N2 Hc=2.4 kOe
Polar Kerr angle (deg)
0.02 0.015 0.01 0.005 0 -20
-10
-0.005 0
10
20
-0.01
-40
-30
-20
-10
0
H (KOe)
-0.015 -0.02 H (kOe)
Kerr, 2 layers, 4nm particles 580C 30 min, N2
Remanence Curves 2 layers, 4 nm Fe58Pt42 particles 650C, 5 min, N2.
Magnetic recording on 3 Layer 4nm FePt Nanoparticles Computer
V
First static read/write test on 3 layer thick sample (low density of 650 fc/mm)
I
Servo Electronics
>
Slider Sample
>
X-Y Piezo Scanning Stage
to piezos
Static read/write tester by A. Moser
S. Anders
Polymer Assisted Layer-by-Layer Assembly - Assembly Thickness Control
Sun, et al, J. Am. Chem. Soc., 124, 2884 (2002)
Mass density of a 3-layer FePt assembly from X-ray reflectivity. Mike Toney, IBM Almaden
Nanoparticle Media Perspective Lubricant
Carbon Co-alloy CrV NiAl
glass substrate
35 Gbit/in2 Media Grains: ~ 9 nm Co-M
>100 Gbit/in2 Grains: ~4 nm FePt Hc 3000-5000 Oe Thickness < 10 nm ……
Terabit/in2 Regime Grains: < 5 nm FePt Each dot as a bit ……