1

Progress Report #5 September 1, 1999 – August 31, 2000

Nanoscale Devices and Novel Engineered Materials DOD/AFOSR MURI

Grant Number F49620-96-1-0026

Prepared by:

S.J. Pearton Department of Materials Science and Engineering

University of Florida Gainesville, FL 32611-6400

Tel: (352) 846-1086 Fax: (352) 846-1182

Email: spear@mse.ufl.edu

Participants:

University of Florida R.K. Singh

Department of Materials Science and Engineering

A.F. Hebard S. Hershfield

F. Sharifi Department of Physics

F. Ren

Department of Chemical Engineering

Florida State University S. Von Molnar

Department of Physics

University of California San Diego R.C. Dynes, F. Hellman and I.K. Schuller

Department of Physics

A.C. Kummel Department of Chemistry

Microelectronics Center of North Carolina

D. Temple and G.E. McGuire

Naval Research Laboratory R.J. Colton and S.A. Syed Asif

2

Table of Contents

Status of Effort (Major Accomplishments)..................................................................................... 3

Research Report

1. Processing Techniques for InGaAs/InAlAs/InGaAs Spin FETs ............................................. 4 2. Development of Chemically Assisted Dry Etching Methods for Magnetic Device Structures. ................................................................................................................................................ 13 3. Effects of UV Illumination on Dry Etch Rates of NiFe-based Magnetic Multilayers........... 23 4. Dry Etching Mechanism of Cu and Magnetic Materials with UV Illumination.................... 36 5. Dry Etching of BaSrTiO3 and LaNiO3 Thin Films in Inductively Coupled Plasmas ............ 50 6. Studies of the Doped Hexaborides......................................................................................... 64 7. Fabrication and Characterization of Magnetic Nano-Particles .............................................. 72 8. Interface Phenomena and Screening Length Determination in Tunnel Junction Capacitors 80 9. Hysteresis and Relaxation Behavior in GMR Multilayers..................................................... 86 10. Superparamagnetic-Ferromagnetic Transition....................................................................... 93 11. Length Scale of Magnetism ................................................................................................... 94 12. Fabrication and Characterization of Magnetotransport in Colossal Magnetoresistive Oxide

Films and Hybrid Structures .................................................................................................. 95 13. Chemically Selective Remote Chemisorption on Metals – ICl/Al(111).............................. 109 14. Abstractive Chemisorption of O2/Al(111) ........................................................................... 111 15. Measuring Mechanical Properties of Nanowires using Depth Sensing and Force Modulation .............................................................................................................................................. 112 16. Size Effects in Magnetic and Superconducting Materials ................................................... 122 17. Theory of GMR.................................................................................................................... 124 18. Ion Beam Sputter Deposition of GMR Materials ................................................................ 125

Activity Report

Technology Transitions................................................................................................... 134

Interactions and Industrial Contracts............................................................................... 135

Publications ..................................................................................................................... 136

Technical Presentations................................................................................................... 141

Personnel Supported........................................................................................................ 149

Degrees Awarded ............................................................................................................ 151

3

Status of Effort (Major Accomplishments) The fifth reporting period was a very productive one for the MURI team. Our work

continues to significantly impact progress in the spintronics and magnetic materials research area. Some of the highlights include:

(i) fabrication of a semiconductor FET with ferromagnetic contacts. (ii) study of the effects of UV illumination on the etch mechanism of magnetic

materials. (iii) long-term stability of dry etched MRAM stacks. (iv) patterning of new storage capacitor materials for DRAM. (v) novel mechanism for ferromagnetism in CaB6. (vi) magnetization reversal in magnetic nano-particle arrays. (vii) screening length determination in tunnel junctions. (viii) hysteresis and relaxation phenomena in GMR multilayers. (ix) particle size dependence of superparamagnetic-ferromagnetic transition in Ni. (x) proximity effects in magnetic/nonmagnetic systems. (xi) tunnelling magnetoresistance in GMR/CMR structures. (xii) chemisorption of halogens on metals as a precursor to etching. (xiii) mechanical properties of nanowires. (xiv) size effects in magnetic and superconducting materials. (xv) optimized ion beam deposition of GMR multilayers. (xvi) theory of maximum-possible GMR effect.

4

1. Processing Techniques for InGaAs/InAlAs/InGaAs Spin Field Effect Transistors

(F. Ren, A. Hebard and S.J. Pearton)

The term ‘Spintronics’, short for spin electronics, refers to electronic devices where the

spin direction of the electron is just as important as its charge. Magnetoresistive devices that

depend on electron spin are already in commercial use. These devices are essentially material

layers whose electrical resistance varies strongly with external magnetic field. They are in

competition with Hall detectors and induction coils to measure magnetic fields. However, spin

electronic devices, which depend on spin transport, have been slow developing. That is, devices

where the current has an imbalance of electron spins, and therefore a net magnetic moment

during operation. The advantage of such devices is that one device could provide magnetic

storage of information and electronic readout. The spin field effect transistor proposed by Datta

and Das in 1989 is such a device.

The fabrication of the spin FET has proved difficult. The primary problem is that many

studies have used permalloy (Ni.8Fe.2), which has been shown to be a poor spin injector (~1%).

Another major device consideration is spin transport length. Device fabrication will be limited

by how long the electron can ‘remember’ its spin during transport through semiconductor

material. Quantum well structures like InGaAs/InAlAs have been shown to enhance spin

transport length. Long spin transport lengths are necessary so that typical photolithography

processes can be used (minimum feature size ~1 µm). This will allow large scale commercial

realization, with efficient spin injection.

We expect that efficient spin injection will become available with the use of technologies

like Mn based dilute magnetic semiconductors. Therefore, the primary purpose of this work is to

outline a reliable fabrication technique using conventional photolithography techniques on an

5

InGaAs/InAlAs/InGaAs system. It will be shown that the inclusion of an ohmic etch mask level

adds great flexibility to the injection geometries that can be tested, with a single mask set. For,

simplicity Fe contacts have been deposited, but the same mask set (and techniques) may be used

for regrowth of semiconductor spin injectors; using SiO2 as deposition mask. This should allow

future work to concentrate solely on spin injection issues rather than processing.

The field effect transistors were fabricated using conventional lithography techniques on

an InGaAs/InAlAs/InGaAs, system on an InP substrate. The top layer was a highly doped 4 ×

1018 cm-3 400 Å InGaAs cap layer, followed by an 3 × 1018 cm-3 300 Å InAlAs donor layer, and

a 35 Å undoped InAlAs buffer layer, and a 5000 Å undoped InGaAs channel layer on an InP

substrate.

Prior to processing the sample was cleaned in 7:1 buffered oxide etch (BOE) for 1 min,

and then rinsed in DI water. The sample was then dehydration baked for 3 min at 120 °C. For

mesa isolation Shipley 1818 (2 µm) was spun at 4000 RPM for 30 seconds, and then hot plate

baked at 90 °C for 90 seconds. Backside clean was followed by edge bead removal. For edge

bead removal the sample was exposed for 30 seconds, then developed for 30 seconds in MF 321,

followed by a DI rinse. Pattern exposure was for 4 seconds (107 mJ/ cm2), and develop was for

90 seconds in MF 321. Photoresist descum in LFE barrel etch at 850 mTorr, 5 watts, 4 min. The

sample was postbaked at 120 °C for 90 seconds. Next was another BOE clean. The mesa was

then etched in 1:4:45 H3PO4:H2O2:H2O (50 Å/sec etch rate on GaAs) with 10% over etch past

InAlAs. Resist removal was with acetone.

Ohmic trench etch followed next, in an effort to directly contact the 2D gas and metal.

The sample was cleaned with BOE, followed by a dehydration bake. The resist, AZ 5214 (1.5

µm), was spun at 5,000 RPM for 30 seconds, and baked at 110°C for 90 seconds. Next was a

6

back side clean and edge bead removal. For edge bead removal the sample was exposed for 45

seconds, and developed for 45 seconds in 1:5 AZ 400k:H2O, followed by a DI rinse. Pattern

exposure was for 7 seconds (190 mJ/cm2), and developed in 1:5 AZ 400k for 60 seconds. The

sample was then rinsed in DI water. Photoresist descum was in LFE barrel etch. The sample was

then etched in 1:4:495 H3PO4:H2O2:H2O (5 Å/sec etch rate on GaAs). For process method 1, the

sample then had 700 Å of Fe and 200 Å of Cr deposited using a low energy ion beam sputtering

(Figure 1 and 2). For process method 2, however, the ohmic etch step and metal deposition were

two separate lithography steps (figure 3). However, the photolithography process for both steps

is the same as the previous, except that the ohmic etch photoresist descum was followed by a

120°C post bake for 90 seconds.

The photolithography process for the gate etch and deposition used AZ5206 (1/2 µm)

photoresist. The sample was cleaned with a 20/1 H2O/NH4OH solution, which was followed by

a dehydration bake. The photoresist was spun at 5,000 RPM for 30 seconds. It was then baked

at 90°C for 90 seconds. Edge bead removal exposed the sample for 30 seconds, with a 60

second develop in 1:5 AZ400K. Pattern exposure was for 2.85 seconds (80 mJ), and developed

for 60 seconds in 1:5 AZ400K. Photoresist descum was again in the LFE barrel etch, followed

by postbake. Since the total donor and buffer layer thickness is only 335 Å thick, a highly

selective etch for InAlAs over InGaAs was necessary for reproducible fabrication. Such an etch

(selectivity ~250) was provided by the following:

1. 1gm adipic acid powder per 5ml DI

2. Add NH4OH until the pH value of the solution is 5.5

3. Add H2O2 (30%) to the pH adjusted solution at a volume ratio of about 6%

7

The gate metallization was Ti/Pt/Au in 200/300/1500 Å. It is important to note that for proper

semiconductor etch profile (for liftoff), that the gate width be oriented along the long axis of the

sample’s oval defects.

The final step was bond pad deposition. The photolithography procedure was an image

reversal process using AZ5214. The sample was cleaned in BOE, then dehydration baked. The

photoresist was spun at 4000 RPM for 30 seconds, and then baked at 110°C for 90 seconds. Next

was backside clean. Edge bead removal exposed the wafer for 45 seconds followed by 45

seconds develop in 1:1.4 MIF 312:H2O. Pattern exposure was for 2.45 seconds (66 mJ). The

wafer was then baked at 110°C for 45 seconds and then flood exposed for 45 seconds. The

pattern was developed in 1:1.4 MIF 312:H2O for 60 seconds. This was followed by LFE

photoresist descum and BOE rinse. Finally, 2000 Å of Au was deposited for the pad metal.

To try to enhance spin injection into the 2D gas of the FET we recess etched the ohmic

contact pads, using the ohmic mask set (process method 1). However, even though the total etch

depth was less than 1000 Å, the lateral undercut of the mask during wet etching forbid ohmic

metal contact to the 2d gas or doped cap layer or donor layer. Process method 1 is shown in

Figure 1.1. The gap in metal deposition was verified by SEM micrographs shown in Figure 1.2

and by source drain current voltage characteristics shown in Figure 1.4.

To promote ohmic contact, the next set of samples had separate levels for ohmic etch and

ohmic metal deposition (process method 2). The ohmic contact level overlapped the ohmic etch

level by ~2 µm. This forced the ohmic metal to contact the 2d gas donor and cap layer. Process

method 2 is shown in Figure 1.3. This result is verified by a much larger total current flow

shown in Figure 1.4.

8

The finished devices using this process method are shown below in Figure 1.5. The

finished devices were 120 µm wide with a 1.5 µm gate. The resultant Ids gate modulated curve

is shown in Figure1.6.

In Figure 1.6 it is seen that there is ~.5 v offset voltage before the onset of current flow.

This is interesting, because it is expected that most of the current will be injected through the

highly doped cap layer; which was ohmic to the Fe contacts. The TLM measurements made to

the Fe contacted to the cap layer indicated an ohmic contact resistance of 5.2 10-5 ohm-cm2. It

may be that the rectifying characteristics of the contact are due to etch of the cap layer beneath

the contact during the rapid etching of the selective gate etch, or oxidation of the ohmic contact.

It may not be desirable to inject spins through the cap layer due to spin scattering events

caused by ionized impurity scattering and interfaces. By modifying the separate ohmic etch and

ohmic metal deposition steps, we can now test several spin injection geometries with a single

mask set that has the extra ohmic etch level. These methods are outlined in Figure 1.7.

Future Fe based contacts should have a thick layer of Au (or oxygen diffusion barrier)

deposited immediately after Fe deposition. Since Fe based contacts have been shown to be

poor spin injectors, these process methods may be best applied to dilute magnetic

semiconductor regrowth in the contact region. The same mask should be useful for both

metal contact and dilute magnetic semiconductor contacts. Careful mask design will allow

testing of several injection geometries with the inclusion of only one extra mask level; the

ohmic etch mask level. This extra ohmic etch level allows plasma etching (and therefore

damage) to be eliminated from the device processing. It also eliminates undercut of the

photoresist during etching that might inhibit contact formation.

9

MESA

FE CONTACTETCH WITH OHMIC

Fe CrDEPOSITIO

SELECTIVGATE RECESSAND Ti/Pt/A

TrencPenetratiointoLaye

PADDEPOSITIO

Wet EtchPhotoresi

Wet EtchPrevents Fewith 2d

CrFe

Au

Photoresi

Cap

Dono

Space BuffeSubstrat

InGaA

InAlA

InGaAsInP

Figure 1.1. Process method 1 for Spin FET fabrication. Use of same mask for ohmic metal

etch and deposition leads to undercut of photoresist layer that prevents ohmic metal contact to 2d gas.

FeFe

2d Gas

Figure 1.2. SEM micrographs of process method 1. The Fe does not contact the 2d gas.

10

Cap

Donor

SpacerBuffer

Substrate

MESA ISOLATION

FE CONTACT TRENCHETCH WITH PAD METALMASK

Fe AND Cr DEPOSITIONWITH OHMICMASK

SELECTIVEGATE RECESS ETCH AND DEPOSITION

Ti/Pt/Au

TrenchPenetrationinto BufferLayer

InGaAs

InAlAs

InGaAs

InP

PAD METAL DEPOSITION

Wet Etch UndercutsPhotoresist

2d Gas NowContacted by Fe

Cr

Fe

Au

Photoresist

Figure 1.3. Process method 2, for spin FET fabrication. The use of separate ohmic etch and

ohmic metal deposition masks allows ohmic metal to contact 2d gas layer.

0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 3 .50

1

2

3

4

5

P r o c e s s M e th o d 1 P r o c e s s M e th o d 2

I ds (m

A)

V d s ( V )

Figure 1.4. Comparison of source drain current between process method 1 and process

method 2.

11

Source

Drain

Gate

Pad Metal2000Å Au

Pad Metal

MesaFe ContactTrench

Figure 1.5. SEM micrographs of finished device.

0 .0 0 .5 1 .0 1 .5 2 .00

5

1 0

1 5

2 0

G a t e V o l t a g e 0 V - .5 V - 1 V - 1 .5 V

I d (mA

)

V d ( V )

Figure 1.6. FET drain current vs. voltage curves for differing applied gate voltages. The

ohmic contact metal is 700 Å Fe.

12

BufferSubstrate

Cap

Donor

Spacer

CrFe

Figure 17. Top left, no ohmic etch level, ohmic mask used for selective etch and metal deposition. Top right, Ohmic etch level used to etch to buffer layer, ohmic mask metal level used to deposit metal. Bottom, ohmic metal used to selectively etch cap layer, ohmic etch level used to etch to buffer layer, ohmic metal level repatterned to deposit metal.

13

2. Development of Chemically Assisted Dry Etching Methods for Magnetic Device Structures

(S.J. Pearton and F. Sharifi)

The push to continually increase bit densities in magnetic storage devices places

emphasis on techniques for patterning submicron metallic multilayer structures. The component

materials within these multilayers may include NiFe, NiFeCo [both are used for structures based

on the giant magnetoresistance (GMR) effect], NiMnSb (a Heusler alloy potentially useful in

advanced spin-valve structures) or the col-lossal magnetoresistance (CMR) materials

LaCaMnO3, LaSrMnO3, and PrBaCaMnO3. A general problem when processing magneto-

resistive materials is their relative invola-tility in conventional dry etching techniques such as

reactive ion etching (RIE).

Practical etch rates may be achieved under high-density plasma (HDP) conditions, where

the high ion flux is able to promote desorption of halogenated etch products. Both inductively

coupled plasma (ICP) and electron cyclotron resonance (ECR) systems have proven capable of

providing the necessary ion-to-reactive-neutral ratio (>0.02). We have completed an

examination of the etch rates of the three basic classes of materials (i.e., NiFe, NiMnSb, and the

perovskite-based CMR materials) in different plasma chemistries and also measured the long-

term magnetic and structural stability of dry etched magnetic multilayer stacks. In the latter case,

we used magnetic random access memory (MRAM) elements as our test vehicle.

The NiFe and NiMnSb layers were deposited on Si (100) substrates by direct current

magnetron sputtering from composite targets. Typical layer thicknesses were 5000Å. Liquid

delivery metalorganic chemical vapor deposition using 2,2,6,6-tetramethyl-3, 5-hepanedionato

(TMHD) precursors [i.e., La(TMHD)3 ,Sr(TMHD)2 ,Mn(TMHD)3 , and Ca(TMHD)2] was

employed to deposit films of La0.41 Ca0.59 MnO3 on Al2 O3 (0001) single crystal substrates at

14

700°C. The precursors were transported by N2 carrier gas, with direct injection of O2 and N2O as

oxidants. Films of Pr0.65 Ba0.05Ca0.3MnO3 were deposited on Si (100) in a pulsed laser ablation

system (248 nm KeF laser, 10 kHz, 2.5 J cm-2) energy density) at a substrate temperature of

700°C and an O2 partial pressure of 250 mTorr. All of the CMR films were in the range 1500–

2500Å thick.

Finally, the MRAM structures consisted of the following layers deposited on 300Å of

SiNx on 8500Å of SiO2 on Si: 80Å NiFeCo, 15Å CoFe, 35Å Cu, 15Å CoFe, 80Å NiFeCo, 200Å

Ta, 550Å TaN, and 800Å CrSi. The deposition was performed by Ar 1 ion-assisted sputtering. A

3000-Å-thick SiO2 mask was patterned by SF6/Ar RIE as the etch mask for subsequent high-

density plasma patterning of the metal layers.

The etching was performed in either Plasma-Therm 790 ICP or Plasma-Therm SLR 770

ECR reactors. In both systems the samples were thermally bonded to a radio frequency powered

(13.56 MHz) chuck which was He-backside cooled. The respective HDP sources were powered

up to 1000 W at either 2 MHz (ICP) or 2.45 GHz (ECR). The gases were injected directly into

the sources through electronic mass flow controllers at a typical load of 15–20 standard cubic

centimeters per minute. We investigated halogen- (Cl2, BI3, BBr3, ICl, IBr, SF6), CH4/H2- and

CO/NH3-based mixtures since these cover the full range of possible etch products (i.e., metal

chlorides, bromides, iodides or fluorides; metalorganics or carbonyls).

Magnetic properties before and after plasma etching were determined using

superconducting quantum interference device magnetometry (Quantum Design MPMS-5S00) at

4.2 K. Scanning electron microscopy (SEM) was used to examine sidewall smoothness on

etched features.

15

Under RIE conditions (i.e., zero watts HDP source power) we invariably saw net

deposition on the samples upon exposure to halogenated mixtures, or essentially no etching with

CH4/H2 and CO/NH3 mixtures. Examination of the halogen-plasma exposed surfaces, by Auger

electron spectroscopy revealed large concentrations of chlorinated residues. Since the

halogenated etch products have larger lattice constants than their pure metal constituent and the

products are essentially involatile under RIE conditions, then one observes a buildup of these

species as shown schematically in Figure 2.1.

Table 2.1 shows a compilation of results for NiFe etching in the different chemistries

investigated. The highest rates were achieved with Cl2/Ar, where the role of the inert gas additive

is to provide ion-assisted desorption of the chlorinated etch products. We found that the mass of

these inert species also played a role, with Xe providing slightly faster rates than either Ar or He

addition. The rates with these Cl2 -based mixtures were approximately a factor of two faster than

with pure Ar sputtering. Bromine or iodine-based plasma chemistries produced lower rates than

with chlorine, and were close to Ar sputter rates. Both CH4/H2/Ar and SF6/Ar led to extremely

low etch rates, while the CO/NH3 mixture had a slight degree (40%) of chemical enhancement.

It has been suggested that the role of the NH3 is to suppress dissociation of the CO so that

carbonyl etch products can form, but an alternative explanation might be that atomic hydrogen

scavenges surface carbon species and prevent carbonization of the NiFe surface. The fact that

Cl2-based plasma chemistries produce the fastest rates for NiFe (and plasma chemistries produce

the fastest rates for NiFe (and NiFeCo) is consistent with the higher vapor pressures of the

chlorinated etch products relative to their brominated or iodidated counterparts.

Table 2.2 shows the corresponding comparisons for NiMnSb. In their cases the Cl2 -based

mixtures produce excellent etch rates (1500–5000Å min-1 for both ICP and ECR tools), but the

16

fastest rates were achieved with SF6 /Ar mixtures. By sharp contrast, NF3 /Ar showed net

deposition rather than etching for source powers >100 W or at high NF3 percentages. The surface

under these conditions showed strong Mn enrichment and were oxidized, with an underlying Sb-

deficient region. With all of the plasma chemistries, careful attention had to be paid to the

removal of the native oxide prior to the commencement of etching to avoid the presence of a

relatively long incubation time.

For the CMR materials, we did not observe any chemical enhancement in etch rate with

any of the plasma chemistries discussed (Table 2.3). The etching was dominated by physical

sputtering under all conditions investigated, with etch yields typically <0.1 and relatively high

ion energies (>150 eV) needed to initial removal of material.

A key issue with the use of corrosive gas mixtures for etching metallic multilayers is that

of postetch stability of the patterned structures. Severe corrosion and delamination of the films is

observed in the absence of preventive measures. We examined use of several different postetch

treatments. The first was simply rinsing the samples in deionized water immediately upon

opening the chamber (which is contained within a N2 dry box). The samples were then

thoroughly dried with filtered N2. In the other three methods, various in-situ plasma cleaning

procedures were examined. After Cl2 /Ar etching was complete, the chamber was evacuated for

15 min, and then a 30 mTorr discharge of either H2, O2 or SF6 (500 W source power, 5 W chuck

power) was used to clean the residual chlorine for 10 min prior to removal of the samples from

the reactor. In these cases, no H2 O rinsing was performed. It should be pointed out that all of

these cleaning procedures have been employed previously for removing etch residues after Cl2-

based plasma etching of Al interconnects in Si microelectronics.

17

Figure 2.2 shows the magnetization of each of the samples over a period of

approximately six months. In each case the samples were simply stored in air between the

measurements and no special precautions were taken to prevent corrosion. Each of the cleaning

procedures produces samples with extremely stable magnetic characteristics. This is also

reflected in their appearance. Figure 2.3 shows SEM micrographs of patterned MRAM elements

three months after Cl2/Ar etching and postetch cleaning. There is no indication of corrosion on

any of the samples and the sidewalls are smooth (to the resolution used in the photos). There is

no indication of striations often observed on dry etched features. Note, however, that in the case

of O2 plasma cleaned samples there was a slight decrease in the magnetization per unit volume

relative to the samples treated in water or H2 or SF6 plasmas. A possible reason for this is that the

feature sidewalls become more oxidized than with other treatments, leading to a degradation in

magnetic properties.

A comprehensive survey of etching results for magnetic materials in different plasma

chemistries has produced the following conclusions:

(i) The optimum chemistry for NiFe is Cl2/Ar, for NiMnSb is SF6/Ar, while no

chemical enhancement of etch rates for CMR oxides was observed.

(ii) Postetch rinsing in H 2 O or in-situ plasma cleaning with H2, O2 or SF6 discharges

are all effective treatments for removing chlorine etch residues. Of these, only O2

plasma exposure appears to degrade the magnetic properties of MRAM stacks.

Once the residues are removed, there is no change in magnetic or visual properties

over a period of ~six months (extent of our study).

(iii) The CO/NH3 chemistry, while being noncorrosive, produces relatively slow etch

rates and is only suitable for patterning of thin (<1000Å) structures.

18

TABLE 2.1. Comparison of plasma chemistries for NiFe etching.

Chemistry

Typical Etch Rates

Corrosive

Comments

10Cl2/5Ar 600Å min-1 ICP

>1000Å min-1 ECR Yes Chemical enhancement of 100%

Etch rate with Xe>Ar>He

13CO/2NH3 250Å min-1 ICP 500Å min-1 ECR

No Chemical enhancement of ~20%–40% CO2 less effective than CO

5CH4/10H2/5Ar <100Å MIN-1 ICP, ECR No Slower than Ar sputtering 10SF6/5Ar <100Å min-1 ICP, ECR No Slower than Ar sputtering 10BI3/5Ar 500Å min-1 ICP, ECR Yes Less effective than Cl2/Ar 10BBr3/5Ar 200Å min-1 ICP, ECR Yes Slower than Ar sputtering 10ICl/5AR 500Å min-1 ICP, ECR Yes Excellent surface morphology 10IBr/5Ar 500Å min-1 ICP, ECR Yes Excellent surface morphology

Table 2.2. Comparison of plasma chemistries for NiMnSb etching.

Chemistry

Typical Etch Rates

Corrosive

Comments

10SF6/5Ar >10,000Å min-1 ICP, ECR No Selectivity ≥20 over Al2O3

10NF3/5Ar 300Å min-1 ICP, ECR No Narrow process window

10Cl2/5Ar 3,000Å min-1 ICP, ECR Yes Selectivity ≥5 over Al2O3

10BCl3/5Ar 5,000Å min-1 ICP, ECR Yes Selectivity ≥5 over Al2O3

Attacks native oxide

10ICl/5Ar 1500Å min-1 ICP, ECR Yes Threshold ion energy 120 eV

10IBr/5Ar 1500Å min-1 ICP, ECR Yes Threshold ion energy 230 eV

19

Table 2.3. Comparison of plasma chemistries for CMR etching.

Chemistry

Typical Etch Rates

Corrosive

Comments

10SF6/Ar 500Å min-1 LaCaMnO3 ICP, ECR No No chemical enhancement

5CH4/10H2/5Ar 200Å min-1 LaCaMnO3 ICP, ECR No No chemical enhancement

10Cl2/5Ar 1500Å min-1 LaCaMnO3 ICP, ECR

900Å min-1 LaSrMnO3 ICP

300Å min-1 PrBaCaMnO3 ICP

Yes Physically dominated under all conditions for all three materials

10BI3/5Ar 500Å min-1 LaCaMnO3 ICP, ECR Yes Etch yield <0.1 threshold ion energy <100 eV

10BBr3/5Ar 500Å min-1 ICP, ECR Yes Threshold ion energy ~150 eV

20

Figure 2.1. Schematic of involatile layer build-up during Cl2-based RIE of NiFe at room

temperature.

NiFeCo substrate

NiFeCo substrate in RIE gases at 25°C (e.g.Cl2)

NiFeCo substrate

NiClX, FeClXdeposition

ioneutral

21

Figure 2.2. Magnetization of MRAM structures, either unetched or etched in Cl2/Ar plasmas

and subsequently cleaned in water, or in H2, SF6 or O2 plasmas, as a function of subsequent storage time in room ambient.

0 40 80 120 1600.0

0.2

0.4

0.6 Magnetization vs. Time

H2 cleaning DI water rinsing SF6 cleaning O2 cleaning Control

M/v

ol (e

mu/

mm

3 )

Time (days)

22

Figure 2.3. SEM micrographs of MRAM elements after etching in ICP Cl2/Ar plasmas and

subsequent cleaning in H2, SF6 or O2 discharges, or by H2O rinsing. The micrographs were taken 3 months after these processes, with the samples having been stored in room ambient.

H2 H2O

SF6

O2

23

3. Effects Of UV Illumination On Dry Etch Rates Of NiFe-Based Magnetic Multilayers

(S.J. Pearton)

There is a strong interest in the development of plasma etching processes for magnetic

multilayer structures of the type used in sensors, magnetic random access memories or read/write

heads for data storage. There are two basic plasma chemistries that have been reported to etch

NiFe and NiFeCo under high ion density conditions, namely Cl2 and CO/NH3. The etch rates are

still low (< 500 Å/min) and are limited by desorption of the reaction products. Recently, several

groups have reported that ultra-violet light irradiation during Inductively Coupled Plasma (ICP)

etching of Cu in Cl2-based discharges lowered the activation energy for etching, and also

enhanced the desorption of the CuCl products. In that case, the UV light was assumed to be

absorbed by the CuCl, promoting nonthermal desorption.

In this paper we report on the effects of UV illumination on ICP etch rates of NiFe and

NiFeCo in Cl2/Ar and CO/NH3 discharges. In the latter chemistry there was no measurable

enhancement in etch rates under a wide range of plasma conditions. In the case of Cl2/Ar,

however, the UV irradiation decreased the NiFe etch rate and more chlorine residues were

detected on the NiFe surface. Only in the case of Cl2/Ar etching of NiFeCo were enhancements

in removal rate obtained with UV irradiation.

The Ni0.8Fe0.2 and Ni0.8Fe0.13Co0.07 layers were deposited on Si(100) substrates by dc

magnetron sputtering from composite targets. Typical layer thicknesses were 5000 Å. Some of

the samples were masked with Apiezon wax for etch rate measurements. The etch depths were

measured by stylus profilometry after removal of the mask in acetone. The morphology and

composition of unmasked samples was examined by Atomic Force Microscopy (AFM) and

Auger Electron Spectroscopy (AES), respectively.

24

The etching was performed in a Plasma-Therm 790 ICP reactor, with the samples

thermally bonded to a rf powered (13.56 MHz) He backside-cooled chuck. The gases were

injected directly into the ICP source (2 MHz, 1000 W) through electronic mass flow controllers

at a total load of 15 standard cubic centimeters per minute (sccm). An unfiltered 400 W Hg arc

lamp was installed on top a 1 inch diameter quartz window on top of the ICP source (~ 20 cm

from the sample position) and provided illumination of the samples during plasma etching. We

believe any sample heating due to the lamp is minimal (< 10 oC) because the samples were

thermally bonded to a Si carrier wafer and the resist mask showed no evidence of thermal

degradation or flow.

Table 3.1 shows thermochemical data for the potential metal chloride or metal carbonyl

etch products for NiFe and NiFeCo in Cl2 or CO/NH3 plasmas. From this data we can calculate

the Gibbs free energies of reactions of Ni, Fe and Co with atomic or molecular chlorine, and with

CO and CO2. There are several important features of this data in Table 3.2. First, the reaction of

the metals is more favorable with atomic chlorine than with Cl2, which emphasizes the need for

efficient dissociation of the feedstock gas in the plasma source. Second, CO is more reactive

with the metals than is CO2, as we have previously reported in a comparison of the two gases.

We emphasize that in a plasma etching environment there will be a strong ion-assisted

component to the etch mechanism and the thermodynamic data provides only a guide to the

reaction pathways.

Figure 3.1 shows the effect of source power (top) and rf chuck power (bottom) in NiFe

etch rates in 10Cl2/5Ar discharges at fixed process pressure (2 mTorr). The source power

controls ion flux and efficiency of plasma dissociation, while the chuck power controls the

25

average ion energy. Note that at this plasma composition the use of UV illumination actually

retards the etching and in fact promotes net deposition on the sample.

Examination of the samples etched either with or without UV illumination by AES

showed more chlorine-related residues in the former case. Figure 3.2 shows AES surface scans

of samples etched in 10Cl2/5Ar discharges at either 500 W source power, 100 W rf chuck power

(top, left and right) or 750 W source power, 200 W rf chuck power (bottom, left and right). The

samples etched without UV illumination (top and bottom right) have much smaller Cl signals on

their surfaces. The effect of the UV in this case is clearly to enhance formation of FeClx and

NiClx species on the surface, but not to enhance their desorption. We have previously found that

a process window exists for Cl2/Ar etching of NiFe in which the formation of the chloride

reaction products must be balanced with their ion-assisted desorption.

The fact that the UV illumination alters their process window is evident from the data in

Figure 3.3. At a lower Cl2 concentration in the discharge (5Cl2/10Ar), the etch rate is now

positive (i.e., no deposition) for all source and chuck powers. However, there are still no

conditions where the UV provides an enhancement in etch rate. Moreover, the etched surface

morphologies were not improved by UV illumination, as shown in the AFM scans of Figure 3.4.

By sharp contrast to the results for NiFe, we did observe significant etch rate

enhancements for NiFeCo with UV irradiation. Figure 3.5 shows the effect of source power (top)

and of chuck power (bottom) on NiFeCo etch rates in 10Cl2/5Ar discharges, either with or

without UV illumination. We obtained a maximum enhancement of approximately a factor of 3

at relatively high source powers or chuck powers.

Figure 3.6 shows similar results for NiFeCo etching at lower Cl2 concentration in the

discharge (5Cl2/10Ar). In this case there was no measurable change in etch rate with UV

26

irradiation. This is consistent with the discussion earlier for NiFe, in that the process window

where etch product formation and desorption are balanced can be shifted by altering the ion-to-

reactive neutral ratio either through changing plasma parameters or adding the UV illumination.

Figure 3.7 shows the influence of the UV light on NiFe etch rates as a function of either

source power (top) or rf chuck power (bottom). There was no significant difference in etch rate

as a result of the illumination. The latter shows that the rates are linearly dependent on chuck

power, indicating a physically dominated etch mechanism, but are also a strong function of ion

flux.

Similar data for NiFeCo are shown in Figure 3.8. Once again there is no significant

difference in etch rate as a result of the UV illumination. For both NiFe and NiFeCo there were

also no differences in surface morphology and root-mean-square roughness due to light

irradiation. with typical values of 2-5 nm under all conditions.

For Cl2/Ar ICP etching of NiFe and NiFeCo, we found that UV illumination either

reduced or had no effect on NiFe etch rates, while it did provide significant enhancement for

NiFeCo at certain plasma conditions. We cannot yet provide an explanation for the enhancement,

but in analogy for the Cu results it may involve transformation of CoClx products to a more

volatile form. No change in etch rates of NiFe and NiFeCo were observed with UV illumination

during CO/NH3 etching. Since magnetic multilayers based on the Giant Magnetoresistance

(GMR) effect are often comprised mainly of NiFeCo/Cu/NiFeCo, the use of UV illumination

may prove useful in patterning these structures.

27

Table 3.1. Thermochemical data for potential etch products (Ref. 11).

Species

∆∆∆∆H0

f.298 K(kJ/mol)

S0298 K (J/mol K)

∆∆∆∆G0

f.298 K (kJ/mol)a

FeCl2 (s) -341.833 117.947 -302.342

FeCl3 (s) -399.405 142.336 -333.930

CoCl2 (s) -312.545 109.266 -269.647

NiCl2 (s) -304.930 98.157 -258.779

Fe(CO)5 (s) -766.090 337.078 -696.975

Fe(CO)5 (s) -727.849 439.286 -389.207

Ni(CO)4 (s) -631.784 319.560 -588.980

Ni(CO)4 (s) -601.576 415.507 -587.378

Co(CO)4 (s) -562.100 -337.442 -535.024

Table 3.2. Thermodynamic data for possible etch reactions.

Fe(s) + 2Cl(g)↔FeCl2(s) ∆G0r = -512.9 (kJ/mol)

Fe(s) + Cl2(g)↔FeCl2(s) ∆G0r = -302.3 (kJ/mol)

Fe(s) + 5CO(g) ↔Fe(CO)5 (1) ∆G0r = -11.2 (kJ/mol)

↔Fe(CO)5 (g) ∆G0r = -3.4 (kJ/mol)

Fe(s) + 5CO2(g) ↔Fe(CO)5(1)+2.502(g) ∆G0r = 1274.9 (kJ/mol)

↔Fe(CO)5(g)+2.502(g) ∆G0r = 1282.7 (kJ/mol)

Ni(s)+2Cl(g)↔NiCl2(s) ∆G0r = -359.4 (kJ/mol)

Ni(s)+Cl2(g)↔NiCl2(s) ∆G0r = -258.8 (kJ/mol)

Ni(s)+4CO(g)↔Ni(CO)4 (1) ∆G0r = -40.3 (kJ/mol)

↔Ni(CO)4(g) ∆G0r = -38.7 (kJ/mol)

Ni(s)+4CO2(g)↔Ni(CO)4(1)+2O2(g) ∆G0r = 988.6 (kJ/mol)

↔Ni(CO)4(g)+2O2(g) ∆G0r = 990.2 (kJ/mol)

Co(s)+2Cl(g)↔CoCl2(s) ∆G0r = -380.3 (kJ/mol)

Co(s)+Cl2(g)↔CoCl2(s) ∆G0r = -269.6 (kJ/mol)

Co(s)+4CO(g)↔Co(CO)4(g) ∆G0r = 13.6 (kJ/mol)

Co(s)+4CO2(g)↔Co(CO)4(1)+202(g) ∆G0r = 1042.5 (kJ/mol)

28

Figure 3.1. Etch rates of NiFe, either with or without UV illumination, in 10Cl2/5Ar, 2 mTorr

ICP discharges, as a function of either ICP source power (top) or rf chuck power (bottom).

ICP Source Power (W )0 200 400 600 800 1000

Etc

h R

ate

(Å/m

in)

-1500

-1000

-500

0

500

dc B

ias (

-V)

150

250

350

450

W / UVW /O UVdc Bias

NiFe10Cl2/5Ar200W rf2m Torr

Etch

Deposition

rf Chuck Power (W )50 150 250

Etc

h R

ate

(Å/m

in)

-400

-200

0

200

400

dc B

ias (

-V)

140

180

220

260

300

W / U VW /O U Vdc Bias

N iFe10C l2/5A r500W IC P2m Torr

EtchEtch

D eposition

29

Figure 3.2. A

ES surface scans of NiFe (top) or N

iFeCo (bottom

) after etching in 10Cl2 /5A

r, 2 m

Torr discharges, either with (tip and bottom

left) or without (top and bottom

right) U

V illum

ination under conditions of 500 W source pow

er, 100 W rf chuck

power (top, left and right) or 750 W

source power, 200 W

rf chuck power

(bottom, left and right).

10 86420

N(E)/E

400800

12001600

2000K

INETIC

ENER

GY

, eV

400800

12001600

2000K

INETIC

ENER

GY

, eV

1086420

N(E)/E

NiFe

500W IC

P 100W

rf W

/ UV

NiFe

500W IC

P 100W

rf W

/O U

V

Cl

Cl

O

O

CC

Ni

Ni

Fe Fe

4080

120160

200K

INETIC

ENER

GY

, eV

10 86420

N(E)/E

1086420N(E)/E

NiFeC

o 750W

ICP

200W rf

W/ U

V

Cl

O

C

Ni

Fe

400800

12001600

2000 K

INETIC

ENER

GY

, eV

NiFeC

o 750W

ICP

200W rf

W/O

UV

Cl

O

C

Ni

Fe

30

Figure 3.3 Etch rates of NiFe, either or without UV illumination, in 5Cl2/10Ar, 2 mTorr ICP discharges, as a function of either ICP source power (top) or rf chuck power (bottom).

ICP Source Power (W)0 200 400 600 800 1000

Etch

Rat

e (Å

/min

)

0

200

400

600

dc B

ias (

-V)

200

300

400

150

250

350

W/ UVW/O UVdc Bias

NiFe5Cl2/10Ar200W rf2mTorr

rf Chuck Power (W)100 200 30050 150 250

Etc

h R

ate

(Å/m

in)

0

200

400

600

800

dc B

ias (

-V)

120

160

200

240

280

W/ UVW/O UVdc Bias

NiFe5Cl2/10Ar500W ICP2mTorr

31

Figure 3.4 AFM scans of NiFe after etching in 5Cl2/10Ar, 500 W source power discharges,

either with or without UV illumination, at different rf chuck powers.

4 8 µm

4 8 µm

4 8 µm

4 8 µm

4 8 µm

4 8 µm

W/O UV W/ UV

100W rf

250W rf

200W rf

RMS: 2.883 nm RMS: 5.415 nm

RMS: 5.195 nm RMS: 6.043 nm

RMS: 9.031 nm RMS: 6.973 nm

32

Figure 3.5 Etch rates of NiFeCo, either or without UV illumination, in 10Cl2/5Ar, 2 mTorr

ICP discharges, as a function of either ICP source power (top) or rf chuck power (bottom).

ICP Source Power (W )0 200 400 600 800 1000

Etch

Rat

e (Å

/min

)

0

400

800

1200

dc B

ias (

-V)

150

250

350

450

W / UVW /O UVdc Bias

Etch

Deposition

NiFeCo10Cl2/5Ar200W rf2mTorr

rf Chuck Power (W )50 150 250

Etc

h R

ate

(Å/m

in)

0

200

400

600

800

dc B

ias (

-V)

100

200

300

W / UVW /O UVdc Bias

NiFeCo10Cl2/5Ar500W ICP2mTorr

33

Figure 3.6 Etch rates of NiFeCo, either or without UV illumination, in 5Cl2/10Ar, 2 mTorr

ICP discharges, as a function of either ICP source power (top) or rf chuck power (bottom).

ICP Source Power (W )0 200 400 600 800 1000

Etc

h R

ate

(Å/m

in)

0

200

400

600

800

dc B

ias (

-V)

200

300

400

150

250

350

W / UVW /O UVdc Bias

NiFeCo5Cl2/10Ar200W rf2mTorr

rf Chuck Power (W)100 200 30050 150 250

Etch

Rat

e (Å

/min

)

-200

0

200

400

600

800

dc B

ias (

-V)

120

160

200

240

280

W/ UVW/O UVdc Bias

NiFeCo5Cl2/10Ar500W ICP2mTorr

Etch

Deposition

34

Figure 3.7. Etch rates of NiFe, either or without UV illumination, in 10CO/5NH3, 2 mTorr

ICP discharges, as a function of either ICP source power (top) or rf chuck power (bottom).

ICP Source Power (W )0 200 400 600 800 1000

Etc

h R

ate

(Å/m

in)

0

100

200

300

400

dc B

ias (

-V)

250

350

200

300

400

W / U VW /O UVdc Bias

NiFe10CO /5NH 3

200W rf2m Torr

rf Chuck Power (W )100 200 30050 150 250

Etc

h R

ate

(Å/m

in)

0

100

200

300

dc B

ias (

-V)

200

300

400

150

250

350

W / UVW /O UVdc Bias

NiFe10CO/5NH 3

500W ICP2mTorr

35

Figure 3.8. Etch rates of NiFeCo, Either with or without UV illumination, in 10CO/5NH3, 2

mTorr ICP discharges, as a function of either ICP source power (top) or rf chuck power (bottom).

ICP Source Power (W)0 200 400 600 800 1000

Etc

h R

ate

(Å/m

in)

0

100

200

300

400

dc B

ias (

-V)

250

350

200

300

400

W/ UVW/O UVdc Bias

NiFeCo10CO/5NH3

200W rf2mTorr

rf Chuck Power (W )100 200 30050 150 250

Etc

h R

ate

(Å/m

in)

150

250

100

200

dc B

ias (

-V)

200

300

400

150

250

350

W / UVW /O UVdc Bias

NiFeCo10CO/5NH 3

500W ICP2mTorr

36

4. Dry Etching Mechanism Of Copper And Magnetic Materials With UV Illumination (S.J. Pearton)

In recent years several research groups have studied dry etching of copper for the next

generation of metallization in the semiconductor industry, focusing on development of new etch

techniques to increase etch rate. They used Cl2 plasmas with or without photon sources using

ultraviolet (UV) laser, UV lamp, illumination and IR light. In contrast to conventional dry

etching that requires relatively high temperatures (> 200oC) in order to produce practical etch

rates, they all reported substantial enhancement of etch rates at low temperatures. Among them

Choi and Han first reported high etch rates of about 3000 Å/min at room temperature with Cl2

discharges in an Inductively Coupled Plasma (ICP) system.

Magnetic materials such as NiFe and NiFeCo are widely used in sensors, magnetic

random access memories (MRAMs) or read/write heads for data storage industry. Due to the

relative inertness of these materials there is a strong interest in the development of high density

plasma etching processes for them. There are two basic plasma chemistries for the etching of

NiFe and NiFeCo under ICP conditions, namely Cl2 and CO/NH3. However, the etch rates are

still low (< 500 Å/min) and are limited by desorption of the etch products such as NiClx, FeClx

and CoClx. Cho et al. first reported the effect of UV illumination on the etch rates of the NiFe-

based magnetic materials

Since the etch mechanism with UV illumination has not been studied in detail, in this

paper we propose an etch mechanism of copper and magnetic materials with UV irradiation

based on subprocesses occurring in the Cl2 -ICP etching system. We also carried out ICP etching

of NiFe and NiFeCo in Cl2/Ar discharges with or without UV illumination. We found that the

chlorination of copper surface is enhanced with UV irradiation and the absorption of UV photons

by metal chlorides is critical to enhance the removal rate of chlorides. The proposed etch

37

mechanism of copper showed good agreement with observed data determined by mass

spectrometry, taken from the literature.

4.1. Etch Mechanism of ICP Etching of Cu with UV Illumination

There are likely five subprocesses involved in etching of copper with UV illumination: 1)

photo-dissociation of Cl2 in gas phase, 2) surface chlorination, 3) absorption of UV photons by

reaction products, 4) photo-assisted removal of reaction products, and 5) Gas-phase reactions

between desorbed species.

(a) Photo-dissociation of Cl2

In addition to formation of Cl radicals by electron-collision in the bulk plasma, more Cl2

molecules are dissociated by collision between photons and molecules:

Cl2 + hν → 2Cl (1)

The above reaction is readily occurred because the bond strength of Cl2 is 2.5 eV and the UV

photon energies are 2 - 4 eV. Hence, compared to the plasmas without UV illumination, the

photo-dissociation reaction provides chlorine-enriched environment, and the reactive chlorine

radicals easily take part in surface reactions.

(b) Surface chlorination

When the copper surface is exposed to UV radiation, electrons are ejected by photo-

electron effect and are captured by chlorine radicals near the surface, leading to chlorination of

the copper surface:

Cu(s) + hν → Cu+ + e (2)

xCl(g) + e → xCl-(g) (3)

Cu(s) + xCl(g) → CuClx(s), ∆Gor = - 226 (kJ/mol) (4)

38

In contrast to dry etching without UV illumination, UV photons promote the chemistry at the

copper surface, resulting in fast deposition of metal chlorides on the surface with low activation

energy. This chlorination reaction is induced by photons and the reaction rate is thus a strong

function of photon flux or UV intensity. Since the unreacted Cu surface absorbs UV photons and

is readily chlorinated by the above reaction, most sites of the copper surface will be chlorinated,

leading to the formation of stoichiometric copper chloride.

Under chlorine-enriched conditions, copper chloride may further react with chlorine

radicals to form CuCl2(s):

CuCl(s) + Cl(g) → CuCl2(s), ∆Gor = - 146 (kJ/mol) (5)

The chlorination reactions of (4) and (5) may produce the copper chlorides having the Cl content

equal to or greater than stoichiometric ratio, i.e. x ≥ 1. If there are no UV photons involved, the

Cl concentration will be strongly dependent on reaction time. The surface chlorination also

weakens Cu-Cu bond strength, resulting in lowering etch threshold.

(c) Absorption of UV radiation by reaction products

The enhancement of etch rate with UV illumination indicates that the UV photons play an

important role in etching mechanism. The clue to this suggestion is the optical properties of the

reaction products, CuClx. Tables 4.1 and 4.2 show the optical constants of some materials

available in terms of reflectivity (Table 4.1) and adsorption depth (Table 4.2). Compared to

metal copper, CuCl has much smaller reflectivity (or longer adsorption depth). This indicates

copper chloride absorbs most UV photon energies, and in turn the photons excite electrons of the

reaction products so that the bond strength becomes weaker. This phenomenon may play a key

role in increasing the etch rates of copper. If a metal chloride has a low absorption capacity of

UV light, the effect of UV illumination will be less significant. Hence, the overall etch process of

39

metals with UV illumination is limited by absorption of UV radiation, which is determined by

optical properties of the metal chlorides.

(d) Photo-assisted removal of metal chlorides at the surface

As soon as the copper chlorides are formed, they absorb UV radiation, resulting in

excitation of valence-electrons in the CuClx layer, weakening bond strengths of CuClx and

CuClx-Cu. This could lead to rupture of surface bonds and subsequent desorption. It is

noteworthy that CuCl has a direct bandgap of 3.26 eV at 300 K. The UV photon energies in the

range 2-4 eV can also be utilized to sublime the copper chloride [1.6 eV for Cu3Cl3(g) and 2.2

eV for CuCl(g)]. Hence, the possible photo-assisted removal process of the chlorinated surface

can be described as:

CuCl(s) + hν → Cu(g) + Cl(g) (6)

3CuCl(s) + hν → Cu3Cl3(g) (7)

CuCl(s) + hν → CuCl(g) (8)

Cu(s) + hν → Cu(g) (9)

Photon-sputtering represented by Eq. (9) could be occurring on the unchlorinated or etched

surface due to lower bond energy of Cu-Cu (1.83 eV) than UV photon energies.

The photon-assisted etch reactions of (6)-(9) are greatly affected by UV intensity because

they do not occur without formation of copper chloride and the formation rate is a function of

photon flux, implicating a higher etch rate with higher UV intensity.

(e) Gas-phase reactions between desorbed species

Some of the desorbed molecules tend to capture electrons and form negative ions, and

react with reactive radicals such as Cl and Cu. There are many possible reactions among

desorbed species, radicals and ions in gas phase. They can be summarized as:

40

CuCl(g) + e → CuCl-(g) (10)

CuCl2(g) + e → CuCl2-(g) (11)

CuCl (g) + Cl(g) → CuCl2 (g) (12)

CuCl-(g) + Cl(g) → CuCl2(g) + e (13)

CuCl (g) + Cu(g) → Cu2Cl(g) (14)

CuCl2-(g) + Cu(g) → Cu2Cl2(g) + e (15)

CuCl2-(g) + Cu(g) + Cl(g) → Cu2Cl3(g) + e (16)

CuCl2-(g) + Cu(g) + Cl(g) → Cu2Cl3(g) (17)

Cu2Cl2(g) + Cl(g) → Cu2Cl3(g) (18)

Cu2Cl2(g) + Cu(g) → Cu3Cl2(g) (19)

Cu2Cl3(g) + Cu(g) → Cu3Cl3(g) (20)

The gas-phase reactions under plasma and UV illumination conditions indicate that the most

favored forms of product gas are CuCl2 (Eqs. (11) - (13)) and Cu2Cl3 (Eqs. (16)-(18)). This may

be confirmed with examining the relative peak intensities of the observed mass distributions of

species using mass spectrometry. Although we may consider photo-assisted dissociation of

CuCl(g) and Cu3Cl3(g), it seems unlikely due to the high bond energy of Cu-Cl, 3.97 eV. The

ionization energies of copper chlorides are also quite high: 10.7 eV for CuCl+, 9.6 eV for

Cu2Cl2+, and 9.7 eV for Cu3Cl3

+. Hence the ionization of copper chloride gases does not occur

under the UV illumination conditions.

4.2. Etch Mechanism of Magnetic Materials with UV Illumination

Successful etching of NiFe and NiFeCo with Cl2 plasmas has been reported. However, due to

the relative involatility of the etch products such as NiClx, FeClx and CoClx the attainable etch

41

rates are quite low (≤ 500 Å/min). The etch rates of these materials in a high density plasma

reactor are function of ion flux, ion energy and plasma composition. To examine the effect of

UV illumination on etch rate we have to know the optical properties of the magnetic materials

and reaction products. However, their optical properties are not available yet. There is however,

a substantial increase in etch rate with UV illumination over etching without UV irradiation.

Table 4.3 summarizes the mass distributions of desorbed gaseous products observed from

thermal etching and photon-induced etching of copper. These mass distributions are normalized

to the Cu3Cl3 intensity. It is seen that for thermal desorption, the major gases are Cu3Cl3 at 860 K

and CuCl at 920 K, but no CuCl2 and Cu2Cl3 are observed. However, UV laser induced etching

showed that the dominant gas species were Cu, Cu2Cl, Cu2Cl3 and Cu3Cl2 with laser fluence of

0.26 J/cm2 per pulse with 532 nm, and Cu, Cu3Cl2 and Cu3Cl3 with 0.66 J/cm2 and 355 nm. The

observation of copper signal is attributed to photon-sputtering due to the strong laser energy. By

contrast, Kwon et al. reported CuCl2 and Cu2Cl3 as the dominant gases observed from ICP

etching of copper with UV illumination regardless of chlorine content in copper chlorides. It is

also interesting to note that the desorbed gases are Cl, CuCl, CuCl2, Cu2Cl, Cu2Cl2, Cu2Cl3,

Cu3Cl2, and Cu3Cl3 in all cases of photon-assisted etching. Furthermore, no observation of CuCl2

and Cu2Cl3 during the thermal desorption indicates that their presence in the etching with UV

illumination is not originated from cracking of Cu3Cl3 by the ionizer of the mass spectrometer. It

seems clear that CuCl2 and Cu2Cl3 are produced in the ICP etching of copper with UV

illumination. Hence, these previous results overall support the etch mechanism proposed in this

paper.

Kwon et.al. also confirmed that the copper chloride layer formed with UV illumination

had a higher chlorine content than stoichiometry (i.e. x > 1.0) regardless of reaction time, and it

42

was composed of CuCl(s) and CuCl2(s). This is contrary to the much lower chlorine content

obtained without UV illumination. However, this result is attributed to the fact that the copper

surface is easily chlorinated under UV illumination because UV photons promote the chemistry

at the surface and lower the activation energy. Choi and Han reported the activation energy of

0.12 eV, which is much lower than the energy of 1.6 eV required to sublime CuCl(s) to

Cu3Cl3(g). This confirms that the dry etching with UV illumination is not a simple thermal

desorption, but a nonthermal etch mechanism due to the presence of UV photons.

Figure 4.1 shows the effect of UV intensity on etch rate and the chlorine content in the

copper chloride, adapted from the experiments by Kwon et al. They measured etch rates and Cl

contents with varying UV intensity at Cl2/N2 = 1.5, 2 mTorr, 500 W ICP source power and room

temperature. The etch rates increased linearly with UV intensity up to certain point and then

remained almost constant, while the Cl concentration in the copper chloride was independent of

UV intensity and maintained at 1.2-1.3, implying coexistence of CuCl(s) and CuCl2(s). The

insensitivity of chlorine atomic ratio to the UV intensity indicates that the UV photon energies

used in their experiment are enough for the surface chlorination to occur and to form CuClx

having x > 1.0. The photon-assisted etch of copper chloride layer occurs very fast since the

CuClx layer has low reflectivity but high absorption depth (see Table 2), and absorbs most UV

photons as soon as the layer is formed. The increase in etch rate with UV intensity is believed to

occur because the photon-assisted removal rate of copper chlorides is faster than the deposition

rate of CuClx, indicating that the deposition rate of copper chloride controls the overall etch

process. However, at higher UV intensity the CuClx formation rate is also increased due to the

increased photon flux, and is thus in equilibrium with the photon-assisted etch rate.

43

Figure 4.2 shows the effect of ICP source power in our reactor on etch rates of Ni0.8Fe0.2

(top) and Ni0.8Fe0.13Co0.07 (bottom) with or without UV illumination in Cl2 plasmas at 10 sccm

Cl2/5 sccm Ar, 200 W rf chuck power, 2 mTorr and room temperature. In these experiments an

unfiltered 400 W Hg arc lamp was used for UV irradiation. Details of the experiment are

described elsewhere. There is net deposition observed on NiFe, indicating the rate of formation

of metal chlorides is greater than their removal rates. This result also implies that reaction

products such as NiClx and FeClx are not absorbing UV photons. The increase in the formation

rate of metal chlorides is attributed to the increased chlorine radicals with UV illumination and

the chemistry promoted at the surface by photons. In contrast to NiFe, NiFeCo showed an overall

increase in etch rate with UV illumination, especially at moderate ICP source powers (500-800

W). This may be attributed to two factors: 1) lower binding energy of NiFeCo than NiFe (for

example, see the Fe-Co phase diagram; addition of Co to Fe lowered the melting point of FeCo

alloy), and 2) greater absorption capacity of UV photons by CoClx than by NiClx and FeClx. The

latter is unlikely because Ni, Fe and Co are elements all in the same period and same group so

that the alloys and metal chlorides have similar optical properties (Tables 2 and 3), and

furthermore the atomic ratio of Co (i.e., 0.07) is too small to affect the overall optical properties

of etch products. However, to clearly understand the effect of UV illumination on the dry etching

of magnetic materials, more systematic studies, in particular, the dependence of etch rates on

optical properties of NiClx, FeClx and CoClx has to be examined.

With the UV-enhanced process, very clearly defined features can be patterned into Cu, as

shown in the scanning electron micrographs of Figure 4.3. The etching was performed at 75°C in

Cl2 / Ar in this case, whereas to achieve similar rates without UV illumination required etch

temperatures ≥ 150°C.

44

An etch mechanism with UV illumination was proposed to better understand the ICP

etching of copper and magnetic materials. The photo-dissociation of Cl2 provides a chlorine-

enriched environment near the surface, and UV photons promote the chemistry at the copper

surface, leading to fast deposition of metal chlorides on the surface with low activation energy.

The proposed model predicts that surface chlorination under UV irradiation produces copper

chlorides having the Cl content equal to or greater than stoichiometric ratio, i.e. x ≥ 1. The

overall etch process of metals with UV illumination is limited by absorption of UV radiation,

which is determined by optical properties of the metal chlorides. The proposed etch mechanism

showed gaseous etch products are CuCl, CuCl2, Cu2Cl, Cu2Cl2, Cu2Cl3, Cu3Cl2, and Cu3Cl3,

verified with reported mass spectrometry data,8 and the dominant gas species are CuCl2 and

Cu2Cl3 in the etching with UV illumination. The Cl2-ICP etching of magnetic materials with UV

illumination showed no enhancement in etch rate for NiFe, but a substantial enhancement for

NiFeCo mainly due to lower binding energy of NiFeCo. However, to clearly understand the

effect of UV illumination on the dry etching of magnetic materials, more systematic studies have

to be carried out in terms of the absorption of UV photons by NiClx, FeClx and CoClx.

45

Table 4.1. Reflectivity, R, at 298 K (wavelength, λ).

λλλλ

(nm) Cu CuCla Ni Fe Co Ni0.8Fe0.2b Ni0.8Fe0.13 Co0.07

b

300 0.36 0.42 0.35 0.41 0.41 0.41

350 0.40 < 0.01 0.44 0.40 0.45 0.43 0.43

400 0.51 0.46 0.48 0.49 0.46 0.46

500 0.59 < 0.05 0.54 0.50 0.56 0.52 0.53

600 0.60 0.61 0.52 0.60 0.59 0.60

700 0.96 0.62 0.53 0.62 0.60 0.61

R = [(n-1)2 + κ2]/[(n+1)2 + κ2], where n and κ are refractive and absorption indexes at 298 K, respectively, and obtained from Refs. 31 and 33. a) From Ref. 1. b) Estimated using n = Σxi ni and κ = Σxi κi, where xi is atomic fraction of element i. Table 4.2. Absorption depth, α−1

(nm), at 298 K (wavelength, λ). λλλλ

(nm) Cua CuClb Ni Fe Co Ni0.8Fe0.2c Ni0.8Fe0.13 Co0.07

c

300 13.9 10.8 13.2 12.1 11.2 11.1

350 14.0 60 12.4 12.6 12.5 12.4 12.3

400 14.4 13.8 12.2 13.0 13.5 13.5

500 15.6 500 14.4 14.2 13.3 14.4 14.3

600 16.2 14.0 16.1 13.8 14.4 14.2

700 13.3 15.5 18.0 14.9 16.0 15.8

Absorption coefficient (nm-1), α = 4 πκ/λ

a) From Ref. 33. b) From Ref. 1. c) Estimated using n = Σxi ni and κ = Σxi κi, where xi is atomic fraction of element i.

46

Table 4. 3. Mass distributions of desorbed gases from thermal etching and photon-induced etching of copper at various conditions

Mass Distribution

Thermal etching28 UV laser1 UV lamp8

Species

860 K 920 K 0.26J/cm2

532 nm

0.66J/cm2

355 nm

CuCl0.4

340 nm

CuCl1.2

340 nm Cl

Cu

CuCl

CuCl2

Cu2Cl

Cu2Cl2

Cu2Cl3

Cu3Cl2

Cu3Cl3

9

20

51

-

83

10

-

25

100

27

68

514

-

82

14

-

27

100

28

273

89

121

284

10

282

141

100

-

47

8

5

6

13

35

60

100

410

-

100

2330

130

100

550

130

100

350

-

100

2350

280

150

2380

270

100

47

Figure 4.1. The effect of UV intensity on etch rate of copper and chlorine content in the

copper chloride at Cl2/N2 = 1.5, 2 mTorr, 500 W ICP source power and room temperature, adapted from Refs. 6 and 8.

UV Intensity (arb. units)0 1 2 3

Etch

Rat

e (Å

/min

)

0

1000

2000

3000

4000

Chl

orin

e C

onte

nt in

CuC

l x

0

1

2

Etch Rate with UVCl content

48

Figure 4.2. The effect of ICP source power on etch rates of NiFe (top) and NiFeCo (bottom) with or without UV illumination in Cl2 plasmas at 10 sccm Cl2/5 sccm Ar, 200 W rf chuck power, 2 mTorr and room temperature.

ICP Source Power (W )0 200 400 600 800 1000

Etc

h R

ate

(Å/m

in)

-1500

-1000

-500

0

500

dc B

ias (

-V)

150

250

350

450

W / U VW /O U Vdc B ias

N iFe10C l2/5A r200W rf2m Torr

E tch

D eposition

ICP Source Pow er (W )0 200 400 600 800 1000

Etc

h R

ate

(Å/m

in)

0

400

800

1200

dc B

ias (

-V)

150

250

350

450

W / UVW /O UVdc Bias

Etch

Deposition

NiFeCo10C l2/5Ar200W rf2m Torr

49

Figure 4.3. SEM micrographs of features etched into Cu layers on Si, using ICP Cl2/Ar discharges

50

5. Dry Etching of BaSrTiO3 and LaNiO3 Thin Films in Inductively Coupled Plasmas

(S.J. Pearton, R.K. Singh, F. Ren and F. Sharifi)

High dielectric constant materials are under intense development as a replacement for

SiO2 as gate materials in metal oxide field-effect transistors or as storage capacitors in advanced

dynamic random access memories (DRAM). Another application is for decoupling capacitors in

device packages. The leading candidates are TaOx and (Ba,Sr)TiO3 (BST), based on their

dielectric breakdown strength, area capacitance and measured leakage current densities. While

dry etching process are well-developed for conventional SiO2-based dielectric structures, there is

much less known about the etching characteristics of the newer materials. In this paper we report

on high density plasma etching of thin film (Ba,Sr)TiO3 and LaNiO3 (LNO) in two different

chemistries, namely Cl2/Ar and CH4/H2/Ar. In a conventional DRAM capacitor technology,

doped polysilicon is generally employed as the electrode material. However with oxide

ferroelectrics this is not feasible because of interfacial reactions to form SiOx, which reduces the

effective dielectric constant of the capacitor stack. There have been two basic classes of

electrode materials employed to date, namely those based on elemental metals, predominantly Pt

(Ir and Ru have also been reported) or those based on metallic oxides such as IrO2, RuO2 or

high-Tc superconductors.(7) The metallic oxides have a potential advantage in improving the

fatigue performance of capacitors. In this work we have chosen LaNiO3 as the metallic oxide for

use with BST films, since it displays several advantages as an electrode material.

Our etching experiments have focussed on comparing the Cl2 and CH4/H2 chemistries for

achieving practical etch rates for the two materials. In the former case the etch products would be

expected to be metal chlorides and O2, while in the latter case metalorganics, metal hydrides,

water vapor and O2 may be expected to form. Under high density conditions, the etching

51

reactions are generally strongly ion-assisted so that fully chlorinated products need not be

formed before ion impingement helps desorb surface species. We find that highly anisotropic

pattern transfer is possible in both materials using ICP etching, with Cl2/Ar providing much

higher removal rates.

The sample preparation has been described in detail elsewhere, but in brief, both

materials were deposited on Si substrates using pulsed laser deposition (KrF excimer laser, 5

Hz pulse frequency) at an O2 partial pressure of 200-300 mTorr and a substrate temperature of

650 °C. Pressed powder targets were used in both cases. The LNO appears to grow

predominantly with (110) orientation even on (100) Si, while the BST is polycrystalline.

LNO/BST/Si capacitor structures produced from companion samples exhibited an interface state

density of ~7x1011 eV·cm-2 without any post-deposition H2 annealing. The leakage current

density was ~10-8 A·cm-2 at 5x104 V·cm-1.

Etching was performed in a Plasma Therm 790 Inductively Coupled Plasma

reactor. The plasma is sustained in a 3-turn, cylindrical geometry source operating at 2 MHz and

powers from 500-1000 W. The samples were thermally bonded on an rf-(13.56 MHz) biased

chuck, at powers of 50-350 W. These conditions produced dc self biases of approximately –50 to

–340 V. In general, dielectric materials have relatively high bond energies and it is necessary to

employ high ion energies during etching to break the bonds so that etch products may form. The

average ion energy is the sum of the dc self-bias through which the ion is accelerated, plus the

plasma potential, which is ~22 eV in our particular system. Two different gas chemistries were

investigated, namely CH4/H2/Ar and Cl2/Ar. Electronic grade gases comprising these mixtures

were injected into the ICP source through mass flow controllers at a total flow rate of either 15

(for Cl2/Ar) or 20 (for CH4/H2/Ar) standard cubic centimeters per minute (sccm).

52

The etch depths were obtained from stylus profilometry after removal of the mask

material, which was Apiezon wax except when we wanted to examine etch anisotropy using

scanning electron microscopy (SEM), in which case we employed lithographically patterned

photoresist (AZ4620E, cured at 150 °C) as a mask. Etch yield (number of atoms of the target

material removed per incident ion) was calculated from a semi-empirical model developed for

this reactor that employs ion fluxes measured by the Langmuir probe technique.

Cl2/Ar Plasma Chemistry - Figure 5.1 (top) shows the influence of rf chuck on the etch

rates of both BST and LNO films at fixed pressure (2 mTorr), source power (750 W) and Cl2/Ar

flow rates (10 sccm/5 sccm). The etch rate of BST increases with the higher ion bombardment

energy up to approximately 250 W rf chuck power and decreases thereafter. This is a commonly

observed trend with high density plasma etching and is usually ascribed to ion-assisted

desorption of the adsorbed chlorine neutrals before the etch products can form. In the case of

LNO we do not observe the decrease in etch rate at high rf chuck powers, suggesting the amount

or stability of the adsorbed chlorine is different than for BST. The etch yields (Figure 5.1

bottom) for both materials are low, and emphasize that the etching is dominated by physical

sputtering.

The role of source power (which controls ion flux) on the material etch rates is shown in

Figure 5.2. Increasing the source power suppresses the dc self-bias because of the higher

conductivity of the plasma, and this leads to two competing effects, namely an increase in ion

flux but a decrease in ion energy. This competition is reflected in an initial increase in BST etch

rate, followed by a decrease when the self-bias falls below approximately –270 V. The latter is

consistent with the data of Figure 5.1. Once again the behavior of the LNO is different, with a

continuing increase in etch rate over the range of source powers we investigated. The etch yield

53

(Figure 5.2 bottom) of LNO does not change as much with ion flux as does etch rate, suggesting

the increased etch rates are mostly due to a higher sputter rate.

CH4/H2/Ar Plasma Chemistry--We did not examine this chemistry as carefully as we did

with Cl2/Ar, because it was quickly apparent the etch rates with CH4/H2/Ar were extremely low.

Figure 5.3 shows the effect of rf chuck power on the etch rates of both BST and LNO-the trend

shows that the etching is sputter-limited, with very low yield (≤ 0.04). There is no apparent

chemical contribution to the etching with this gas mixture, with results similar to those obtained

with pure Ar plasmas.

The effect of source power is shown in Figure 5.4. Once again, the rates are low (≤

100 Å·min-1) under all conditions, and ≥ 25 ions are required for removal of one atom of both

materials. One problem with trying to use this plasma chemistry for etching of high bond

strength materials is that polymer deposition from the CH4 may act to shield the surface from ion

bombardment.

Comparison of Plasma Chemistries: Figure 5.5 shows a comparison of the etch rates

obtained for both BST and LNO as a function of source power in the two plasma chemistries.

The maximum rates with Cl2/Ar are roughly one order of magnitude higher than with

CH4/H2/Ar. This has consequences in terms of mask erosion when etching device features,

because since the CH4/H2/Ar shows no chemical contribution to the etch mechanism, there will

be no selectivity over common mask materials such as SiO2, SiNx or photoresist. By contrast

there is some ion-assisted chemical component to the etching with Cl2.

Since the etching is ion-driven under all conditions for both materials, highly anisotropic

features can be formed provided mask erosion is minimized. Figure 5.6 (top) shows a cross-

sectional SEM view of a narrow (≤1 µm) feature created in BST using a 10 Cl2/5 Ar discharge

54

for 9 mins (5 mTorr, 750 W source power, 200 W rf chuck power). In this case a 0.5 µm thick

SiNx layer and 1 um resist bilayer was used as a mask and all the resist and the SiNx 0.15 µm of

the SiNx was lost during the etch process. The side walls are slightly sloped from facetting of the

edges of the mask during exposure to the plasma. Figure 5.6 (center and bottom) show features

etched into BST (center) or LNO (bottom) using the same plasma conditions as above, but with a

single 7 µm thick photoresist mask (AZ4614). About one-third of the resist remained at the

completion of the etching. This is a simpler masking procedure than the dielectric/resist bilayer,

and is still able to produce anisotropic pattern transfer. We were unable to achieve acceptable

etch anisotropy with the CH4/H2/Ar chemistry because of severe mask facetting that led to

sidewall slopes ≥ 30° from vertical.

Figure 5.7 shows some typical AFM scans of BST and LNO surfaces before and after

exposure to either Cl2/Ar or CH4/H2/Ar discharges. For CH4/H2/Ar etching the surfaces became

slightly rougher, as evidenced by the change in root-mean-square (RMS) roughness. This may

result from non-equal rate removal of one or more of the lattice constituents (probably the lighter

Ti). By sharp contrast, the surfaces of both materials exhibit a degree of smoothing after

exposure to the Cl2/Ar plasma, as seen in the raw data of Figure 5.7 and the tabulated RMS

values of Figure 5.8. This can result from the angular dependence of ion-milling rate in

physically dominated chemistries, whereby sharper surface features are removed faster than flat

features.

In summary, two common semiconductor plasma chemistries etching, namely

Cl2/Ar and CH4/H2/Ar, have been examined for dry etching of thin films of (Ba,Sr)TiO3 and

LaNiO3. The etching in both chemistries is physically-dominated, but only Cl2/Ar produces

reasonable removal rates. Although not presented here, under typical conditions of 750 W of ICP

55

source power and 250 W of rf chuck power (-275 V chuck bias) in our tool, the etch selectivity

for BST and LNO over Si is ~16 (BST) and ~7 (LNO) when using Cl2/Ar. The surfaces of both

materials become smoother with exposure to these discharges, and highly anisotropic pattern

transfer can be achieved using simple resist masks.

56

Figure 5.1. Etch rates (top) and etch yields (bottom) for BST and LNO in 10Cl2/5Ar, 5mTorr, 750W source power discharges, as a function of applied rf chuck power.

RF chuck power (W)0 100 200 300 400

Etc

h ra

te (Å

/min

)

0

200

400

600

800

1000

DC

bia

s (-V

)

100

200

300

400BST(E/R) LNO(E/R) DC

750W ICP2m Torr10Cl2/5Ar

RF chuck power (W)0 100 200 300 400

Etc

h yi

eld

0.00

0.15

0.30

0.45

Ion

flux

(x10

16 c

m-2

· s-1

)

1.49

1.50

1.51

1.52

1.53BST(Yield) LNO(Yield) Ion flux

750W ICP2m Torr10Cl2/5Ar

57

Figure 5.2. Etch rates (top) and etch yields (bottom) for BST and LNO in 10Cl2/5Ar, 5mTorr,

250W rf chuck power discharges, as a function of source power.

RF chuck power (W)50 150 250

Etch

rat

e (Å

/min

)

0

50

100

DC

bia

s (-V

)

0

100

200

300

BST(E/R) LNO(E/R) DC

750W ICP5mTorr5CH4/10H2/ 5Ar

RF chuck power (W)50 150 250

Etc

h yi

eld

0.00

0.03

0.06

Ion

flux

(x10

16 c

m-2

· s-1

)

1.2

1.5

1.8BST(Yield) LNO(Yield) Ion flux

750W ICP5mTorr5CH4/10H2/5Ar

ICP source power (W)200 400 600 800 1000

Etch

rat

e (Å

/min

)

0

300

600

900

DC

bia

s (-V

)

200

250

300

350

400BST (E/R) LNO(E/R) DC

250W RF5mTorr10Cl2 / 5Ar

58

Figure 5.3. Etch rates (top) and etch yields (bottom) for BST and LNO in CH4/10H2/5Ar,

5mTorr, 750W source power discharges, as a function of rf chuck power.

ICP source power (W)200 400 600 800 1000

Etch

yie

ld

0.0

0.1

0.2

0.3

0.4

Ion

flux

(x10

16 c

m2 .

s-1)

0

1

2

3BST(Yield) LNO(Yield) Ion flux

250W rf2m Torr10 Cl2/5Ar

59

Figure 5.4. Etch rates (top) and etch yields (bottom) for BST and LNO in CH4/10H2/5Ar,

5mTorr, 250W rf chuck power discharges, as a function of rf chuck power.

ICP source power (W)400 600 800 1000

Etch

rat

e (Å

/min

)

0

100

200

300

DC

bia

s (-V

)

200

250

300

350

400BST(E/R) LNO(E/R) DC

250W RF5mTorr5CH4/10H2/5Ar

ICP source power (W)400 600 800 1000

Etc

h yi

eld

0.01

0.02

0.03

0.04

0.05

Ion

flux

(x10

16 c

m-2

· s-1

)

0.5

1.0

1.5

2.0

2.5BST(Yield) LNO(Yield) Ion flux

250W RF5m Torr5CH4/10H2/5Ar

60

Figure 5.5. Comparison of BST and LNO in Cl2/Ar and CH4/H2/Ar ICP discharges (5mTorr, 250W rf chuck power), as a function of source power.

ICP source power (W)200 400 600 800 1000

Etc

h ra

te (Å

/min

)

0

300

600

900

1200 BST (Cl2/Ar) LNO(Cl2/Ar)

BST(CH4/H2/Ar) LNO(CH4/H2/Ar)

250W RF5m Torr

61

Figure 5.6. SEM micrographs of features etched into BST (top and center) using 10Cl2/5Ar, 5mTorr, 750W source power, 250W rf chuck power discharges using either a resist/SiNx bilayer mask (top) or a single layer resist mask (center). About 0.35µm of the SiNx remains in the top micrograph. The SEM at bottom shows features etched into LNO using similar plasma conditions and a resist mask, which has been removed.

62

Figure 5.7. AFM scans of BST and LNO surface before and after dry etching in either Cl2/Ar

or CH4/H2/Ar .

RMS=7.4

RMS=5.0

RMS=0.7

RMS=2.1

RMS=0.8

RMS=0.6

BST LNO

Control

250 rf 1000 ICP CH4/H2/Ar

250 rf 750 ICPCl2/Ar

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

um um

um um

um um

nm 100

nm100

nm 100

nm100

nm 100

nm100

63

Figure 5.8. RMS roughness measured over 5x5 µm2 area for BST and LNO samples before and after etching in either Cl2/Ar or CH4/H2/Ar discharges.

Etch condition

RM

S ro

ughn

ess (

nm)

0

2

4

6

8

BSTLaNiO

(Control) (CH4/H2/Ar) (Cl2/Ar)

64

6. Studies of the Doped Hexaborides

(F. Sharifi)

Over the last year, our efforts have focused on a new class of ferromagnetic materials that

have been discovered in studies of doped hexaborides of the alkaline earths. A surprisingly

robust weak ferromagnetism has been found both in electron and hole doped CaB6 at a carrier

concentration of approximately 0.005 e/unit cell. These materials raise not only interesting

fundamental questions regarding the nature of magnetic ordering, but suggest as well unusual

device possibilities.

The alkaline earth and light rare earth hexaborides are highly refractory materials that

crystallize in a simple cubic CsCl-type array of regular B6 octahedra and metal atoms. These

materials possess a low carrier density, high carrier mobility, and long mean free path. Band

structure calculations suggest that the hexaborides of the divalent metals are semi metallic, with

very small direct overlap of a conduction and a valence band of different symmetry at the X-

points in the Brillouin zone. In the presence of ferromagnetism, this band structure may result in

a large degree of spin-polarized carriers at the Fermi level. The internal Zeeman splitting of the

upper band can enhance the direct overlap of the majority band and cause the minority band to

be semiconducting [Figure 6.1]. Magnetization studies with bulk single crystals have shown that

in CaB6, both the hole-doped compound Ca1-δB6 and the electron doped compound

Ca0.995La0.005B6 are ferromagnetic. In electron-doped Ca0.995La0.005B6, the ferromagnetism has a

maximum moment of 0.07µB/La with a Curie temperature of 600 K. This is of course entirely

unexpected and falls outside the usual models for magnetic ordering in metals. Since these

materials couple low carrier densities with ferromagnetism, they may be excellent candidates for

field gating experiments, where the ferromagnetism is controlled directly through modulation of

carrier density [Figure 6.2]. We have performed a series of transport, magneto-transport, and

65

Hall measurements to characterize this novel material. The results of these measurements are

described below.

(a) Transport Results

The transport characteristics of these materials are extremely sensitive to the degree of

band overlap (if any) and the position of the Fermi level. Our measurements indicate that the

undoped material is essentially a small band gap semiconductor with a temperature dependent

resistivity shown in Figure 6.3c. In the La-doped material, the addition of electron carriers raises

the Fermi level to the upper band, leading to metallic behavior with a weak temperature

dependence, as shown in Figure 6.3a. Similarly, hole-doping this material through removal of

Ca decreases the Fermi level, placing it in the lower band, again leading to a weak metallic

behavior as shown in Figure 6.3b.

For the metallic compounds, the rather weak temperature dependence implies that

electron-phonon scattering may be quite constrained. This is not surprising in view of the

peculiar band structure of this material, where the metallicity arises from quite narrow bands.

Any scattering mechanism would have a constrained phase space since the final momentum

values of the scattered electrons have to lie within this narrow pocket for the process to be

allowed.

(b) Magneto-transport Results

Magneto-transport studies are a useful tool for elucidating the mechanisms behind

ferromagnetism in many materials. For example, the magnetoresistance of an itinerant

ferromagnet, the category into which we would suspect doped CaB6 to fall, is negative and linear

with magnetic field. As illustrated in Figures 6.4 - 6.6, data taken on these compounds show a

negative magnetoresistance that exhibits curvature at low fields but becomes linear at higher

66

fields. These findings imply additional interactions that distinguish these materials from

ordinary itinerant systems.

(c) Hall Measurements

The Hall effect is useful in the determination of the carrier sign and concentration. In a

Hall measurement, current is injected longitudinally while a magnetic field is applied normal to

the plane of the sample. Charge builds up on one side of the sample as the Lorentz force acts on

the charge carriers. As a result, a potential can be measured in-plane and perpendicularly to the

current path, as illustrated in Figure 6.7.

The sign of this Hall voltage indicates the sign of the carriers and can be used to calculate

the carrier concentration as follows:

neRwhere

diHRV H

HH

1, ==

A compensated band structure complicates the expression for the Hall coefficient, RH.

When both electrons and holes contribute to the electrical transport of a material, the Hall

coefficient must include the electron and hole densities, as well as the corresponding mobilities:

( )2

22

np

npH npe

npR

µµµµ

+

−=

Hall effect measurements are useful to the determination of the band structure of CaB6.

As shown in Figure 6.8, data already taken indicate a compensated band structure in which both

holes and electrons play a role in transport. The magnitude of the carrier density is clearly

determined by the position of the Fermi level, where a reduction in the carrier density occurs as

the Fermi level is shifted downwards in the band in this material.

67

The anomalous Hall effect is commonly seen in many ferromagnetic systems. In addition

to the conventional Hall resistivity, a term proportional to the magnetization appears:

MRBRR SoH π4+=

This term arises due to the internal field present within the ferromagnet. In ferromagnets that

exhibit hysteresis in magnetization with field, manifestation of this term in Hall data is marked

by a break in the slope accompanied by a non-zero intercept.

Since the magnetization data taken on doped CaB6 features hysteresis with field, a

hysteretic anomalous Hall voltage is expected. In our data, hysteresis is seen in hole-doped

CaB6, as shown in Figure 6.9. This effect is only observed when the sample is zero-field cooled

and only during the first sweep in field. The saturation field is in good agreement with that

indicated in the magnetization curve shown in Figure 6.10. To observe the anomalous signal in

such a weakly ferromagnetic system requires extremely careful measurement techniques over

small magnetic field ranges.

(d) Summary

Our measurements indicate an unusual mechanism for magnetism exists in these

materials. Due to the low carrier densities, the hexaborides present a clear opportunity for a new

class of materials where the ferromagnetism can be directly controlled through external field

gating. As such, these materials can be classified as novel magnetic semiconductors, allowing

for their use in the newly emerging field of spin-based electronics.

68

Figure 6.1 Figure 6.2

Figure 6.3 (a-c)

69

Figure 6.4. Resistance versus magnetic field for stoichiometric CaB6.

Figure 6.5. Resistance versus magnetic field for electron-doped CaB6.

70

Figure 6.6. Magnetoresistance versus magnetic field for hole-doped, or Ca-deficient, CaB6.

Figure 6.7. Conventional Hall geometry for a rectangular sample.

71

Figure 6.8. Effective carrier concentration versus temperature for electron-doped, hole-doped, and undoped CaB6. Note that the carrier concentration shown here is defined as RH/e and may be complicated by compensation, i.e. may not represent only electron-like transport.

Figure 6.9. Hysteresis in Hall voltage versus magnetic field for Ca1-B6, which shows the

saturation field to be roughly 2500 Oe.

72

Figure 6.10. Magnetization versus magnetic field for Ca1-B6, which shows the saturation field

to be roughly 2000 Oe, in good agreement with Figure 6.9.

73

7. Fabrication and Characterization of Magnetic Nano-Particles

(S. von Molnár)

During the past year, our major focus has been on the development of more sensitive Hall

devices to study magnetization reversal mechanisms in greater detail and in ever smaller

numbers of particles. The goal is to characterize single magnetic nanoparticles in arbitrarily

large fields and over temperature ranges spanning liquid Helium to room temperature.

Theoretical considerations (see Figure 7.1) show that the sensitivity of Hall devices is greatest

when the array or magnetic object to be measured covers maximal areas of the Hall cross itself.

A demonstration device is shown in Figure 7.2.

The measured Hall voltages due to this array are shown in Figure 7.3 as a function of the

applied magnetic field in a direction almost perpendicular to the long axis of the particles (almost

parallel to the 2DEG film). These data indicate, as reported in last year's MURI summary, that

magnetization reversal occurs in different modes depending on the direction of the applied field.

Because the Hall voltages measured exceed earlier measurements by more than an order of

magnitude with noise levels as low as 0.04 - 0.07µV/ Hz , it has been possible to analyze these

data with confidence as arising from two contributions, an irreversible part which depends on

reversal mode, and reversible part from which one can extract the anisotropy distribution due to

the slight variation in particle shape (see solid line, Figure 7.3). Furthermore, it has been

possible to measure the magnetic viscosity, Figure 7.4, and to derive from this the activation

volume for irreversible processes.

Finally, from the measured voltage response and noise level, it is possible to predict that

a single 10 nm particle grown onto a 400 nm2 Hall cross may be characterized in detail. As a

first step in this direction, we have produced a working Hall cross of areal dimension 1•m2 (see

74

Figure 7.5) onto which a portion of a 4x4 particle array has been grown. This device is perfectly

capable of measuring the ~7 particles which contribute to the signal and detailed measurements

are underway. We have also, over the past year, installed and made operational a new electron

beam lithographic tool with line resolution of order 20 - 30 nm. Attempts to fabricate Hall

devices of these dimensions are underway in GaAs/GaAlAs heterostructures and will be

extended to InAs quantum well structures operable at room temperature.

75

Figure 7. 1. Comparison of Hall voltages. Calculations were performed by assuming Hall

crosses of different sizes (circles, lower axis). Array and Hall cross were either assumed to be aligned with their centers (closed circles) or with one of their corners (open circle). All triangles represent experimental results (see text). In addition, the calculated Hall voltage for different separations between particles and 2DES is shown (crosses, upper axis).

76

Figure 7.2. SEM picture of an array of 420 particles grown onto a Hall cross. The etched Hall cross of about 3.2 x 2.8 µm2 is clearly visible. The image shows an area of 4.5 x 4.5 µm2.

77

Figure 7.3. Hysteresis curve (x) of particle array (in Figure 7.2) for field applied 86 deg. with respect to the particles' easy magnetization direction. Fits account for reversible (first and third quadrant, •) and reversible & irreversible(second and fourth quadrant, •) magnetization processes. From the "difference, the irreversible contributions, i.e. the anisotropy distribution, can be evaluated (line, right axis).

78

Figure 7. 4. Measured time dependence of the total polarization under different conditions.

The line presents the result of a fit to a logarithmic law.

79

Figure 7.5. SEM images of a small array grown onto a Hall cross of 1 x 1 •m2 (images size

1.1 x 1.5 •m2, one current leg can be recognized on the top.

80

8. Interface Phenomena and Screening Length Determination in Tunnel Junction Capacitors

(A. F. Hebard, K. T. McCarthy, D. Temple)

As electronic devices shrink to nanoscale size and their components become increasingly

close spaced, it becomes imperative to understand the new physics that becomes manifest at

reduced length scales. For example, when two parallel metal plates are placed in close proximity

(see inset of Figure 8.1) with a dielectric of thickness d separating them, the capacitance deviates

sharply from the value that would be expected from purely geometrical considerations. This

deviation occurs because, in addition to the voltage drop across the insulator separating the

plates, there are also voltage drops across the two metal-insulator interfaces. The extra

capacitance associated with these voltage drops is called the interface capacitance, Ci, and it

manifests itself as an additional capacitance in series with the geometrical capacitance, Cg.

The interface capacitance can be experimentally determined by plotting the reciprocal of

the measured capacitance, Cm, versus the dielectric spacing, d, and then extrapolating the linear

dependence to d=0 to find the intercept, Ci-1. Since capacitors adding in series obey the relation,

C C Cm i g− − −= +1 1 1 , the linear dependence arises because Cg is inversely proportional to d and Ci

depends only on interface properties. An example of this linear dependence is shown in the

Figure 8.2 plot of the thickness dependence of the reciprocal capacitance of seven different

Si/SiO2/Ni trilayer sandwiches with thermally grown (950°C) oxide thicknesses ranging from 35

to 400Å. The substrate is a (100) oriented n-type device grade silicon wafer with resistivity on

the order of 0.1Ωcm.

An extrapolation of the linear fit in Figure 8.2 to the negative intercept on the abscissa

gives a direct measure of the “thickness” d0 of the interface capacitance. This thickness includes

contributions from both interfaces. Accordingly, if the actual dielectric thickness d is less than

81

d0, then the interface capacitance dominates over the geometrical capacitance and properties such

as the ac loss (dispersion) are dominated by interface processes rather than bulk processes

associated with the dielectric. Such a situation is shown in Fig. 1, which reveals considerable

dispersion in the capacitance of an Al-AlOx-Al tunnel junction capacitor having a thermally

oxidized barrier with thickness on the order of 20 Å. This is considerably thinner than the d0=50

Å inferred from a set of measurements on similar structures with different dielectric spacings.

We note that the dispersion is pronounced down to frequencies as low as 0.001 Hz and that there

is no evidence of a loss peak at these low frequencies. We surmise that the filling and emptying

of interface trap states is responsible for this dispersion. This conclusion is supported by our

observation that for thicker dielectrics the dispersion is considerably less, as would be expected

since the capacitance in this regime is dominated by the frequency-independent properties of

bulk aluminum oxide.

Frequency dependence at such low frequencies can have considerable implications for

the temporal stability of devices that incorporate tunnel junctions and related structures.

Accordingly, it is important to understand in detail the interface phenomena that contribute to the

magnitude of d0 and the accompanying losses associated with time-dependent excitations. The

Thomas-Fermi screening length, which depends on the density of states in the metal electrodes,

is an important starting point. Film roughness, impurities, and the number and type of interface

trapping sites are also important and are presently being investigated.

If magnetic electrodes are involved, then there is an additional contribution to the

interface capacitance that is predicted to arise from the exchange interactions between the spin-

polarized carriers of the magnetized electrodes. Our preliminary attempt to confirm this

prediction is shown in Figure 8.3, which is a plot of the d0’s obtained from a series of Si/SiOx/M

82

(M=Fe, Ni, Co) trilayer structures, all made on the same n-type silicon wafer. Each point on this

plot represents a series of measurements similar to that shown in Fig. 2 on seven different

samples. The values for d0 are obtained by extrapolation in the same way as discussed above.

The thickness of the depletion layer in the doped silicon depends on voltage bias. Accordingly,

d0 is minimum for each series with different counter electrode M at positive voltage where the

depletion capacitance is maximum (strong accumulation). Assuming that the differences in d0

arise only from the differences in the counter electrodes (assuming everything else remains

equal), we tentatively conclude that the screening lengths of Fe, Ni and Co are significantly

different. Further work must be done to show that these differences arise from the predicted spin-

dependent surface screening in the ferromagnetic electrodes. We anticipate that these studies

may have important implications for tunneling magnetoresistive (TMR) devices and related

structures with magnetic electrodes. The effects will be enhanced with low-carrier density

magnetic materials having relatively large screening lengths.

83

10-2 100 102 104 1061.00

1.50

2.00

2.50

3.00AL-AL2O3-AL

C1 (µ

F/cm

2 )

Frequency (Hz)

V /σ

0 d

d + d o

Figure 8.1. Frequency dependent capacitance of an Al-Al2O3-Al tunnel-junction trilayer structure. The inset shows schematically the voltage drops across the two interfaces and the dielectric (shaded).

84

Figure 8.2. Inverse areal capacitance as a function of dielectric thickness, d, for a series of seven Si/SiO2/Ni samples with d ranging from 35 to 400Å

0 50 100 150 200 250 300 350 4000

2

4

6

8

10

12

d0=12.8 ± 4.4 ÅΚ=4.11 ± 0.07CI=2.84 ± 0.98 µF/cm2

A/C

1 (cm

2 /µF)

d (Å)

85

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6

0

10

20

30

40

50

Fe Ni Co

f = 100 HzVac = 50 mV

d 0 (Å)

Vbias (V)

Figure 8.3. Dependence of interface thickness d0 on dc bias voltage. The differences between the curves reflect differences in the screening of the magnetic electrodes identified in the legend

86

9. Hysteresis and Relaxation Behavior in GMR Multilayers

(A. F. Hebard, N. Theodoropoulou, A. K. Majumdar, D. Temple)

The work reported in this section is a continuation of collaborative work between UF and

MCNC on the characterization of GMR multilayer stacks prepared by ion beam sputter

deposition. The effort at MCNC has focused on deposition and structural characterization and the

effort at UF has been on providing rapid feedback to MCNC on the electrical

(magnetoresistance) and magnetic (magnetization) properties of the fabricated multilayer

structures. This collaboration has produced GMR structures with properties comparable to the

best structures fabricated using sputter deposition techniques. The magnetoresistance of a typical

Fe/Cr multilayer [Fe(20Å)/Cr(10 Å)]×30 made by ion beam sputter deposition (Xenon, 900V,

20mA) is shown in Figure 9.1. The chromium spacer thickness is chosen to insure that the

ferromagnetic domains in adjacent iron films are antiferromagnetically coupled. Thus in low

field, when the spins of adjacent layers are antiparallel, spin-flip scattering gives rise to a higher

resistance than occurs at high fields when the spins are parallel.

In our magnetization studies of these structures, we have observed a heretofore-

unreported hysteretic behavior in the low field (H<1000 Oe) region. The evidence for hysteresis

is shown in the temperature-dependent magnetization data of Figure 9.2. We note from these

data that the magnetization at a given temperature and field is history dependent, i.e., the value

depends on whether it has been zero-field-cooled (ZFC) or field-cooled (FC). For a given field

there is an irreversibility temperature Tc(H), below which there is hysteresis (irreversible

behavior) and a difference between the FC and ZFC curves, and above which, there is no

hysteresis, and the FC and ZFC curves overlap. The irreversibility temperature is field

87

dependent and thus defines a critical field, which separates irreversible from reversible behavior

and which has the temperature dependence shown in Figure 9.3.

Relaxation dynamics in the hysteretic region are extremely slow. This is shown in Figure

9.4 where the sample has been field cooled to 55K in a field of 200 Oe. After turning off the

field, the relaxation towards the ZFC state is observed to have logarithmic time dependence as

shown in the figure. Assuming the persistence of this logarithmic rate, the time taken to

complete the transition is calculated to be greater than the age of the universe.

The temperature dependence of the logarithmic rates at selected fields is shown in Fig. 8.

As temperature increases the logarithmic creep rate decreases and approaches zero near Tc(H).

This “critical slowing down” with increasing temperature is common to spin-glass and related

systems where disorder and frustration are prominent. This behavior is distinctly different from

the “magnetic viscosity” effects observed when a field is applied to a bulk material such as

Alnico and the logarithmic creep rate is observed to increase with temperature. With logarithmic

rates it is straightforward to show that the time, tF, the system takes to get from the ZFC to the

FC state (or vice verse) is exponentially sensitive to the difference (FC-ZFC) between the two

states divided by the logarithmic creep rate. Since with increasing temperature the separation

between states decreases faster than the logarithmic rate, the time tF, which at low temperatures

can be longer than the age of the universe, decreases toward zero at the irreversibility

temperature. In the language of Zeno, the tortoise always wins the race with the hare.

The observed behavior is different from conventional spin glasses where the

irreversibility temperature usually delineates the boundary between a low-temperature frustrated

system of exchange-coupled spins and a high-temperature paramagnetic system of independent

spins. In the GMR samples the boundary is between antiferromagnetically coupled

88

ferromagnetic domains that are predominantly antiparallel at low fields (T<Tc) and parallel, but

still antiferromagnetically coupled, at higher fields (T>Tc). Alternatively, these dynamical effects

are related to thermally blocked relaxation of superparamagnetic grains in our multilayers.

Understanding these dynamics is essential for the achievement of long time stability in devices

based on GMR effects.

-20 -10 0 10 20

-0.20

-0.15

-0.10

-0.05

0.00

10 K

300 K

Mag

neto

resi

stan

ce (%

)

H (kOe)

Figure 9.1. Percent change in resistance of an iron-chromium 30-layer GMR stack [Fe(20Å)/Cr(10Å)] at the indicated temperatures as a function of magnetic field applied in a direction parallel to the plane of the film.

89

0 50 100 150 200 250 300 3500.0

2.0

4.0

6.0e-4 ZFC (solid symbols)FC (open symbols)

H(Oe)

600

100200400

Mag

netiz

atio

n(em

u)

T(K)

Figure 9.2. Temperature dependent magnetization for field cooled (FC) and zero-field-cooled (ZFC) curves at the indicated fields.

90

0 50 100 150 200 250 3000

100

200

300

400

500

Mag

netic

Fie

ld (O

e)

Temperature (K)

Figure 9.3. Temperature dependence of the irreversibility field.

91

10 100 1000 100004.48

4.50

4.52

4.54

4.56

4.58e-4FC in 100 Oe to 55KTurn off field at t=0

M (e

mu)

Time (sec)

Figure 9.4. Time dependence of the magnetization after cooling the sample to 55K in a field of 100Oe and then setting the field to zero.

92

0 50 100 150 2000.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0e-2

Loga

rithm

ic R

ate

Temperature(K)

B50 B100 B150 B200 B300

Figure 9.5. Logarithmic rate (magnetic viscosity) as a function of temperature at the indicated magnetic fields.

93

10. Superparamagnetic-ferromagnetic Transition

(R.C. Dynes)

We have studied the superparamagnetic-ferromagnetic transition of small Ni particles as

a function of T and particle size. The transition has been detected in magnetoresistance and

magnetization of thin discontinuous films of Ni quench-condensed on an isolating substrate.

With increasing coupling between grains (

≈100 Ao

in size) we observe a crossover from non

hysteretic to hysteretic behavior in magnetoresistance and magnetization. This crossover

correlates with a transition from isolated grains to larger clusters. With increasing cluster size,

the blocking temperature increases and the material becomes a ferromagnetic. This 2

dimensional crossover allows a study of the interactions between magnetic-nanoparticles.

(a) Micro SQUIDS

We have developed the technology to manufacture micro SQUIDS to be used to study

magnetic nanoparticles. We have fabricated SQUID loops approximately equal to 2 microns on

a side using e-beam lithography techniques. The sensitivity of these SQUIDS is such that they

are capable of detailing a single ferromagnetic nanoparticle (a few hundred angstroms in

dimension) and will be used to study the switching characteristics of ferromagnetic particles.

Attached is a figure describing the geometry of these SQUIDs and a typical Ic(B) pattern at 1.5K

for an Al SQUID.

(b) Superconducting and Spin-dependent Tunneling Using native Oxide Barriers

on CoFe Thin Films

Spin dependent tunneling has been demonstrated in native oxide barriers of CoFe. Using

a superconductor as a counter electrode, it has been clearly demonstrated that the dominant

conduction mechanism for CoFe to the superconductor is electron tunneling through the native

94

oxide. Replacing the superconductor with the ferromagnetic Co results in a magnetoresistance of

≈ 4%. This is the largest value to date in a spin-dependent tunnel junction using a native oxide

barrier on a ferromagnet. Junction quality depends strongly on the deposition condition for the

top electrode.

11. Length Scale of Magnetism

(F. Hellman)

The goal of this work is to understand the length scale and magnitude of the magnetism

induced in a non-magnetic or nearly magnetic material by proximity to a magnetic material,

particularly in metallic materials. This is an effect that is well understood in superconducting

materials and has led there to practical devices. In magnetism, the effect is less well understood

and is generally believed to be extremely local. However, it is clear that in materials, which are

nearly magnetic, such as Pd or Pt, this length scale will be much longer. As little as 1% Co in Pt

leads to a ferromagnetic Curie temperature of over 5K, indicative of the ability of Co to induce

magnetism in Pt on a relatively long length. We have used UHV techniques to prepare vapor-

deposited Co-Pt and related Ni-Pt and Co-Pd alloys of various compositions and Curie

temperatures. These alloys spontaneously form a nanostructured incoherent multilayer, causing

perpendicular anisotropy to occur in a nominally fcc material. This compositional

nanostructuring has to date not been directly measured, although asymmetric short-range

chemical order has been seen in EXAFS. By controlling the growth parameters, we can control

the development of this nanostructure and measure its effect on the magnetic properties. In the

most extreme case, the magnetization versus temperature exhibits a broad (several hundred

degrees in width) and nearly linear temperature dependence. We are working at present on

preparing multilayers in order to determine the nature of the nanostructure which could give rise

95

to the observed broad M(T) in the spontaneously nanostructured material. We are also at present

completing work on thermodynamic characterization of several nanostructured magnetic

materials.

12. Fabrication and Characterization of Magnetotransport in Colossal Magnetoresistive Oxide Films and Hybrid Structures

(Srinivas V. Pietambaram, Jaeyoung Choi, D. Kumar, Hyoung-June Kim, Rajiv K. Singh and S. J. Pearton)

The continually increasing demand for magnetic information storage and retrieval has

driven a significant worldwide effort to improve the performance of relevant hardware

components. As the areal density continues to increase, more sensitive materials and innovative

structures will be required to detect the decreasing magnetic fringe-fields emanating from the

media. Doped LaMnO3 is being examined as a possible next generation magnetoresistance

sensor material. Over an appreciable range of doping, these materials exhibit a very large

magnetoresistance effect at temperatures close to where they undergo a ferromagnetic-

paramagnetic transition. The basic manganite, LaMnO3, is antiferromagnetic and insulating.

Partial doping of La sites with alkaline earth cations in LaMnO3 results in a Mn3+/Mn4+ mixed

valence state which is responsible for both metallic conductivity and ferromagnetism in

accordance with double exchange interaction. Although, structural and magnetotransport

properties of manganite thin films have been found to vary significantly with change in process

parameters, detailed and systematic study is missing in the literature. Besides, the effects of

crystallinity and oxygen content of the films on the properties is not clearly resolved. The highly

spin polarized nature of these materials can be used for developing new structures having a high

MR ratios.

96

(a) Oxygen Content and Crystallinity Effects in Pulsed Laser Deposited Lanthanum Manganite Thin Films

The effects of oxygen in manganite film have been a focus since the discovery of

extraordinary magnetotransport in this kind of materials. There is no quantitative relation

between the oxygen content and the magnetic and transport properties in thin films till date due

to the difficulty in the determination and control of oxygen content. Substrate temperature,

oxygen partial pressure, and deposition rate could have an effect on the oxygen content of the

film severely. Vacuum annealing can produce a deficiency in oxygen while excess oxygen can

be produced by annealing in oxygen. All these indicate that oxygen content can be flexible in

manganese oxides. In the early studies of CMR materials, a post deposition anneal in oxygen at

high temperatures was critical for achieving large magnetoresistance. However, it was not known

whether the improvement was due to grain refinement through grain growth and enhancement in

the crystallinity of the films or due to oxygen incorporation. Improvement in the crystallinity and

grain growth enhances the properties of the oxide thin films. Systematic post deposition heat

treatments were performed to deconvolute the effects of oxygen content and grain growth in

Lanthanum Manganite thin films.

Bulk La0.7Ca0.3MnO3 (LCMO) was prepared by a ceramic method. The required

quantities of respective oxide or carbonate powders were mixed and sintered at 1400°C for 24

hours. Six LCMO films with a thickness of 1500 Å were grown in situ on (100) LaAlO3

substrates using a pulsed laser ablation system. To eliminate other external effects, all the

substrates were placed side by side on sample holder. All the films were characterized to see if

the properties were identical before performing any post deposition heat treatments. A 248 nm

KrF pulsed laser with 5 Hz repetition rate and 1.6 J/cm2 energy density was used. A substrate

temperature of 700°C and oxygen pressure of 250 mTorr were used during the deposition of the

97

films. Following the deposition, the films were cooled down to room temperature at a rate of

10°C/min in 400 Torr of oxygen. After the initial characterization to check the identical

properties, the films were subjected to the following post deposition anneals – (i) annealing in

oxygen at 900°C for 4 hrs, (ii) annealing in argon at 900°C for 4 hrs, (iii) annealing in oxygen at

500°C for 12 hrs, (iv) annealing in argon for 12 hrs and (v) annealing in vacuum at 850°C for

half-an-hour.

XRD patterns of the films subjected to various anneals are shown in Figure 12.1. From

the Figure it is clear that all the films have single phase with (00l) peaks with l = 1 and 2. The

presence of only sharp (00l) peaks indicates the highly textured growth of all the films on (100)

LaAlO3 substrate. The lattice parameters for the films with no anneal, 900°C oxygen anneal and

900°C argon anneal were found to be 3.8632, 3.8337, and 3.8444 Å respectively. XRD patterns

of 500°C oxygen and argon annealed films were similar to that of as-deposited films. The films

annealed in oxygen and argon show a decrease in the full width at half maximum (0.1°)

compared to the as-deposited film (0.25°) indicating these films are more crystalline.

The variation of electrical resistance in zero and applied field (5T) as a function of

temperature for the films subjected to various anneals are shown in Figure 12.2. All the films

were grown under identical conditions so that film thickness, oxygen contents of all the films and

other external effects (before anneals) could be kept identical. It is important to keep these

parameters identical in order to reveal the effects of the magnetotransport properties in these

films. According to the variation of resistance shown in Figure 12.2., all the films (except

vacuum annealed one) have similar qualitative magnetotransport behavior. That is, all the films

undergo an insulator-to-metal (I-M) transition as the temperature is lowered down and the

resistance of all the films is suppressed significantly with the application of magnetic field. The

98

suppression in film resistance in each case is maximum near the resistivity peak in zero field as

observed frequently by others in several manganite systems. The MR ratios of the films were

calculated using the data in Figure 12.2. The MR ratios obtained are plotted in Figure 12.3 as a

function of temperature at 5T.

The as-deposited films show an I-M transition at 260 K and a MR ratio of 190% in 5T.

The films subjected to 500°C oxygen anneal and 500°C argon anneal show little change in I-M

transition and MR ratio. This shows that 500°C is not a sufficiently high enough temperature to

affect the oxygen content, grain growth and crystallinity. The films subjected to 900°C oxygen

anneal show an improvement in transition temperature (290 K) but a marginal enhancement in

the MR ratio (225%). The films subjected to 900°C argon anneal show slight increase in the

transition temperature (270 K) and significant enhancement in the MR ratio (525%). The films

subjected to vacuum anneal have shown deteriorated properties (insulating down to 10 K).

Annealing of the films at high temperature lead to two simultaneous effects – removal of

oxygen from the film and grain growth and improvement in crystallinity. Films subjected to

500°C oxygen and argon anneals show little effect on the transition temperature or MR ratio.

This temperature is not high enough for either oxygen incorporation or to cause grain growth.

Further, no observable differences were found from the XRD patterns of these films as compared

to the as-deposited films. Hence these anneals have no effect on the properties of the film.

Films annealed in oxygen at 900°C show an increase in the transition temperature from

260K to 290K. The resistance of the films also decreased compared to as deposited films. These

effects can be understood from the increase in the oxygen content of the films. The

ferromagnetic transition is strongly determined by the number of Mn4+ ions. The mixed

Mn3+/Mn4+ valence is believed to give rise to both ferromagnetism and metallic behavior in

99

LCMO films, and to be responsible for the occurrence of colossal magnetoresistance. As oxygen

is incorporated in the film two distinct reactions occur: contraction of the lattice, as evidenced by

x-ray diffraction; and gain in the O2- ions. Gain of oxygen ions should lead to changes in

magnetotransport similar to those resulting from the application of external pressure. Under

applied pressure the lattice contracts while transition temperature (Tc) increases and resistance

decreases. These results can be explained by an enhancement in the Mn-Mn electron transfer

probability as the average lattice spacing is decreased. An increase in the transfer probability

should lead to an enhanced ferromagnetic correlations and increased carrier mobility between

adjacent Mn ions. This leads to a higher Tc and lower resistance. The second effect relating to

oxygen arises from the requirement of charge neutrality within each unit cell. The chemical

formula for LCMO can be written as La1-x3+Cax

2+Mn1-x+2δ3+ Mnx-2δ

4+O3-δ2-. Therefore, each

oxygen incorporated into LCMO should lead to a conversion of Mn3+ ions to two Mn4+ ions. The

carriers in LCMO are holes whose concentration is proportional to the Mn4+ concentration.

Therefore, the incorporation of oxygen should increase Mn4+ concentration, which leads to an

increase in the carrier concentration, and hence decrease in the resistance.

However, annealing at high temperatures also causes grain growth which affects the

domain size. These effects can be isolated by annealing the films at elevated temperature in an

ambient other than oxygen such as argon. Films subjected to argon annealing at 900°C show

marginal increase in the transition temperature. The resistance of the films is higher than that of

the oxygen annealed films. The increase in the resistance of the films can be attributed to the loss

of the oxygen from these films. The interesting feature of high temperature argon annealed films

is significantly high MR ratio observed in these films. This may be attributed to the increase in

the domain size. Near the domain-wall boundaries, the pairs of spins of Mn3+ and Mn4+ may not

100

be parallel. As a result the electron transfer between pairs of Mn3+ and Mn4+ ions across the

domain wall is difficult and the resistance is high. Increase in the domain size reduces the

amount of domain-wall boundaries in a specified area. As a result, smaller fields are necessary to

align these domains; hence the suppression of resistance is much higher for a given magnetic

field and as such higher MR ratios. The marginal improvement in the transition temperature may

also be related to an increased domain size due to the smaller amount of domain-wall boundaries

and reduced resistance for the ferromagnetic alignment. The lattice parameter of argon annealed

films is higher than that of oxygen annealed films which indicates that the films have less

oxygen. However, the decrease in the lattice parameter in argon annealed films compared to as-

deposited films is not clearly understood. It seems that an increase in the domain size causes a

decrease in the lattice parameter.

Vacuum annealed samples undergo the most severe heat treatment. The loss of oxygen

from the film leads to the conversion of two Mn4+ ions to Mn3+ ions. The presence of critical

oxygen content is essential for the occurrence of metal to insulator transition (MI) in the LCMO

films. Observations made by others indicate that MI transition will disappear in the compound

La1-xCaxMnO3 with a Mn4+/Mn3+ ratio less than 0.17. In the vacuum annealed samples, there

may be a depletion of oxygen to the extent in which Mn4+/Mn3+ ratio is less than 0.17. Hence

these films show insulating behavior down to 10 K.

Low temperature (500°C) oxygen and argon anneals show little effect on the transition

temperature or MR ratio. High temperature (900°C) oxygen anneal shows a significant

improvement in the transition temperature while high temperature (900°C) argon anneal shows a

substantial increase in the MR ratio. The transition temperature is related to the oxygen content

while the MR ratio is affected by the domain size.

101

(b) Fabrication of Multi-Layered Structures

One of the distinguished features of La0.7Ca0.3MnO3 (LCMO) is a ferromagnetic half-

metallic state, where only the single spin band crosses the Fermi level. Such a half metallic state

can lead to 100% spin polarization of conduction carriers, which is much higher than that of

typical ferromagnetic metals such as Co, Ni, or Fe and their alloys. Nearly 100% of spin

polarization of various manganites such as LCMO and La0.7Sr0.3MnO3 (LSMO) have been

reported from spin polarized photoemission studies. One way of taking advantage of this highly

spin polarized nature of these manganites is a spin-dependent tunneling magnetoresistance

(TMR) of a magnetic tunnel junction (MTJ), since the TMR ratio is defined and given by

TMR (%) = (RAP – RP)/ RAP X 100 = 200 X P1P2/(1+ P1P2)

Where RAP and RP are the resistances in the antiparallel and parallel magnetic

configuration respectively and P1and P2 are the spin polarizations of the two electrodes. Spin

polarization P is the difference between the density of states (DOS) of spin-up (parallel to the

magnetization) and spin-down (antiparallel to the magnetization) electrons at the Fermi level.

Recently, TMR ratios up to 400% have been reported in LSMO/STO/LSMO trilayer

structure at low temperature (10 K). Despite of these large TMR values, the obstacle in these

systems is that large TMR values are observed at temperatures below 100 K, rapidly vanishing at

higher temperatures. Since this temperature is far below the Curie temperature of these

manganites (Tc of LCMO and LSMO are approximately 270 K and 370 K, respectively), this

premature decrease of TMR brings an important issue for understanding as well as practical

applications of these systems. Even though the cause of the rapid TMR loss with temperature is

not understood at this moment, the existence of surface dead layer with depressed magnetic order

at elevated temperatures is suspected. Since the magnetic properties of these materials are highly

102

sensitive to local crystal properties and the extrinsic strain field induced by lattice mismatch

among the grown layers, combination of selected material species, their crystal integrity, and

thickness of insulators will be critical for improving the TMR values as well as their temperature

dependency.

With this background, we have started investigating the characteristic TMR effects in

various LCMO/LAO/FM (FM = NiFe, etc) systems mainly as a function of deposition conditions

and layer thickness. In the first study, we have chosen La0.7Ca0.3MnO3 as a primary

ferromagnetic (FM) material, since it shows good ferromagnetic properties and is well optimized

in our lab. We have decided to use LaAlO3 as the barrier material because it gives good lattice

match with the LCMO base. NiFe was used as the top FM electrode. The choice of NiFe as a top

electrode could claim the following advantages. NiFe is a typical FM metal, which shows

superior soft magnetic properties. Low Coercivity value of NiFe, usually less than 10 Oe, as well

as typically observed squareness of M-H hysteresis give the wide range of antiparallel alignment

with the other electrode LCMO, which shows a coercivity values of 200 - 400 Oe. Also NiFe

grows pretty well on any substrate.

Thin films of Lanthanum calcium manganite (La0.7Ca0.3MnO3) were deposited on LaAlO3

substrates in a pulsed laser ablation system at a substrate temperature of 850°C and oxygen

pressure of 250 mTorr. Following the deposition, thin films of LaAlO3 were grown on

La0.7Ca0.3MnO3 in the same system. After the deposition, the films were cooled down to room

temperature at a rate of 10°C/min in 400 Torr of oxygen. Then NiFe films were sputter deposited

using UHV sputtering chamber.

The first requirement to observe TMR in a junction is the magnetic decoupling of the

electrodes, which allows the existence of an antiparallel configuration in a certain magnetic field

103

range as well as the parallel configuration at high field. We have measured the field dependence

of magnetization at 10 K for the structures with varying thickness of the barrier layer (shown in

Figure 12.4 (a), (b), (c) and (d)). We found that at least 10 A of barrier layer is necessary for

decoupling the two electrodes

We have then fabricated some shadow mask magnetic tunnel junctions to study the TMR

behavior. The shadow mask junctions were as shown in Figure 12.5. TMR values observed with

a shadow mask as described above were poor. The actual reason for this is not clearly

understood. One of the probable reasons may be that the interface resistance or the junction

resistance is lower compared to that of the sheet resistance of the electrode. In such a case CPP

resistance is no longer representative of the interface and would contain complex, current-

distribution limited resistance that is associated with the details of junction geometry. This is

especially true in case of magnetic oxide materials such as manganites whose resistivity can vary

two orders of magnitude between 4.2 K and ambient temperature. Another reason may be due to

the local pinhole shorts, which occur, when the thickness of the barrier layer is small. We need to

further investigate to understand these effects clearly.

104

Further studies are underway in which we are planning to fabricate smaller junctions

(thus having higher junction resistances) and study the TMR characteristics.

Inte

nsity

(arb

itrar

y un

its)

2-θθθθ

(degrees)30 40 50

(a)

(b)

(c)

(100

) film

(100

) sub

stra

te

(200

) film

(200

) sub

stra

te

(100

) film

(100

) film

(100

) sub

stra

te(1

00) s

ubst

rate

(200

) film

(200

) film

(200

) sub

stra

te(2

00) s

ubst

rate

Figure 12.1. XRD patterns of LCMO films, (a) no anneal, (b) 900 C oxygen anneal,and (c) 900 C argon anneal

105

Res

ista

nce

(Ohm

)

0

250

500

0100200300400

0300600900

1200

0

300

600

900

Temperature (K)0 100 200 300

0300600900

1200

(a)

(b)

(c)

(d)

(e)

Figure 12.2. Variation of electrical resistance with temperature for (a) noanneal, (b) 900 C oxygen anneal, (c) 900 C argon anneal, (d) 500 C oxygen anneal and (e) 500 C argon anneal (filled circle(0 T), filled diamond (2 T) and filled triangle (5 T))

106

Temperature (K)0 50 100 150 200 250 300 350

MR

ratio

(%)

0

100

200

300

400

500

600no anneal9000C O2 anneal9000C Ar anneal 5000C O2 anneal 5000C Ar anneal

Figure 12.3. Variation of MR ratio with temperature for the films

107

Magnetic Field (Oe)-1000 -500 0 500 1000

Mag

netiz

atio

n (e

mu)

-0.0015

-0.0010

-0.0005

0.0000

0.0005

0.0010

0.0015

no spacer

Magnetic Field (Oe)-1000 -500 0 500 1000

Mag

netiz

atio

n (e

mu)

-0.0015

-0.0010

-0.0005

0.0000

0.0005

0.0010

0.0015

5 A spacer

Magnetic Field (Oe)-1500 -1000 -500 0 500 1000 1500

Mag

netiz

atio

n (e

mu)

-0.0015

-0.0010

-0.0005

0.0000

0.0005

0.0010

0.001510A LAO

Magnetic Field (Oe)-1500 -1000 -500 0 500 1000 1500

Mag

netiz

atio

n (e

mu)

-0.0015

-0.0010

-0.0005

0.0000

0.0005

0.0010

0.0015

20 A LAO

Figure 12.4 Field dependence of magnetization of 10 K for LCMO/LAO/NiFe heterostructure with (a) no barrier, (b) 5Å barrier, (c) 10Å barrier and (d)20Å barrier.

108

LAO LAO

LCMLA

NiF

LCMLA

NiF

LCMO/L

Figure 12.5 Shadow mask junctions used for making TMR measurements.

109

13. Chemically Selective Remote Chemisorption on Metals - ICl/Al(111):

(Andrew C. Kummel)

The principle method of converting magnetic films into device is etching by halogen.

The first step in this process is the reaction of the halogen with the metal surface. We have bee

investigating this process using molecular beams techniques. The chemisorption of diatomic

halogens on aluminum and other low work function substrates is described as remote

dissociation; this process is analogous to the carefully studied alkali + ICl reactions. As the

halogen molecule approaches the surface, a harpooning electron is ejected from the surface to the

halogen molecule causing it to dissociate instantaneously. This process is associated with exoion

emission (ejection of negatively charged halogen atoms) that has a probability of 10-6 - 10-12

and exoelectron emission that has a probability of 10-3 - 10-9. However, little is known about

the dominant process. We have hypothesized that abstractive chemisorption should be very

common since an ejected halogen atom needs only a little kinetic energy to escape the attractive

potential of the surface compared the energy required for a halogen ion to escape its image

charge.

We have studied the chemisorption of ICl on Al(111) using non-resonant multiphoton

ionization, Auger spectroscopy, and sticking measurements to determine the chemisorption

dynamics. Using the reflection technique, we have determined that the sticking probability is

60% at high incident translational energy and is independent of the surface temperature. Using

Auger spectroscopy, we have determined that the ratio of iodine to chlorine on the surface is 5:1

for 0.05 to 1.2 eV ICl dosing of 110 K Al(111) at low coverage (<0.02 ML). This strongly

suggests that the dominant chemisorption process is abstractive chemisorption in which the

surface harpoons an electron preferentially to ICl molecules oriented with the iodine atoms

110

directed toward the surface. Upon formation of ICl-, the ICl dissociates to I- and Cl; the I-

chemisorbs to the surface while the Cl atoms are ejected back into the gas phase. To confirm

this hypothesis, we have used non-resonant MPI to detect the reaction products for 1.2 eV ICl on

110 K Al(111). As shown in Figure 13.1, in the incident beam we detect ICl+, the impurity I2+,

and the photodissociation products of I+ and Cl+, but in the scattered beam we only detect Cl+.

Therefore, when abstractive chemisorption occurs, it always results in chemisorption of iodine

and ejection of chlorine back into the gas phase. Harpooning preferentially occurs to the iodine

end of the molecule where the lowest unoccupied molecular orbital, *, is concentrated. When

ICl- dissociates in free space, it usually forms I + Cl- because Cl has a higher electron affinity

than I. However, since the iodine-end of ICl is closer to the surface during harpooning, the

iodine's electron affinity is increased by its image charge; therefore, the ICl + Al(111) results in

the preferential adsorption of iodine

Figure 13.1. MPI Spectra of ICl/Al(111) MPI Time of Flight Spectrum of the 1.2 eV ICl beam

reflected from the 110 K Al(111) surface. The left peak is just from photoelectrons. The only ion peak (Cl+) originates from the Cl abstractive chemisorption product. There is no detectable I+ peak. This proves that ICl/Al(111) abstraction always occurs by formation of an Al-I site and ejection of Cl back into the gas phase.

111

14. Abstractive Chemisorption of O2/Al(111)

(Andrew C. Kummel)

After a metallic magnetic film is deposited and exposed to air, an oxide layer forms. We

are using STM to determine how this process occurs. Our first studies have been on aluminum

since it is reactive and absorbates can be imaged with atomic resolution at low coverage at 300K

on Al(111). The chemisorption of O2 on Al(111) has been a source of great controversy because

the original STM experiment by Ertl's group showed that adsorption of thermal O2 on 300 K

Al(111) forms isolated oxygen atoms more than 80 Å apart. This was originally interpreted as

being the result of long range dissociation, but several theorists felt this was impossible due to

the large corrugation in the O/Al(111) potential. We have repeated this experiment using high

translational energy O2 molecular beams. We observe that at high translational energy, instead

of observing just single sites, we observe a 1:1 mixture of single and double sites (see Figure

14.1). We can be assured of the identity of the atomic O sites because we observe the exact

same O sites in the islands of oxygen atoms whose structure is known from LEED. We feel the

single sites are due to abstractive chemisorption while the double sites are due to dissociative

chemisorption; this would be consistent with theoretical investigations predicting that at high

translational energy the momentum towards the surface is sufficiently large that both abstraction

and long range dissociation are suppressed.

112

Figure 14.1. 50 Å x 50 Å STM Image of 0.5 eV O2/Al(111). This image shows the

chemisorption of O2 at very low coverage. There are three types of oxygen adsorbates, but all the oxygen appears at bright spots under these tip/tunneling conditions. There are single (“S”), isolated sites from abstractive chemisorption. There are double sites (“P” pairs) from dissociation. Finally there are islands of oxygen atoms. It seems that the chemisorption probability is greater near existing islands so the islands dominate with increasing coverage. At low translational energy the double sites are absent.

15. Measuring Mechanical Properties of Nanowires using depth Sensing and Force

Modulation

(S.A. Syed Asif, K.J. Wahl and R.J. Colton)

With the development of new nanostructured materials and continuing

miniaturization of engineering and electronic components, thin films and surface coatings,

there is a need to understand the mechanical properties of materials at the nanoscale.

Conventional uniaxial testing, macro or micro indentation techniques cannot be used for such

small volumes, and optical measurements of indent sizes are not possible. A logical approach

is to replace the optical microscope of a microindenter by an electron microscope. However

imaging of indents using electron microscopy suffers from the disadvantage of being time

113

consuming and can give large errors if the imaging conditions are not correct. To overcome

these problems, two different techniques and instruments have been developed: (1) the

atomic force microscope (AFM) and (2) the depth-sensing nanoindentation technique. While

both instruments can be used to determine materials properties at the nanoscale, each

technique has distinct advantages and disadvantages. However, we find that coupling the

AFM with depth-sensing indentation and AC force modulation can provide the best of both

techniques. In this report we will present our recent progress in developing quantitative,

surface sensitive nanoindentation techniques for nanostructures and thin films as well as the

applications of the force modulation technique to study adhesion and elastic properties of

thin compliant materials.

In this report we present a typical application of combined depth sensing and imaging

to study the deformation mechanics of nanostructures. Figure 15.1 shows an array of 500 nm

wide and 50 nm thick nickel nanowires on a silicon substrate. This topographic image was

obtained under displacement feedback. To study the mechanical response of this structure,

the image was used to locate and center the indentations on the Ni nanowire.

Figure 15.2 shows the load-depth response for both the Si substrate and Ni wire on Si.

From the unloading curve the modulus (130 GPa) and hardness (11 GPa) of the Si substrate

were calculated via the Oliver and Pharr method. The calculated value of hardness and

modulus agree with reported literature values. The load-depth curve for the Ni nanowire on

Si substrate shows substantial plastic deformation. To study the depth dependence of the

mechanical properties of the nanowires multiple indentation experiments were carried out at

different locations.

114

Figure 15.3 shows the superimposed load-depth response for Ni nanowire at different

loads and different locations. From each unloading curve the modulus and hardness were

calculated as a function of contact depth.

Figure15.4 shows the variation of modulus as a function of contact depth for the Ni

nanowire. The measured modulus is almost constant around 120 GPa, which is close to the

modulus of the Si substrate. The influence of the substrate, even for contact depths ~ 20 nm,

can be clearly seen. Figure15.5 shows the variation of hardness as a function of contact

depth. Unlike modulus, the hardness variation has a clear trend. At shallow depths the

measured hardness is close to that of Ni (1.5-2 GPa); and as the depth increases, the hardness

increases. To understand the possible reason for the increased hardness, an image of the

deformed zone was obtained using the hybrid nanoindenter. Figure15.6 shows the

indentation on a single Ni nanowire. The image clearly shows pile-up and flow of material

along the sides due to the lack of constraint. The procedure used to calculate the hardness is

based on analytical solutions of the elastic contact problem that assumes use of an

axisymmetric rigid punch indenting an isotropic elastic half space. This assumption is not

valid in the present experiment because the stress is not constrained at the side wall and

surface of the nanowire (i.e., there is evidence for flow and pile-up of material). To fully

understand the role of constraint, FEM simulation of the deformation is necessary along with

experiments.

To reduce the substrate influence and to measure the modulus of the 50-nm thick Ni

nanowire, the depth of indenter penetration must be less than 10 nm. In order to perform such

a measurement, the instrument should have good surface sensitivity to detect the contact of

the tip with the specimen surface without causing any damage to the surface. The surface

115

sensitivity of the instrument can be enhanced using force modulation, which is the topic of

the following section.

For force modulation, a small sinusoidal AC force is superimposed on the DC applied

load and the resulting displacement amplitude and the phase shift between the force and

displacement is measured using a DSP lock-in amplifier. The displacement amplitude and the

phase shift are processed to obtain contact stiffness along with the DC load and depth. Figuer

15.7a shows a typical load-depth curve for a Ag nanowire on silicon. The corresponding

contact stiffness as a function of load, measured from force modulation, is shown in Figure

15.7b. The load-depth curve shows a depth discontinuity (pop-in) around 40 µN load. Pop-

in generally occurs due to sudden release of strain energy which could be due to nucleation

and multiplication of dislocations or fracture and debonding of a thin film. From measured

contact Stiffness( sK ), load DCF and depth h, the modulus ∗E and hardness H can be

calculated from following equations.

ππ5.2422 ** hEAEK s == (1)

2

2*4

s

DC

KEFH

π= (2)

Figure 15.8a shows the variation of modulus as a function of depth for Ag nano wire on Si

substrate coressponding to data presented in Figure 15.7a&b. The modulus of the Ag wire is

~75 GPa which agrees with the modulus of bulk Ag [ref]. The modulus remained constant

for depths ranging from 5 nm to 10 nm and began to increase above 10 nm as a result of

substrate influence. The data for depths less than 5 nm are not reliable as the tip shape

calibration is very difficult below 5 nm. The hardness of the Ag nanowire is ~3GPa at depths

of 5-10 nm and increases after that showing behavior similar to the modulus. This

116

experiment clearly demonstrates the cabability of the force modulation technique as it is now

possible to measure the mechanical properties of the nanowires below the 10 nm length scale

and avoid the substrate influence and constraint effect. As the load is further increased to

200 µN, an additional pop-in could be seen (~75 µN, Figure 15.9). The modulus and

hardness calculated from stiffness and load-depth curve is shown in Figure 15.10a&b. The

modulus variation as a function of depth clearly shows the substrate influence at higher

depths. At depth ~45nm the modulus is ~120GPa, which is closer tothe modulus of the Si

substrate. With force modulation we continuosly measure the modulus and hardness variation

as a function of depth with a single indentation experiment. This not only improves the

sensitivity but also reduces the number of experiments needed to get the same information

using the depth sensing indentation technique alone. We speculate that the pop-in behaviour

observed on the Ag nanowires is due to the debonding or failure of the nanostructures. The

pop-in first occurs at a contact stress of ~3-4GPa. From the data, the yield stress of this Ag

nanostructure can be estimated to be ~1Gpa ( H/3, as hardness is actually the mean contact

stress).

Similar experiments were carried out on Ta nanowire. Figure 15.11a shows the load-

depth curve for a Ta nanowire for a maximum load of 100 µN. The load-depth curve looks

nominally elastic without any residual deformation. However the stiffness-load curve shows

hystereses, suggesting some permanent deformation. The modulus and hardness calculated

from the load, stiffness and depth are shown in Figure 15.12a&b. The modulus at 5 nm

depth is ~180 GPa and decreases as the depth of penetration increases, reaching the modulus

of Si at depths greater than 12 nm. This clearly indicates that if the modulus of the film is

higher than the substrate, then the substrate influence can be felt at much shallower depths.

117

The hardness of the Ta nano wire at 5 nm depth is ~ 6GPa and increases with depth of

penetration, approaching the hardness of the Si substrate. The Ta nanowire has higher

modulus (180 GPa/120GPa) and lower hardness (6GPa/11GPa) compared to the Si substrate.

By implementing the force modulation technique we have improved the sensitivity of

the measuring technique and have demonstrated that it is now capable of measuring the

mechanical properties of nanostructures. At shallow depths (below 10nm) we can avoid the

issue of constraint effects to some extent. However at higher depths constraint will be an

additional effect along with the substrate. The deformation mechanics of semi-infinite

surfaces is different from two-dimentional or one-dimentional nanostructures. Although it is

possible to use our experimental technique to measure the mechanical properties of much

smaller nanostructures (10 nm or below), we are limited by the probe tip radius. At this scale

the tip radius is the same size as the nanostructure. Hence for this case, the experiment is no

longer a nanoindentation experiment, it is a nanocompression experiment. To our

knowledge, no uniaxial compression experiment has been done at this scale. We are currently

exploring the possiblity of conducting a nanocompression experiment to measure the

mechanical properties of nanostructures. We have also developed a quantitative stiffness

imaging technique to map the surface mechanical properties of nanostructures, polymer

materials and thin films with a lateral resolution less than a micron [ref]. Finally, we have

also extended the force modulation technique to understand the pre- and apparent contact of

compliant polymer materials, enabling simultaneous measurement of polymer materials

properties (e.g. storage and loss moduli), as well as providing new and improved capabilities

for examining adhesive contacts at the nanoscale (e.g. measuring work of adhesion, strain

energy release rate).

118

Figure 15.1. An array of 500 nm wide and 50 nm thick Ni nanowires on Si substrate.

0

1000

2000

3000

4000

5000

0 50 100 150 200

Load

(µN

)

Depth (nm)

Si Substrate

500 nm Si wire

Figure 15.2. Load-depth response for Si substrate and Ni wire on Si substrate.

119

0

1000

2000

3000

4000

5000

0 50 100 150 200

Load

(µN

)

Depth (nm)

Figure 15.3. Load-depth response for Ni nano-wire at different loads and different locations.

0

50

100

150

200

0 20 40 60 80 100 120

Mod

ulu

s, E

(GP

a)

Contact depth, nm

Figure 15. 4. The variation of modulus as a function of contact depth for the Ni nanowire.

0

5

10

15

20

25

30

0 20 40 60 80 100 120 140

Har

dnes

s, G

Pa

Contact depth, nm

Figure 15. 5. The variation of hardness as a function of contact depth for the Ni nanowire.

120

Figure 15. 6. Image of the indent on a Ni nanowire showing material flow and pile-up.

Figure 15.7. (a)The load-depth and (b) load-stiffness curve for Ag nanowire on Si substrate

Figure 15.8. (a) The variation of modulus and (b) hardness of Ag nanowire on Si Substrate as a function.

1000

2000

3000

4000

5000

6000

7000

0 10 20 30 40 50

Con

tact

Stif

fnes

s, N

/m

Load, µ Νµ Νµ Νµ Ν

500 nm wide and 50nm thick Ag wire on Si5%RH

0

50

100

150

200

250

0 5 10 15

Mod

ulus

, GP

a

Depth, nm

500 nm wide and 50nm thick Ag wire on Si5%RH

0

10

20

30

40

50

0 5 10 15 20

Load

, µΝ

µΝ

µΝ

µΝ

Depth, nm

Pop-in500 nm wide and 50nm thick Ag wire on Si5%RH

121

Figure 15.9. (a)The load-depth and (b) load-stiffness curve for Ag nanowire on Si substrate at 200µN load

Figure 15.10. (a)The modulus and (b) hardness of Ag nanowire on Si Substrate showing

substrate influence at higher load (200µN)

Figure 15.11. (a)The load-depth and (b) load-stiffness curve for Ta nanowire on Si substrate

at 200µN load

0

20

40

60

80

100

120

0 2 4 6 8 10 12

Load

, µµ µµ

ΝΝ ΝΝ

Depth, nm

Ta w ireThickness 50nmWidth 500nm

0

50

100

150

200

0 10 20 30 40 50

Load

, µΝ

µΝ

µΝ

µΝ

Ddepth, nm

50 nm thick Ag film on Si5%RH

Pop-in

2000

4000

6000

8000

1 104

1.2 104

1.4 104

1.6 104

0 50 100 150 200

Con

tact

Stif

fnes

s, N

/m

Load, µµµµ

ΝΝΝΝ

0

50

100

150

200

250

300

350

0 10 20 30 40 50

Mo

dulu

s, G

Pa

Depth, nm

50 nm thick Ag film on Si5%RH

0

2

4

6

8

10

12

0 10 20 30 40 50

Har

dnes

s, G

Pa

Depth, nm

50 nm thick Ag film on Si5%RH

6000

7000

8000

9000

1 10 4

1.1 10 4

1.2 10 4

1.3 10 4

0 2 4 6 8 10 12

Con

tact

Stif

fnes

s, N

/m

Depth, nm

122

Figure 15.12. (a) The modulus and (b) hardness of Ta nanowire on Si Substrate showing

substrate influence at shallower depths

16. "Size Effects in Magnetic and Superconducting Materials"

(Ivan K. Schuller)

Summary of Research

We have developed the technology for the fabrication of nanostructures by several

different methods: electron beam, diblock copolymer, and self assembly. The electron beam

lithography technique has been used to prepare a large number of different structures for

studies of flux pinning in superconductors and for investigations of magnetic hysteresis.

Flux pinning has been extensively studied in a variety of geometrical arrangements of

magnetic dots. In particular, square arrays of dots allow for the first time investigation of the

pinning mechanism. At present, there are several experiments under way which will give a

direct measure of the pinning energy for various magnetic materials and in different magnetic

states (i.e. single or multi domain). Interestingly these experiments also give information on

the magnetic state of the dots, which is being used to investigate purely magnetic

phenomena. Many of these measurements and techniques were developed and performed in

collaboration with Prof. Sharifi's group.

0

100

200

300

400

0 2 4 6 8 10 12

Mod

ulus

, GP

a

Depth, nm

Ta wire Thickness 50 nmWidth 500 nm

0

5

10

15

20

25

0 2 4 6 8 10 12

Har

dnes

s, G

Pa

Depth, nm

123

In collaboration with Profs. S. von Molnar, and S. Pearton we are currently

investigating exchange bias in bilayers of an antiferromagnet (FeF2, initially) and a

nanostructured ferromagnet (Fe, initially). Bilayered films and the electron beam lithography

is performed in Prof. Schuller's laboratory, selective height lithography is done using

techniques developed by Prof. Pearton and collaborators, and measurements using a novel

2D electron gas Hall probe is performed by Prof. von Molnar. We found that

nanostructuring the ferromagnet changes the exchange bias and hysteresis behavior of the

ferromagnetic film. This has important implications for applications in the field of magnetic

sensors and memories which invariably use exchange bias as an integral part of the device.

In collaboration with Prof. S. Hershfield, we are performing a theoretical-

experimental study of the anisotropic magnetotransport in magnetic superlattices (Fe/Cr

initially). We have performed extensive studies of the structure, magnetic and

magnetotransport properties of Fe/Cr superlattices in a large variety of configurations and for

different preparation methods. A particularly unique capability in our lab is the measurement

of the perpendicular resistivity (i.e. with the current perpendicular to the interfaces). This

measurement is more amenable to theoretical calculations. In addition, together with our

quantitative studies, this is probably the most complete study on the interconnection between

structure and magnetotransport in metallic superlattices. In this fashion, we are able to obtain

quantitative data of structural parameters such as roughness, interdiffusion, length scales,

resistivity, magnetoresistance, magnetization, etc. These can be then fed into theoretical

models and obtain quantitative results regarding microscopic parameters.

124

17. Theory of GMR

(Selman Hershfield)

Last year we successfully completed calculations of the CPP GMR in materials like

FeCr. These calculations included realistic electronic structure and different kinds of

scattering - both surface and bulk. This theory compared well with experiments done on the

CPP GMR so we are confident that it is an accurate description.

In this, the final year of this project we wish to carry these calculations to the next

level and use the theoretical model to try optimize the experimental parameters and produce

the maximum possible GMR. This is the so-called inverse band-structure problem of finding

an atomic configuration with given electronic properties (see Franceschetti and Zunger,

Nature 402, 60 (1999).) For our case the desired electronic property is the maximum possible

GMR.

My senior graduate student on this project, Tat-Sang Choy, is in the process of

optimizing our GMR code so that many different runs can be done. Each run contains a

slightly different configuration and an optimization technique called simulated annealing is

used to choose the next atomic configuration. Note that simulated annealing as an

optimization tool and should not be confused with real physical annealing.

Tat-Sang Choy has already improved the running speed of the code by several orders

of magnitude so that what used to take a day to run now only takes minutes. The

developments he has made have been well documented and will form the major part of his

thesis so we should be able to transfer this knowledge effectively to others in the field.

125

18. Ion Beam Sputter Deposition of GMR Materials

(D. Temple)

I. Summary

MCNC's Role in the Program

During the second phase of the MURI program (years 4 and 5), the statement of work for

MCNC, a subcontractor to the University of Florida, included two major tasks:

1. Development of ion beam sputter deposition (IBSD) techniques for deposition of GMR

multilayers, using the IBSD system designed, constructed, and automated in-house.

2. Device processing and photolithographic mask/design and fabrication support for other

organizations in the MURI program.

Progress During the Last Reporting Period

Task 1:

• Demonstrated deposition of polycrystalline Fe/Cr multilayers exhibiting the GMR effect,

with GMR ratios comparable to the best ones obtained by RF sputtering.

• Examined effects of variations of the primary ion beam energy and the type of ions on GMR

values; the investigated primary ion energy range was 700 eV- 1200 eV for Ar ions and 900

eV - 1200 eV for Xe ions. Demonstrated that the GMR ratio is greater for films deposited

using Xe ions than for films deposited using Ar ions, and that for both types of ions the GMR

ratio increases as the primary ion beam energy decreases.

• Modeled the observed dependence of the GMR ratio on the primary ion beam energy value

and the type of ions via a correlation of the measured GMR values with the average energy

transferred to the substrate during the film growth. This energy was calculated using the

Transport of Ions in Matter (TRIM) algorithm based on the Monte Carlo method.

• Modified the configuration of the IBSD equipment to expand the range of primary ion beam

energies to include lower deposition energies, potentially optimizing GMR ratios of the Fe/Cr

multilayer system. These modifications included: a) changes in the vacuum pumping system

126

to provide lower base and operating pressures as well as shorter pumping and pump

regeneration cycles; b) a rebuild of the ion gun consisting of the replacement of the quartz

chamber and installation of curved accelerator grids designed to reduce the ion beam

spreading; and c) the installation of magnetic sensors to monitor the position of metal targets

as they rotate into place during the deposition of multilayers.

Task 2:

• A series of Si/SiO2/metal capacitors were fabricated for studies of the screening length of thin

magnetic films; the studies were conducted in Art Hebard's group at the University of

Florida.

• Magnetic metal multilayer electrodes were fabricated for spin field-effect transistor studies;

the studies were conducted in Fen Ren's group at the University of Florida.

• Three sets of photolithographic masks were designed and fabricated for the University of

Florida, Florida State University and University of California in San Diego.

II. Detailed Description - Ion Beam Sputter Deposition Development

Effect of Processing Parameters on Magnetic Properties of Fe/Cr Multilayers

The configuration of the IBSD system has been described in detail in progress report #4.

For this study, we have deposited the following Fe/Cr multilayer combinations:

Si/Cr(50Å)/[Fe(20Å)/ Cr(t(Å))]x30/Cr(50Å- t(Å)), where t was varied from 8 to 14 Å; this range

surrounds the first antiferromagnetic maximum in the Fe/Cr multilayer system. The deposition

rates varied from 5 to 30 Å/min depending on the primary ion beam energy, the type of ions, and

the target material. The films were deposited at room temperature.

Deposited multilayers were characterized using Transmission Electron Microscopy

(TEM), Atomic Force Microscopy (AFM), Auger Electron Spectroscopy (AES), X-ray

Photoelectron Spectroscopy (XPS) as well as resistivity, magnetic hysteresis loop and

magnetotransport measurements. Film resistivity was determined from sheet resistance values

obtained using a four-point probe. Magnetic hysteresis loop measurements were performed

using a vibrating sample magnetometer (VSM). Magnetotransport measurements were obtained

127

at the University of Florida (in the group of Dr. Arthur Hebard, one of the investigators in the

MURI program) using two different instruments. The first of these, a Physical Properties

Measurement System (PPMS, Quantum Design) allows measurements as a function of

temperature (2-400K), magnetic field (0-7T), and angle (0-360 degrees). The second instrument,

a Magnetic Properties Measurement System (MPMS, Quantum Design), can also be used for

magnetoresistance measurements. Its temperature and field ranges are similar to those of the

PPMS.

Among other structural measurements, low angle X-ray diffraction (XRD) was performed

on selected films to verify the composition of the multilayers. As an example, Figure 1 shows a

low angle XRD spectrum for a film with a nominal composition of

Si/Cr(50Å)/[Fe(20Å)/Cr(10Å))]x30/Cr(40Å), obtained using a BEDE D3 diffractometer using

Cu radiation from a rotating anode source operating at 40 kV and 120 mA. The scan was run

from 0 to 16,000 arcsec with a stepsize of 25 arcsec and a counting time of 30 sec/step. The

structural parameters, such as density, thickness and roughness, were fit using a commercially

available autofit algorithm (BEDE Mercury REFS). All parameters were allowed to vary

independently. The only constraint was that the roughness for a layer could not be greater than

the nominal thickness of that layer. The fit resulted in the value of 30 Å for the modulation

period, in perfect agreement with the nominal value, and in the average interface roughness value

of 5 Å.

128

1x10-1

1x100

1x101

1x102

1x103

1x104

1x105

1x106

0 2500 5000 7500 10000 12500 15000 17500

2 Theta (sec)

Figure 1. Low angle XRD spectrum of a Si/Cr(50Å)/[Fe(20Å)/Cr(10Å))]x30/Cr(40Å) multilayer showing first and second low angle Bragg peaks, the angular position and intensity of which is dependent on the multilayer period and average interface roughness.

Figure 2 shows experimental data for the GMR ratio and zero-field resistivity for Si/Cr(50Å)/[Fe(20Å)/Cr(10Å))]x30/Cr(40Å) multilayers, deposited with Ar ions of varying primary energy. The shaded area covers the primary energy range outside of the original optimum operation range of the ion gun in the IBSD system. As expected, depositing films with the primary ion beam energy in this range resulted in incorporation of impurities into the multilayer films due to beam spreading beyond the confines of the sputtering targets. This beam spreading caused an increase in the film resistivity as well as a strong decrease in the GMR ratio due to the segregation of the Al impurities to the interfaces of the growing film. The films deposited outside the shaded primary ion beam energy range, i.e., between 700 eV and 1200 eV, were free of impurities, as evidenced by chemical characterization of the films using AES or XPS. The resistivity of these films was constant, and the trend in the GMR ratio was clear: the GMR ratio increased as the primary ion beam energy decreased.

As discussed in detail in the previous report, using Xe ions in place of Ar ions resulted in a dramatic increase in the GMR ratio. Si/Cr(50Å)/[Fe(20Å)/Cr(10Å))]x30/Cr(40Å) multilayers deposited using Xe ions exhibited a GMR ratio of 26%, in comparison with the 11% value obtained for multilayers of the same composition deposited using Ar ions of the same energy.

1st Bragg Peak

2nd Bragg Peak

129

0

2

4

6

8

10

12

0 200 400 600 800 1000 1200

Primary Ion Beam Energy (eV)

GM

R (%

)

0.0E+0

2.0E-5

4.0E-5

6.0E-5

8.0E-5

1.0E-4

1.2E-4

Res

istiv

ity (o

hm c

m)

Si/Cr(50A)/[Fe(20A)/Cr(12A)]x30/Cr(38A) Ar ion beam

Figure 2. GMR ratio and zero-field resistivity for Fe/Cr multilayers deposited using Ar ions with

varying primary ion beam energy. The shaded area corresponds to the "forbidden" energy range for the then-current configuration of the ion gun.

The fact that the GMR ratio is dependent on both the type of ions as well as on the primary ion beam energy points to a possible correlation between the GMR values and energies of species arriving at the substrate. These energetic species include sputtered atoms and reflected neutrals resulting from neutralization of the primary ion beam as it approaches the sputtering target. To explore this correlation, we have calculated energy distributions for sputtered Fe and Cr atoms, as well as for the reflected neutrals. The energy distributions vary as functions of the type of primary ions and the primary energy of the ions. These calculations were performed using a commercial algorithm (Transport of Ions in Matter, TRIM) based on the Monte Carlo method. Based on the energy distributions, we have calculated the average energy per sputtered atom deposited onto the substrate during the multilayer growth. Figure 3 presents a plot of the measured GMR ratio for the Fe/Cr multilayers as a function of the calculated average energy per atom. As can be seen from the figure, as the average energy increases, the GMR ratio decreases. This result can be attributed to the increased probability of creating defects that form non-spin-

130

dependent scattering centers, and the increased degree of intermixing at interfaces associated with an increase of an average energy per atom.

05

1015202530

10 15 20 25 30

Calculated Average Energy Per Atom (eV/atom)

GM

R ra

tio (%

)

Figure 3. Correlation between experimentally determined GMR ratios for

Si/Cr(50Å)/[Fe(20Å)/Cr(10Å))]x30/Cr(38Å) multilayers and calculated values of the average energy per sputtered atom deposited into the substrate during the multilayer growth. This energy is a function of the type of ions and the primary ion beam energy.

II2. IBSD System Modifications

The necessary modifications of the IBSD system performed during the reporting period included changes in the vacuum pumping system to provide lower base and operating pressures as well as shorter pumping and pump regeneration cycles. Figure 4 shows a schematic diagram of the IBSD system before and after the modifications performed on the pumping configuration. The "before" configuration relied on the roughing pump to achieve a pressure of about 300 mTorr in the deposition chamber. At that point in the pumpdown sequence, the gate valve isolating the roughing pump from the rest of the system was closed, and the pressure in the chamber was decreased further using a cryogenic pump. This cryopump was responsible for reaching and maintaining ultra-high vacuum (UHV) before and after deposition, as well as maintaining a pressure in the 10-4 Torr range during the deposition. Since the cryopump was not isolated from the rest of the system in the "before" configuration, the regeneration of the pump, a part of the routine maintenance, required a shutdown of the system. When Xe is used as the source gas, the regeneration of the cryopump must be done frequently, because Xe is susceptible to freezing on the entry baffle of the pump, inhibiting its pumping efficiency.

131

Figure 4. Schematic diagrams for the IBSD system before and after the modification.

Chamber

Cryo pump

Sample Intro

“Small” Turbo

Load Rod

Roughing Pump

Diaphragm Pump

Isolation Valve

BEFORE

AFTER

Chamber

Cryo pump

Sample Intro

“Small” Turbo

Load Rod

Roughing Pump

Diaphragm Pump

GateValv

e

(future

“Large” Turbo 500 L/sec

Roughing Pump

53 L/sec

132

The "after" configuration has several improvements. First, the addition of a high throughput turbomolecular pump (500 liters/second) decreases the amount of time for pumping the chamber down in preparation for the deposition. This turbo-pump is capable of bringing the base pressure in the chamber down to the 10-8 Torr range. The combination of the turbo pump and the cryopump is then responsible for lowering the pressure down to the 10-9 Torr regime.

The gate valve in front of the turbopump is another important addition. It remains in the open position unless there is a problem with the turbopump or the roughing pump. In that case, the pumps can be isolated from the chamber. The gate valve at the cryopump offers a similar advantage. In addition, the cryopump can be isolated from the chamber during most depositions which eliminates its exposure to high pressures of the source gases and extends the period of time between the regeneration of the pump, minimizing the system downtime.

The ion gun rebuild involved replacing the quartz chamber with a new chamber and the installation of a set of focusing accelerator grids. The grids have a curved surface that causes focusing of the ion beam and reduces the beam spreading between the ion source and the target. This modification was implemented to enable operation of the ion gun with lower primary energy values than previously possible.

As an example of the effects of the performed modifications, Figure 6 shows an XPS spectrum from a monitor Cr film deposited at a beam energy of 500 eV using Xe ions in the modified system. As can be seen, the spectrum indicates that the film is free from Al impurities. The aluminum impurities were present in films deposited in the previous configuration of the system.

Figure 5. XPS spectrum of a Cr film deposited using 500 eV Xe+ ions. Note the absence of Al

in the film, indicating a more focused ion beam, as desired.

Another modification that is scheduled to be implemented in the future is the addition of a second ion gun. A second ion gun would provide a means to execute dual ion beam deposition. This will allow for intentional, controlled intermixing of individual films, incorporation of magnetic dopants at the interfaces, and modification of the underlayer structure. These experiments, along with others that might be added, are designed to take advantage of the IBSD's flexibility in varying deposition parameters. This flexibility is an important asset in the quest for

133

a greater understanding of the mechanisms involved in the GMR effect. Another modification that will enhance the capabilities of the current IBSD system is the installation of the shutter assembly. The electrically actuated shutter will protect the substrate during the initial beam current rise time, as well as during precleaning of the targets.

III. Plans For the Remainder of the Program

• Continue investigations of the effect of the primary ion beam energy on magnetoresistive properties of Fe/Cr multilayers using the extended ion beam energy range made possible by the recent system modifications.

• Optimize the GMR ratio for Co/Cu multilayers and/or spin valves as a function of IBSD deposition conditions. Correlate the GMR values with the average energy metric to arrive at material-specific, but deposition-system-independent process condition windows.

• Continue to support University of Florida and other participating institutions by providing wafers with high resolution liftoff patterns for patterning of magnetic multilayers, performing TEM analyses of deposited materials, designing masks/reticles for use in photolithographic processes and providing device fabrication services.

• Optimize processes for formation of contacts to patterned magnetic multilayers using single and dual level metallization schemes.

134

Technology Transitions

1. Ultra-high density patterning process (> 64 Mbit·in-2) based on e-beam lithography for

future generations of rad-hard, infinitely rewritable Magnetic Random Access Memories (MRAMs) – work done in collaboration with Honeywell Solid State Technology Center, MN.

2. Growth of high quality CoFeB and low temperature NiMnSb thin films (< 250oC), both of which are attractive candidates for next generation information storage devices – work done in collaboration with IBM Almaden Research Center, San Jose, CA.

3. Development of the non-corrosive Co/NH3 plasma chemistry for shallow feature etching (e.g. MRAMs) – work done in conjunction with Plasma-Therm, St. Petersburg, FL and Seagate, MN.

4. Development of new Kerr probe for ultra-high B field magneto-optical measurements – work done in conjunction with National High Magnetic Field Laboratory, Tallahassee, FL, where the probe is available as a user facility.

5. Maskless approaches to ultra-high density patterning for future generations of rad-hard, infinitely rewritable Magnetic Random Access Memories (MRAMs) – work initiated in collaboration with Honeywell Solid State Technology Center, MN.

6. Discovered a dry etching process for NiFe and related alloys based on a balance between chemical surface reaction with Cl, and ion-assisted desorption of the reaction products, followed by in-situ removal of Chlorine resides to prevent corrosion. This is being transitioned to Honeywell for use in fabricating high density, rad-hard, non-volatile magnetic Random Access Memories (MRAMs).

7. Growth of the first high quality, potentially 100% spin-polarized NiMnSb thin films at low (250oC) temperatures. These are attractive candidates for advanced magnetic storage devices with improved Giant Magneto-Resistance response.

8. Development of techniques for reducing the ultimate limits of e-beam lithography, and resultant achievement of individual features with dimension < 300Å, and dense arrays (> 10 Gbits·in-2) of 500Å features for next generation information storage devices.

9. Growth of improved nanoscale SrS:Ce, F thin films which emit in the blue, using addition of GaS during rf-magnetron deposition. These have application for full color electroluminescent displays. A plasma etching process for patterning of these films has also been discovered (work done in collaboration with Planar Systems, the only domestic supplier of EL displays).

135

Interactions and Industrial Contacts 1. IBM Almaden Research Center, San Jose, CA – new nanoscale materials and processes

for computer hard devices 2. Honeywell Solid State Electronics Center, Plymouth, MN – new processes for radiation-

hard memory. 3. Honeywell Space Systems, Clearwater, FL – next generation MRAM for use in

submarine-based missile systems. 4. Plasma-Therm, St. Petersburg, FL – new etch processes for magnetic materials. 5. Seagate, Minneapolis, MN – new etch processes for magnetic material. 6. IBM T.J. Watson Research Center, Yorktown heights, NY - new etch processes for

magnetic materials. 7. Motorola, Tempe, AZ – advanced processing for MRAM. 8. Bell Laboratories, Lucent Technologies – deposition/etching of advanced dielectrics for

memory devices. 9. Piezo Technology, Inc., Orlando, FL – characterization/etching of new oxides for ultra-

precise timing applications (oscillators). 10. Corning, Corning, NY – advanced dielectrics for magnetic sensors. 11. Naval Research Laboratory, Washington, DC – methods for measuring mechanical

properties at the nanoscale, and fabrication of spin-FETs. 12. Sandia National labs, Albuquerque, NM – high rate etching processes for magnetic

multilayers. 13. Allied Signal, Columbia, MD – e-beam lithography for ultra-dense patterning. 14. American Crystal Technology, Fremont, CA – growth of dilute magnetic semiconductors. 15. US Army Research Lab, Adelphi, MD – magnetic sensors.

136

Publications “Large Magnetic Entropy in GMR a-Gdx Si1=x,” B. Zink, E. Junard, K. Allen and F. hellman, Phy. Rev. Lett. 83, 2266 (1999). “Low Temperature MR in Insulating a-Gdx Si1-x Alloys,” P Xiong, B.L. Zink, S. Apllebaum and F. Hellman, Phys. Rev. B59, 3929 (1999). “Parametric Study of NiFe and NiFeCo High Density Plasma Etching Using CO/NH3,” K.B. Jung, H. Hong, H. Cho, S. Onishi, D. Johnson, Y.D. J.R. Childress and S.J. Pearton, J. Electrochem. Soc. 146, 2163 (1999). “ICP Etching of CoFeB, CoZr, CoSm and FeMn Thin Films in Interhalogen Mixtures,” H. Cho, K.B. Jung, D.C. Hays, Y.B. Hahn, T. Feng, Y.D. Park, J.R. Childress, F.J. Cadieu, R. Rami, X.R. Qian, L. Chen and S.J. Pearton, Mat. Sci. Eng. B60, 107 (1999). “Cl2-based ICP Etching of CoFeB, CoSm, CoZr and FeMn,” K.B. Jung, H. Cho, Y.B. Hahn, D.C. Hays, T. Feng, Y.D. Park, J.R. Childress and S.J. Pearton, Mat. Sci. Eng. B60, 101 (1999). “Selective Dry Etching using ICP, Part I: GaAs/AlGaAs and GaAs/InGaP,” D.C. Hays, H. Cho, K.B. Jung, Y.B. Hahn, C.R. Abernathy, S.J. Pearton, F. Ren and W.S. Hobson, Appl. Surf. Sci. 147, 125 (1999). “Selective Dry Etching using ICP, Part II: InN/GaN and InN/AlN,” D.C. Hays, H. Cho, K.B. Jung, Y.B. Hahn, C.R. Abernathy, S.J. Pearton, F. Ren, J. Han and R.J. Shul, Appl. Surf. Sci. 147, 134 (1999). “Effect of Inert Gas Additive Species on Cl2 High Density Plasma Etching of Compound Semiconductors Part I: GaAs and GaSb,” Y.B. Hahn, D.C. Hays, H. Cho, K.B. Jung, C.R. Abernathy, S.J. Pearton and R.J. Shul, Appl. Surf. Sci. 147, 207 (1999). “Effect of Inert Gas Additive Species on Cl2 High Density Plasma Etching of Compound Semiconductors Part II: InP, InSb, InGaP and InGaAs,” Y.B. Hahn, D.C. Hays, H. Cho, K.B. Jung, C.R. Abernathy, S.J. Pearton and R.J. Shul, Appl. Surf. Sci. 147, 215 (1999). “Comparison of Cl2 and F2 based Chemistries for the ICP Etching of NiMnSb Thin Films,” J. Hong, J. Caballero, E.S. Lambers, J.R. Childress and S.J. Pearton, J. Vac. Sci. Technol. A17, 1326 (1999). “Damage to III-V Devices During ECR-CVD,” J.W. Lee, K. MacKenzie, D. Johnson, R.J. Shul, Y. Hahn, D.C. Hays, C.R. Abernathy, F. Ren and S.J. Pearton, J. Vac. Sci. Technol. A17, 2183 (1999).

137

“Effect of Inert Gas Additive on Cl2-based ICP Etching of NiFe and NiFeCo,” K.B. Jung, H. Cho, Y.B. Hahn, D.C. hays, E.S. Lambers, Y.D. Park, T. Feng, J.R. Childress and S.J. Pearton, J. Vac. Sci. Technol. A17, 2223 (1999). “Study of ICP NH3 Plasma Damage on GaAs Schottky Diodes,” L.C. Meyer, J.W. Lee, D. Johnson, M. Huang, F. Ren, T.J. Anderson, J.R. LaRoche, J.R. Lothian, C.R. Abernathy and S.J. Pearton, J. Electrochem. Soc. 146, 2717 (1999). “Cl2-based Dry Etching of Doped Manganate Perovskites: PrBaCaMnO3 and LaSrMnO3,” K.P. Lee, K.B. Jung, H. Cho, D. Kumar, S.V. Pietambaram, R.K. Singh, P.H. Hogan, K.H. Dahmen, J.B. Hahn and S.J. Pearton, J. Electrochem. Soc. 146, 2748 (1999). “Novel In-Situ Ion Bombardment Process for a Thermally Stable (7800oC) Plasma Deposited Dielectric,” Electrochemical and Solid-State Letters 2, 537, (1999). “Dry Etching of BaSrTiO3 and LaNiO3 Thin Films in Inductively Coupled Plasmas,” K.P. Lee, K.B. Jung, A. Srivastava, D. Kumar, R.K. Singh and S.J. Pearton, J. Electrochem. Soc. 146, 3778 (1999). “Inductively Coupled Plasma Etching of Ta2O5,” K.P. Lee, K.B. Jung, R.K. Singh, S.J. Pearton, C. Hobbs and P. Tobin, J. Electrochem. Soc. 146, 3794 (1999). “Mechanism of High Density Plasma Etch Processes for Ion-driven Etching of Materials,” J.W. Lee, J.F. Donahue, K.D. MacKenzie, R. Westerman, D. Johnson and S.J. Pearton, Solid-State Electron. 43, 1769 (1999). “Fabrication and Magneto-transport and SQUID Measurements of Sub-micron Spin-valve Structures,” Y.D. park, D. Temple, K.B. Jung, D. Kumar, P.H. Holloway and S.J. Pearton, J. Vac. Sci. Technol. B17, 2471 (1999). “Development of Chemically-assisted Dry Etching Methods for Magnetic Device Structures,” K.B. Jung, H. Cho, K.P. Lee, J. Marburger, F. Sharifi, R.K. Singh, D. Kumar, K.H. Dahmen and S.J. Pearton, J. Vac. Sci. Technol. B17, 3186 (1999). “Plasma Etching of Magnetic Multilayers-effect of Concurrent UV Illumination,” H. Cho, K.P. Lee, Y.B. Hahn, E.S. Lambers and S.J. Pearton, Mat. Sci. Eng. B67, 145 (1999). “Long-term Stability of Dry Etched MRAM Elements,” K.B. Jung, J. Marburger, F. Sharifi, Y.D. Park, E.S. Lambers and S.J. Pearton, J. Vac. Sci. Technol. A18, 268 (2000). “Dry Etch Selectivity of Gd2O3 to GaN and AlN,” D. Hays, K.P. Lee, B.P. Gila, F. Ren, C.R. Abernathy and S.J. Pearton, J. Electron. Mater. 29, 285 (2000).

138

“Comparative Study of Ni Nanowires Patterned by e-beam Lithography and Fabricated by Life-off and Dry Etching Techniques,” Y. Park, K.B. Jung, M. Overberg, D. Temple, S.J. Pearton and P.H. Holloway, J. Vac. Sci. Technol. B18, 16 (2000). “Thermal Stability and Etching Characteristics of e-beam Deposited SiO and SiO2,” J. LaRoche, F. Ren, J. Lothian, J. Hong, S.J. Pearton, E. Lambers, C.H. Chu, C.S. Wu and M. Hoppe, J. Vac. Sci. Technol. B18, 283 (2000). “Ultraviolet Light Enhancement of Ta2O5 Dry Etch Rates,” K.P. Lee, H. Cho, R.K. Singh, S.J. Pearton, C. Hobbs and P. Tobin, J. Vac. Sci. Technol. B18, 293 (2000). “Effect of UV Illumination on SiC Dry Etch Rates,” P. Leerungnawarat, H. Cho, S.J. Pearton, C.-M. Zetterling and M. Östling, J. Electron. Mater. 29, 342 (2000). “Low Temperature SiNx and SiO2 Film Processing by ICP-CVD,” J.W. Lee, K.D. MacKenzie, D. Johnson, J.N. Sasserath, S.J. Pearton and F. Ren, J. Electrochem. Soc. 147, 1481 (2000). "Magnetization Behavior of Nanometer-scale Iron Particles," S. Wirth, M. Field, D. D. Awschalom and S. von Molnár, Phys. Rev. B 57 (1998), R 14028-14031. "Magnetism of Nanometer-scale Iron Particle Arrays, " S. Wirth, M. Field, D. D. Awschalom and S. von Molnár, J. Appl. Phys. 85 (1999), pp.5249-5254. "Magnetic Interactions In Nanometer-Scale Particle Arrays Grown Onto Permalloy Films," S. Wirth and S. von Molnár presented at MMM'99, J. Appl. Phys. 87 (2000) pp.7010-7012. "Thermally Activated Magnetization Reversal In Nanometer-Size Iron Particles," S. Wirth, A. Anane and S. von Molnár (to be published). “Synthesis and Characterization of Silica-Coated Iron Oxide Nanoparticles in Microemulsion: The Effect of Non-Ionic Surfactants,” S. Santra, R. Tapec, N. Theodoropoulou, A. F. Hebard, W. Tan, submitted to Langmuir (6/20/2000). “Superparamagnetism in Discontinuous Ni Films,” A. Frydman, T. Kirk and R.C. Dynes, Solid State Communication 114, 481 (2000). “Magnetoresistance of Granular Ferromagnets - Observation of a Magnetic Proximity Effect,” A. Frydman and R.C. Dynes, Solid State Communications 110, 485 (1999). ‘Superconducting Tunneling as a Probe of Sputtered Oxide Barriers,” C.L. Platt, A.S. Katz, R.C. Dynes and A.E. Berkowitz, Appl. Phys. Lett. 75, 127 (1999). “Disorder-induced Andreev Reflections in Granular Metals,” A. Frydman and R.C. Dynes, Phys. Rev. B 59, 8432 (1999).

139

“Superconducting and Spin-Dependent Tunneling Using Native Oxide Barriers on Co-Fe Thin Films,” C.L. Platt, A.S. Katz, E.P. Price and R.C. Dynes, Phys. Rev. B 61, 68 (2000). “Nanoscale Surface Mechanical Property Measurements: Force Modulation Techniques Applied to Nanoindentation,” S.A.S. Asif, R.J. Colton, and K.J. Wahl, accepted for publication in Interfacial Properties on the Submicron Scale, R. Overney and J. Frommer, eds., ACS/Oxford Press, Jan. 2000. “Quantitative Study of Nanoscale Contact and Pre-Contact Mechanics Using Force Modulation,” S.A.S. Asif, K.J. Wahl, and R.J. Colton, Thin Films: Stresses and Mechanical Properties VIII, , R. Vinci, O. Kraft, N. Moody, P. Besser, E. Shaffer II, eds., Vol. 594 (Materials Research Society, Warrendale, PA, 2000). “The Influence of Oxide and Adsorbates on the Nanomechanical Response of Silicon Surfaces,” S.A. Syed Asif, K.J. Wahl, and R.J. Colton, J. Mater. Res. 15 (2000) 546-553. “Nanoindentation – Quantitative Study of the Nanomechanical Properties of Materials Using Depth Sensing and Force Modulation,” S.A. Syed Asif, K.J. Wahl, O. Warren, and R.J. Colton, to be published in “SXM Industrial Use,” R.J. Colton and H. Fuchs, Eds. (Springer Verlag, submitted).

"Suppression of Growth-Induced Perpendicular Magnetic Anisotropy in Co-Pt Alloys by Trace Amounts of Si", A. L. Shapiro, O. Vajk, B. M. Maranville, and F. Hellman, Appl. Phys. Lett. 75, 4177 (1999). "Growth-Induced Perpendicular Anisotropy and Clustering in NixPt1-x alloys.” A. L. Shapiro, D. Vasumathi, B. M. Maranville, and F. Hellman, paper submitted to J. Magn. Magn. Mat. (2000). "Growth-Induced Anisotropy on Vicinal Substrates,” B. M. Maranville, A. L. Shapiro, D. Vasumathi, and F. Hellman, paper in preparation, to be submitted to J. Appl. Phys. (2000). “Ion-Beam Assisted Growth of Perpendicular Anisotropy Materials, “D. Vasumathi, B. M. Maranville, and F. Hellman, paper in preparation, to be submitted to J. Appl. Phys. (2000). “Magnetic and Thermodynamic Features Of Antiferromagnetic Nanoparticles In A Metallic Matrix,” R. Sappey, E. P. Price, F. Hellman, A. E. Berkowitz, and D. J. Smith, in preparation; to be submitted to Physical Review. “Oxygen Content and Crystallinity Effects in Pulsed Laser Deposited Lanthanum Manganite Thin Films,” Srinivas V. Pietambaram, D. Kumar, Rajiv K. Singh, and C. B. Lee, Proceedings of the 2000 MRS Spring Meeting, San Francisco, CA, USA “Artificially Induced Reconfiguration of the Vortex Lattice by Arrays of Magnetic Dots,” Jose Martin, M. Velez, A. Hoffmann, Ivan K. Schuller, and J.I. Vicent, Phys. Rev. Lett. 83, 1022 (1999).

140

“Periodic Vortex Pinning with Magnetic and Non-Magnetic Dots: Does the Size Matter?,” A. Hoffmann, P. Prieto, and Ivan K. Schuller Phys. Rev. B 61, 6958 (2000). “Metallic Superlattices,” Ivan K. Schuller, Physics Today – Invited Article (To appear, 2000). "Processing Techniques for InGaAs/InAlAs/InGaAs Spin Field Effect Transistors," J.R. LaRoche, F. Ren, D. Temple, S.J. Pearton, J.M. Kuo, A.G. Baca, P. Cheng, Y.D. Park, Q. Hudspeth, A.F. Hebard, and S.B. Arnason, to be published in Solid State Electronics (2000). "Comparative Study of Ni Nanowires Patterned by E-Beam Lithography and Fabricated by Lift-off and Dry Etching Techniques," Y.D. Park, K.B. Jung, M. Overberg, D. Temple, S.J. Pearton and P.H. Holloway, J. Vac. Sci. Technol. B18 (2000) 16. "Fabrication and Magneto-Transport and SQUID Measurements of Sub-Micron Spin-Valve Structures," Y.D. Park, D. Temple, K.B. Jung, D. Kumar, P.H. Holloway, and S.J. Pearton, J. Vac. Sci. Technol. B17 (1999) 2471.

141

Technical Presentations “Electrical Noise from Phase Separation in PrCaMnO3 Single Crystal,” A. Prange, S. von Molnar, L. Pinsaid-Godast and A. Revcolevschi, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Magnetic Interactions in nm-Scale Particle Arrays Grown onto Permalloy Films,” S. Wirth and S. con Molnar, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Investigating Artificial Barriers in Spin-Dependent Tunnel Junctions with Superconducting Electrodes,” C. Platt, A. Berkowitz, A.S. Katz and R.C. Dynes, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Anisotropic Magnetotransport and Microstructural Analysis of Fe/Cr Superlattices,” M. Cyrille, M.E. Gomez, C. Leighton and I.K. Schuller, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Effect of the Antiferromagnetic Spin-Hop in Exchange Bias,” J. Nogues, C. Morellon, M. Ibarra, C. Leighton and I.K. Schuller, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Exponential Thickness Dependence and Nonlinear I-V Curves: Do they Establish Tunneling,” B. Fousson, R. Escadero, C. Leighton, A. Romero, S. Kim, I.K. Schuller, M. Grossman and D. Rabson, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “45o Exchange Coupling Across the Fe-Fe2 Interface,” M. Fitzsimmons, D. Yashar, C. Leighton, I.K. Schuller, J. Nogas and J. Dura, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Coercivity in the Positive Exchange Bias Regime,” C. Leighton, I.K. Schuller and J. Nogues, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, November 1999. “Spin Polarized Photoemission Study of Magnetic Films – Evidence for Half-Metallic Ferromagnetic Behavior,” S. Morton, G. Waddill, J. Tobin, S. Kim and I.K. Schuller, Spring MRS, San Francisco, April 2000. “Transient and Steady State Electrochemical Effects and its Correlation to CMP Removal Rates During Metal Removal,” R.K. Singh, U. Mahajan, S.-M. Lee, Z. Chen and D. Lamholdt, Spring MRS, San Francisco, April 2000. “Fundamental Studies on the Mechanisms of Oxide CMP,” U. Mahajan, S. Lee and R.K. Singh, Spring MRS, San Francisco, April 2000. “Particulate Effect in Cu CMP,” S. Lee, U. Mahajan, A. Nagory and R. Sing, Spring MRS, San Francisco, April 2000.

142

“In-Situ Friction-Force Measurements in CMP,” V. Mahajan, S. Lee and R. Singh, Spring MRS, San Francisco, April 2000. “Fundamental Studies on Iodate Slurry Chemistries During CMP pf Cu,” S. Lee, U. Mahajan, V. Cracium and R.K. Singh, Spring MRS, San Francisco, April 2000. “Epitaxial ZnO Films Grown by UV-Assisted PLD,” V. Cracium, N. Bassim, R.K. Sing, J. Perriere and D. Crasium, Spring MRS, San Francisco, April 2000. “Role of O Content and Crystallinity in Magnetoresistance Behavior of CLMO Then Films,” S. Pietambaram, D. Kumar, R.K. Singh and C.B. Lees, Spring MRS, San Francisco, April 2000. “Low Temperature Growth of BaSrTiO3 Thin Films by UV-Assisted PLD,” V. Crasium, J. Howard, N. Bassim, R.K. Singh and J. Pierre, Spring MRS, San Francisco, April 2000. “UV-Assisted PLD of Thin Films,” V. Crasium and R.K. Singh, Spring MRS, San Francisco, April 2000. “Magnetic and Magnetoresistance Properties of PLD LaCaMnO3 on Si,” D. Kumar, J. Narayan, R.K. Singh, C.B. Lee and J. Sankar, Spring MRS, San Francisco, April 2000. “Modelling of Interfacial Scattering Effects During Light Emission from Phosphor Then Films for Field Emission Displays,” R.K. Singh, K. Cho, Z. Chen and D. Kumar, Spring MRS, San Francisco, April 2000. “Nanunctionalized Sulfide-Band Powders for Flat Panel Display Applications,” M. Ollinger, V. Crasium and R.K. Singh, Spring MRS, San Francisco, April 2000. “The Effect of Microstructure on the Brightness of PLD Y2O3:Eu Then Film Phosphors,” K. Cho, D. Kumar, R. Singh, G. Russel and B.K. Wagner, Spring MRS, San Francisco, April 2000. “Stoichiometry Effects of Li on the Electrochemical Porperties of LiMn2O4 Films Grown by Laser Ablation,” D. Singh, H. Hofmann, V. Carcium, R.K. Singh and J. Pierre, Spring MRS, San Francisco, April 2000. “Room Temperature Growth of ITO Films by UV-Assisted PLD,” V. Cracium, R.K. Singh and D. Cracium, Spring MRS, San Francisco, April 2000. “Effects of Co-Dopants on the EL Properties of ZnS:Tb,” P. Holloway, J. Kim, M. Davidson and B. Speck, Spring MRS, San Francisco, April 2000. “Electron Tunnelling Measurements on the CMR Perovskites,” H. Hudspeth, P. Xiong, S. von Molnar and F. Sharifi, 2000 March Meeting of APS, Minneapolis, March 2000.

143

“Discovery of a New Undoped nFi System – U2Co2Sn,” G. Stewart, J. Kim, S. Getty and F. Sharifi, 2000 March Meeting of APS, Minneapolis, March 2000. “Transport Measurements of Electron and Hole-doped CaB6,” S. Getty, F. Sharifi, D. Young and Z. Fisk, 2000 March Meeting of APS, Minneapolis, March 2000. “Dependence of the CTP-GMR on Spin-Independent Scattering in Fe/Co Superlattices,” T. Choy, S. Hershfield and J. Chen, 2000 March Meeting of APS, Minneapolis, March 2000. “A Database of Fermi Surfaces in Virtual Reality Modelling Language,” T. Choy, J. Naset, S. Hershfield, C. Stanton and J. Che, 2000 March Meeting of APS, Minneapolis, March 2000. “Magnetic Bound States of SrCu2(BO3)2,” V. Kobor and S. Hershfield, 2000 March Meeting of APS, Minneapolis, March 2000. “Zero-bias Anomalies in Magnetic Hexabodies,” S. Hershfield and V. Kobor, 2000 March Meeting of APS, Minneapolis, March 2000. “Systematic in the Behavior of Co60 Monolayers Deposited Linearly onto Then Films and Doped by Electron Transfer,” Q. Hudspeth, S. Arason and A.F. Hebard, 2000 March Meeting of APS, Minneapolis, March 2000. “Experimental Determination of Screening Length in Thin Magnetic Films,” K. McCarthy, N. Theodoropoloa and D. Temple, 2000 March Meeting of APS, Minneapolis, March 2000. “The Influence of Percolation on Quantum Coherence in Coalescing Ag Films,” S. Arason and A.F. Hebard, 2000 March Meeting of APS, Minneapolis, March 2000. ‘Spin-Peierls Transition in NaV2O5 in High Magnetic Fields,” A.F. Hebard, S. Bornpudre, V. Kotov, D. Hall, V. Bass and T. Palstra, 2000 March Meeting of APS, Minneapolis, March 2000. “Direct Measurement of the g-Factor in Crystalline Bi at High B/T,” S. Bompadre, C. Biagini, D. Maslov and A.F. Hebard, 2000 March Meeting of APS, Minneapolis, March 2000. “Microcalorimetry: Wide Temperature Range, High Field and Small Sample Measurements,” F. Hellman, 2000 March Meeting of APS, Minneapolis, March 2000. (Invited) “Critical Phenomena of LaSrMnO3,” D. Kim, F. Hellman and J. Coey, 2000 March Meeting of APS, Minneapolis, March 2000. “Ion Beam Assisted MBE Growth of Magnetic CoPt3,” D. Vasumalli, B. Marauville and F. Hellman, 2000 March Meeting of APS, Minneapolis, March 2000. “Perpendicular Magnetic Anisotropy of CoPt3 on Vicinal Substrates,” B. Maranville, A. Shapiro, F. Hellman and E.T. Yu, 2000 March Meeting of APS, Minneapolis, March 2000.

144

“Magnetic Filed Driver Change of the Density of States of a-GdxSi1-x at the Metal-Insulation Transition,” W. Teitzer, F. Hellman and R.C. Dynes, 2000 March Meeting of APS, Minneapolis, March 2000. “Magnetic and Thermodynamic Features of Antiferromagnetic Nanoparticles in a Metallic Matrix,” R. Sappey, E. Price, F. Hellman and A. Berkowitz, 2000 March Meeting of APS, Minneapolis, March 2000. “Local Movements and Localized Conduction Electrons in a-GdxSi1-x,” B. Zink, D. Queen, R. Sappey, E. Janod and F. Hellman, 2000 March Meeting of APS, Minneapolis, March 2000. “Infrared Spectroscopy of NdCeCuOx,” E. Singley, A. Katz, S. Woods, R.C. Dynes and K. Yamada, 2000 March Meeting of APS, Minneapolis, March 2000. “Epitaxial Growth of CrO2 Films by CVD from Cr8O21 Precursors,” P. Ivanov, S. Watts, D. Lind and S. von Molnar, 2000 March Meeting of APS, Minneapolis, March 2000. “Electrical Noise Tide of the Percolative Conduction in LaCaMnO3,” A. Anane, B. Baquet and S. von Molnar, 2000 March Meeting of APS, Minneapolis, March 2000. “Temperature Evolution of Magnetic Scattering in Half-Metallic Chromium Dioxide,” S. Watts, S. Wirth, S. von Molnar, A. Barry and J. Coey, 2000 March Meeting of APS, Minneapolis, March 2000. “Electron Tunneling Measurements on the CMR Perovskites,” F. Sharifi, P. Xiong and S. von Molnar, Fall MRS, Boston, December 1999. (Invited) “Quantitative Study of Nanoscale Contact and Pre-Contact Mechanics Using Force Modulation,” S. Syed Asif, K. Wahl and R.J. Colton, Fall MRS, Boston, December 1999. “Relationship Between Microstructure and Luminescent Properties of Epitaxially Grown Y2O3:Eu Thin Films on LaAlO3,” H. Guo, S. Pennycook, D. Kumar and R.K. Singh, Fall MRS, Boston, December 1999. “Low Temperature Growth of BaSrTiO3 by UV-Assisted PLD,” V. Cracium, J. Howard, R.K. Singh and J. Perriere, Fall MRS, Boston, December 1999. “Characteristics of ZnO Films Grown by UV-Assisted PLD,” V. Cracium, J. Howard and R.K. Singh, Fall MRS, Boston, December 1999. “Dielectric Passivation/Oxides on Compound Semiconductors,” F. Ren, M. Hong, S.J. Pearton, C.R. Abernathy, G. Dang and J.R. Lothian, 46th Int. Symp. AVS, Seattle, Oct. 1999.

145

“Advanced Selective Dry Etching of GaAs/AlGaAs in High Density Inductively Coupled Plasmas,” J.W. Lee, M. Devre, B. Reelfs, D. Johnson, J. Sasserath, F. Clayton and S.J. Pearton, 46th Int. Symp. AVS, Seattle, Oct. 1999. “Comparison of Plasma Chemistries for Dry Etching of Ta2O5,” K.P. Lee, K.B. Jung, R.K. Singh, S.J. Pearton, C.C. Hobbs and P. Tobin, 46th Int. Symp. AVS, Seattle, Oct. 1999. “Effects of UV Illumination on Dry Etch Rates of NiFe-Based Magnetic Multilayers,” H. Cho, K.P. Lee, K.B. Jung, Y.B. Hahn, S.J. Pearton and R.J. Shul, 46th Int. Symp. AVS, Seattle, Oct. 1999. “High Breakdown Voltage (Au/Pt/GaN Schottky Diodes,” J.I. Chyi, J.M. Lee, C.C. Chuo, G.C. Chi, G. Dang, A.P. Zhang, X.A. Cao, MM. Mshewa, F. Ren, S.J. Pearton, S.N.G. Chu and R.G. Wilson, 46th Int. Symp. AVS, Seattle, Oct. 1999. “ICP-Induced Etch Damage of GaN p-n Junctions,” R.J. Shul, A.G. Baca, L. Zhang, C.G. Willison, S.J. Pearton and F. Ren, 46th Int. Symp. AVS, Seattle, Oct. 1999. “Schottky Diode Measurements of Dry Etch Damage in n- and p-type GaN,” X.A. Cao, Z.P. Zhang, G. Dang, F. Ren, S.J. Pearton, R.J. Shul and L. Zhang, 46th Int. Symp. AVS, Seattle, Oct. 1999. “Effect of N2 Discharge Treatment on AlGaN/GaN HEMT Ohmic Contact Using ICP,” A.P. Zhang, G. Dang, X.A. Cao, F. Ren, S.J. Pearton, J.M. Van Hove, P.P. Chow, R. Hickman and J.J. Klaassen, 46th Int. Symp. AVS, Seattle, Oct. 1999. “High Density Plasma Etching of Ta2O5-Selectivity to Si and Effect of UV Light Enhancement,” K.P. Lee, h. Cho, R.K. Singh, S.J. Pearton, C. Hobbs and P. Tobin, MRS Fall Meeting, Boston, Nov. 1999. “High Density Plasma Etching of (BaSr)TiO3 and LaNiO3,” K.P. Lee, K.B. Jung, A. Srivastava, D. Kumar, R.K. Singh and S.J. Pearton, MRS Fall Meeting, Boston, Nov. 1999. “Ion Enhanced Dry Etching of Magnetic Multilayers: Post-Etch Cleaning and Effects of UV Illumination,” H. Cho, K. Lee, K.B. Jung, S.J. Pearton, F. Sharifi and J. Marburger, MRS Fall Meeting, Boston, Nov. 1999. “A Unified Approach to Modelling of Etched Profiles in an ICP Etching System,” Y. Im, Y.B. Hahn and S.J. Pearton, 5th Intl. Workshop on Advanced Plasma Tools & Process Engineering, Santa Clara, CA, Feb. 2000. “Corrosion-Free Dry Etch Patterning of MRAM Stacks – Effects of UV Enhancement,” H. Cho, K.-P. Lee, K.B. Jung, S.J. Pearton, J. Marburger, F. Sharifi and J.R. Childress, 44th Ann. Conf. Magnetism and Magnetic Materials, San Jose, CA, Nov. 1999.

146

“Dry etching of MRAM Structures,” S.J. Pearton, H. Cho, K.B. Jung, J.R. Childress, F. Sharifi and J. Marburger,” 2000 Spring MRS Meeting, San Francisco, CA, April 2000. “Low DiT Dielectric/GaN MOS Systems,” M. Hong, H. Ng, J. Kwo, A. Korkan, J. Baillargeon, S. Chu., J. Mannaerts, A.Y. Cho, F. Ren, C. Abernathy, S.J. Pearton and J.I. Chyi, 197th Meeting of the ECS, Toronto, May 2000. "Magnetism Of Nanometer-Scale Iron Particle Arrays," S. Wirth MPI for Chemical Physics of Solids, Dresden, January 19, 1999. "Magnetisierungsverhalten Regelmäßiger Anordnungen Von Ferromagnetischen Nanometerteilchen," S. Wirth, IMW Dresden (IFW), January 20, 1999. "Magnetism Of Nanometer-Scale Iron Particle Arrays," S. Wirth, Laboratoire Louis Neel, CNRS Grenoble, January 25, 1999. "Magnetism Of Nanometer-Scale Iron Particle Arrays," S. Wirth, CMRR (UCSD), San Diego, February 9, 1999. "Magnetism of Nanometer-Scale Iron Particle Arrays," S. Wirth and S. von Molnár APS March Meeting 2000, Minneapolis, USA. (Invited) “Experimental Determination of the Screening Length in Thin Magnetic Films,” K. T. McCarthy, N. A. Theodoropoulou, A. F. Hebard (Department of Physics, Gainesville Fl 32611-8440), Dorota Temple (MCNC, Electronics Technologies Division, Research Triangle Park, NC 27709-2889), APS March meeting (3/20-3/24). "Growth-induced perpendicular anisotropy and clustering in NixPt1-x alloys.” A. L. Shapiro, D. Vasumathi, B. M. Maranville, and F. Hellman, Conference presentation March APS (2000). "Growth-induced anisotropy on vicinal substrates,” B. M. Maranville, A. L. Shapiro, D. Vasumathi, and F. Hellman, Conference presentation March APS (2000). “Ion-beam assisted growth of perpendicular anisotropy materials, “D. Vasumathi, B. M. Maranville, and F. Hellman, Conference presentation March APS (2000. “Sliding Transitions and Dissipation in Nanoscale Contacts,” K.J. Wahl and S.A. Syed Asif, American Vacuum Society National Symposium, Seattle, WA, 25-29 October, 1999. “Measuring and imaging contact stiffness quantitatively at the nanoscale using force modulation,” S.A. Syed Asif, K.J. Wahl and R.J. Colton, American Vacuum Society National Symposium, Seattle, WA, 25-29 October, 1999.

147

“Quantitative Study of Nanoscale Contact and Pre-Contact Mechanics using Force Modulation,” S.A. Syed Asif, K.J. Wahl and R.J. Colton, Materials Research Society Fall Meeting, Boston, MA, Dec 1999. “Quantitative Nanoscale Surface Mechanical Properties of Polymers and Thin Films,” K.J. Wahl, S.A.S. Asif, and R.J. Colton, SPM in Biomaterials Conference, Bristol, UK, June 23 2000. (Invited) “Mechanics, Sliding Transitions and Dissipation in Nanoscale Contacts,” K.J. Wahl presented at Gordon Research Conference on Tribology, Holderness, NH, July 2000. (Invited) “Nanoscale surface mechanical properties of polymer thin films and organic monolayers,” K.J. Wahl, S.A.S. Asif, and R.J. Colton, 219th National Meeting of the American Chemical Society, San Francisco, CA, 26-31 March 2000. (Invited) “Nanoscale surface mechanical properties of polymer thin films and organic monolayers,” K.J. Wahl, S.A.S. Asif, and R.J. Colton, 220th National Meeting of the American Chemical Society, Washington DC, 20-24 August 2000 (Invited) “Quantitative Imaging of Dynamic Mechanical Properties by Hybrid Nanoindentation,” S.A. Syed Asif, K.J. Wahl, and R.J. Colton, American Vacuum Society National Symposium, Boston, MA, 2-6 October 2000. “Force-modulated nanoindentation of fluorinated polymer thin films grown by PECVD,” S.A. Syed Asif, E.J. Winder, K.K. Gleason, and K.J. Wahl, American Vacuum Society National Symposium, Boston, MA, 2-6 October 2000. “Quantitative Study of Nanoscale Mechanical Properties of Nanostructures,” S.A. Syed Asif, K.J. Wahl, and R.J. Colton Symposium T: Fundamentals of Nanoindentation and Nanotribology II, Fall MRS Conference, Boston, MA, 27 November-1 December 2000. “Sliding Transitions, Mechanics and Dissipation in Nanoscale Contacts,” K.J. Wahl, S.A.S. Asif, and R.J. Colton, Symposium T, Dynamics in Small Confining Systems VI, Fall MRS Conference, Boston, MA, 27 November-1 December 2000. (Invited) “Flux Pinning in a Superconductor by an Array of Submicron Magnetic Dots,” Ivan K. Schuller, Euroconference on Vortex Matter…, Crete, Greece, September 18-24, 1999. (Invited)

"Dependence of the CPP-GMR on Spin-independent Scattering in Fe/Cr Superlattices," Tat-Sang Choy, Selman Hershfield, and Jian Chen, March Meeting of the American Physical Society, Minneapolis Minnesota, March 20-24, 2000. "Database of Fermi Surfaces in Virtual Reality Modeling Language," Tat-Sang Choy, Jeffery Naset, Selman Hershfield, Christopher Stanton, and Jian Chen, March Meeting of the American Physical Society, Minneapolis Minnesota, March 20-24, 2000.

148

"Zero-bias anomalies in magnetic hexaborides," Selman Hershfield and Valeri Kotov, March Meeting of the American Physical Society, Minneapolis Minnesota, March 20-24, 2000. "Ion Beam Sputter Deposition of GMR Materials,"G. E. McGuire, D. Temple, M. Ray, J. Lannon, and A.F. Hebard, invited talk at the Annual Symposium of the Mexican Vacuum Society, Mexico, 1999. "Giant Magnetoresistive Films Grown by Ion Beam Sputter Deposition," G.E McGuire, D. Temple, J. M. Lannon, C.C. Pace, and M.A. Ray, invited talk to be presented at the International Workshop on Smart and Functional Film Deposition for VLSI Applications, November 2000, Nagoya, Japan.

149

Personnel Supported Postdoctoral Associates and Visiting Researchers

S. Wirth Post Doctoral Fellow at FSU (now at the Max Planck Institute, Dresden, Germany)

H. Cho Post Doctoral Fellow at UF Y.B. Hahn Visiting Professor at UF M. Coey Visiting Professor at FSU Jian Chen Post Doctoral Fellow at UF Y. Park Post Doctoral Fellow at UF (now at NRL)

J.I. Martin Visiting Professor at UCSD M. Velez Visiting Professor at UCSD J. Nogues Visiting Professor at UCSD J.L. Vincent Visiting Professor at UCSD J.-M. George Visiting Professor at UCSD E.M. Gonzalez Visiting Professor at UCSD P. Prieto Visiting Professor at UCSD Y. Jaccard Post Doctoral Fellow at UCSD

M.C. Cyrille Post Doctoral Fellow at UCSD Aviad Frydman Post Doctoral Fellow at UCSD W. Teizer Post Doctoral Fellow at UCSD

A.S. Katz Post Doctoral Fellow at UCSD Axel Hoffmann Post Doctoral Fellow at UCSD D. Kumar Research Scientist at UF

Bernard Revaz Visiting Scholar at UCSD Graduate Students

D. Kent with Dr. Pearton X. Cao with Dr. Pearton (now at Alpha Photonics) K.P. Lee with Dr. Pearton D.C. Hays with Dr. Pearton (now at Sony Corporation) H. Hudspeth with Dr. Sharifi (now at GE) J. Marburger with Dr. Sharifi T.S. Choy with Dr. Hershfield S.A. Getty with Dr. Sharifi S. Khan with Dr. Sharifi J. Howard with Dr. Singh G.T. Dang with Dr. Ren E. Price with Dr. Hershfield A. Shapiro with Dr. Hershfield Zhihong Chen with Dr. Hershfield Tat-Sang Choy with Dr. Hershfield A. Hoffmann with Dr. Schuller S.V. Pietambaram with Dr. Singh B. Maranville with Dr. Hellman

150

A. Shapiro with Dr. Hellman N. Theodoropoulou with Dr. Hebard K.T. McCarthy with Dr. Hebard Casey Pace with Dr. Temple (a student at the Department of Physics of the University of

North Carolina in Chapel Hill)

Undergraduate Students

T. Plew with Dr. Pearton T. Kirk with Dr. Dynes

151

Degrees Awarded

Heather Hudspeth Ph.D. University of Florida 2000

“Electron Tunnelling Measurements on Ferromagnetically-Doped Lanthanum Manganite Films”

K.B. Jung Ph.D. University of Florida 1999 “High Density Plasma Etching of Magnetic Materials” A. Srivistava M.S. University of Florida 1999

“Growth and Characterization of BaSeTiO3 Films with Enhanced Electrical Properties using PLD”

D.C. Hays MS University of Florida 1999 “Selective Etching of Compound Semiconductors” X.A. Cao Ph.D. University of Florida 2000 “Advanced Processing for Novel Devices” K.P. Lee MS University of Florida 1999 “Dry Etching of Novel Dielectric Films” K. Majumdar Ph.D. University of Florida 1999 “Study of Transport Properties in Magnetic Nanostructures” A. Hoffmann Ph.D. University of California – San Diego 1999 “Fundamental Studies of Magnetism” A.S. Katz Ph.D. University of California – San Diego 1999

"Fabrication, Characterization and Analysis of Nanofabricated Ion Damage High Temperature Josephson Junctions"