Artifacts in ballistic Magnetoresistance measurements J' APPL' Phys' Vol'95' 2004 Lubna Rafiq Shah D

1
Artifacts in ballistic Magnetoresistance
measurements( J. APPL. Phys. Vol.95. 2004
)Lubna Rafiq ShahDepartment Of
PhysicsUniversity Of Delaware
2

• OUTLINE
• 1- What is Ballistic Magnetoresistance (BMR)
• 2- Study of factors that can mask real BMR effect
• 3- Designing Artifact-free experiments
• 4- Summary

3

• 1- Ballistic Magnetoresistance
• In the high resistance state a domain wall is
  presumed to lie in a nanocontact. If the spatial
  extent of the domain wall lt spin-flip mean
  free-path of electrons
• ? The electrons trying to cross the wall would
  have a high reflection probability due to poor
  matching of the Fermi surfaces of spin-up and
  spin-down electrons.
• ? Reflection would manifest as higher electrical
  resistance than when both sides of the
  nanocontact were magnetized in parallel.

4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7

• Experimental results reported on BMR
• 1- BMR over 3000 in Ni nanocontact at room
  temperature (PRB 66, 2002)
• 2- 100,000 BMR in stable Ni nanocontacts at
  room temperature (PRB, 67,2003)
• 3- BMR effects exceeding 500 are found at room
  temperature in a field of 7 mT in nanocontacts
  between Fe3O4 crystallites. (PRL, 87, 2001)
• 4- Ballistic magnetoresistance in a magnetic
  nanometer sized contact An effective gate for
  spintronics. The contacts exhibit
  magnetoresistance of up to 700 at room
  temperature and low applied fields .(APL,
  80,2002)

8

• WAIT..
• ARE these results real
• ????

9

• 2- Possible factors for masking real BMR
• Lengthening due to magnetostatic effect.
• Stretching or contracting due to
  Magnetostriction.
• Applied field gradient-Magnetomechanical effect.
• Bowing-out effect due to magnetostriction.

10

• Magnetostatic effect
• Consideration of field energy in the gap between
  the ends
• Magnetostatic attractive force between the ends
  of the magnetized wire
• Produces lengthening ?l in the wire of length l
  and Y is the modulus of elasticity

11

• For flat ends, ?l3nm
• For hemispherical ends, ?l1nm
• For anti-parallel magnetized wires, equal
  contraction occurs due to repulsive magnetostatic
  force.
• Between parallel and anti parallel states, change
  in separation between the ends of the wire 4-12
  nm.
• Physical dimensions of BMR1-10nm ? profound
  effect on the structure (resistivity) due to
  displacement 4-12nm

12

• Magnetostriction effect
• Magnetostriction is the changing of a material's
  physical dimensions in response to changing its
  magnetization. In other words, a magnetostrictive
  material will change shape when it is subjected
  to a magnetic field.

13
In the absence of field, Ni wire ?
multidomain In the presence of axial applied
field ? magnetize axial wire? magnetostriction
makes axial wire shorter ? pull on nanocontact?
data having low R at low field and a constant
higher R at high field.
14

• Field gradient-magnetostatic force
• Transverse wire can be mounted in such a way
  that it experiences field gradient of the magnet.
• Magnetostatic force would tend to break the
  nanocontact.

15

• Field increases? transverse wire experiences
  attraction toward the pole.
• At higher field, contact is re-established due to
  magnetostatic attraction between wires.

16

• Bowing out effect
• Magnetostriction effect? shorting of axial wire
  and lengthening of transverse.
• If the transverse wire is straight and epoxy is
  rigid
• Bowing out is required to accommodate the
  increase in length.
• Distort the nanocontact.
• Bowing out effect can occur in any direction,
  increasing or decreasing the resistance of
  nanocontact.

17

• 3- Designing Artifact-free experiments
• To suppress physical displacement due to above
  mentioned effects,
• Nanocontact electrodeposited between
  ferromagnetic films
• Nanocontacts formed by ion-beam etching
• Nanocontacts between ferromagnetic films
  connected by a pinhole through an insulating film
• Ferromagnetic nanocontacts electrodeposited on Cu
  films anchored on wafers.

18

• None of these samples gave credible evidence for
  a BMR

19

• 5- Summary
• Previous attempts to observer BMRserious
  Artifact.
• Carefully designed experiments? no evidence for
  real BMR.
• Entirely possible that there is no real BMR in
  previously published data.