NanoSbudurcd Materiels. Vol. 9. pp. 13-22.

1997
Elsevia Science Ltd
Q 1997 Acta Mdallurgica Inc.
Printed in the USA. All rights maved
0965~9773D7 $17.00 + .oo
PII so9659773(97)00014-7

SYNTHESIS OF NANOSTRUCTURED MATERIALS BY

MECHANICAL MILLING: PROBLEMS AND OPPORTUNITIES

C. C. Koch
MamialsscienceandEngineeringDepartment
North Carolina State University, Raleigh, N. C. 27695, USA

Abstract - Mechanical attritionas a method toproduce nanocrystalline(nc) materials

is reviewed Its advantages include the fat thatall classes of materials- including brittle
compounds - are amenable to the method; it can be easily scaled up to tonnage quantities.
The phenomenology and suggested mechanisms for formation of nc microstructures are
discussed for ball milling of single component powders, mechanical alloying of multi-
component powders, and mechanical crystallization of amorphous alloys. The
phenomenology is well documented but microscopic mechanisms await better unaktanding
of the nature of &formation processes in nc materials. The problems of contaminationand
pow&r consolkiationare brieflyconstaked 01997 Acta MetallurgicaInc.

INTRODUCI’ION

A wide variety of techniques are being used to synthesize nanostructured materials

including inert gas condensation (1) rapid solidification (2), electrodeposition (3), sputtering
(4), crystalkation of amorphous phases (S), and chemical processing (6). Mechanical attrition
-ballmilling- which induces heavy cyclic deformation in powders, is a technique which
has also been used widely for preparation of nanostructured materials (7). Unlike many of the
above methods, mechanical attrition produces its nanostructures not by cluster assembly but by
the suuctural decomposition of coarser-grained structures as the result of severe plastic
deformation. This has become a popular method to make nanoaystalline materials because of
its simplicity, the relatively inexpensive equipment (on the laboratory scale) needed, and the
applicability to essentially all classes of materials. The major advantage often quoted is the
possibility for easily scaling up to tonnage quantities of material for various applications.
Similarly, the serious problems that are usually cited are 1) comammation from milling media
and/or atmosphere, and 2) the need (for many applications) to consolidate the powder product
without coarsening the nanocrystalline microstructure. In fact, the contamination problem is
often given as a reason to dismiss the method_at least for some materials.
This paper will review the mechanisms presently believed responsible for formation of
nanocrystalline structures by mechanical attrition of single phase powders, mechanical alloying
of dissimilar powders, and mechanical crystallixation of amorphous materials. The two
important problems of contamination and powder consolidation will be briefly considered.

13
14 CC KOCH

THE EVOLUTION OF NANOCRYSTALLINE STRUCTURES IN SINGLE

PHASE MATERIALS BY MECHANICAL ATTRITION

The present understanding of the development of a nanocrystalline microstructure by ball

milling has been reviewed previously (7,8,9). This information comes mainly from a TIM
and high resolution TEM (I-IREM)study of ball-milled Ru and AlRu microstructures as a
function of milling time (10,ll). Using x-ray line broadening to measure “grain” sixe it is
commonly observed that the crystallite size decreases with milling time. In the TEM and
HREM study (11) of AlRu it was observed that at the early stages of milling the deformation
was localked within shear bands that am approximately 0.5 to 1 em wide. The shear bands
contain a high dislocation density. At a given strain these dislocations annihilate and
recombine to small angle grain boundaries separating individual grains. Small grains, 8-12
mn diameter, are seen within the shear bands and electron diffraction patterns suggest the
n&orientation angles between the grains are relatively small. At longer milling times the
grain size decreased steadily, consistent with the x-ray results, and the shear bands coalesced.
The small angle boundaries were replaced by higher angle boundaries, implying grain rotation,
as reflected by the disappearance of texture in the electron diffraction patterns as well as the
random orientation of the grains observed from the lattice fringes in HREM. The minimum
grain size obtainable by milling has been attributed to a balance between the defect/dislocation
structure introduced by the plastic deformation of milling and its recovery by thermal processes
(12). It has been found that the minimum grain size induced by milling scales inversely with
the melting temperature of the group of fee metals studied (12). These data are plotted in
Figure 1 along with data for bee and hcp metals (lO,ll), Si (13,14), and C (graphite (15)). For
these data, only the lower melting point (Al, Ag. Cu, Ni) fee elements show a clear inverse
dependence of minimum grain size on melting temperature. The minimmn grain size data for
the bee and hcp metals, and for the fee metals with the higher melting temperatures (r TM for
Pd), exhibit esseutially constant values with melting temperature. For these elements it
appears that &in. is iu the orderz fce&cc&zp. However, before explanations for the above
based on the strain hardening response or other fundamental differences in deformation behavior
for the various metallic crystal structures can be considered, it should be pointed out that a
number of variables can influence the values of the minimum grain size attained by ball
milling. First of all, most of the measurements reported are based on analysis of x-ray
diffraction line broadening. Such analysis is subject to difficulty in terms of absolute
quantitative values for grain sixe (16). However, most of the data in Figure 1 are from the Cal
Tech group (10,11,12)or from Oleszak and Shingu (17). Both these groups used the approach
of Williamson and Hall (18) to estimate the crystallite sixe and mot-mean-square (rms) strain
from the x-ray line broadening. In spite of the facts that the mill energies and contamination
levels were very different in these studies, mmarkable agreement for minimum grain size, d is
found for Al, Cu, Ni, and Fe. The value of d for W is lower (5.5 nm) for Oleszak and Shingu
(17) compared to that (9 nm) of Hellstem et al (11). The Cal Tech group (10,11,12) used a
high energy shaker mill (Spex 8000) while Oleszak and Shingu (17) used a conventional
horizontal low energy ball mill. The metallic impurity level (Fe) from the shaker mill is
much larger (2 2 at.% Fe) than that from the conventional ball mill (< 0.1 at.% Fe).
Conversely, the reported oxygen concentrations were about 4 times higher for powders milled
in the conventional mill than for those from the shaker mill. This recent observation
 SYNTHESIS
OF NANOSTRUCTUAED
MATERIALS
BYMECHANICAL
MILUNG 15

26

16 Al \\
16 I
\ Ni
14 ‘ii%
12
10
6i

ot I I I I I
1000 2000 3000 4000 5000

Melting Temperature (K)

Figure 1. Minimum grain size for nc elements vs. their melting temperatures (open
symbols (10,11,12); fiiled circles (17); Si (13.14); C (15)).

(17) that the minimum nanocrystalline grain sizes for a number of elements milled in a low
energy mill are comparable to those milled in a high energy mill (10,11,12) is contrary to
conclusions made previously on the minimum grain size in milled TiNi (19). The
nanocrystalline grain size was compared after milling TiNi in a high energy Spex shaker mill
and a lower energy vibratory mill. After about 10h in the Spex mill the grain size (determined
by TEM and x-ray line broadening analysis) decreased to about 4-5 nm. At longer milling
times an amorphous structure was observed. Milling for 1OOhin the vibratory mill resulted in a
grain size of about 15 nm. We originally assumed this value represented a saturation to the
minimum grain size obtainable in the lower energy mill. However, in light of the work of
Oleszak and Shingu (17) it is likely that the &. was not obtained at 1OOhin the lower energy
mill and that continued milling may have reduced the grain size further. These new results
suggest that total strain, rather than milling energy or ball-powder-ball collision frequency, is
responsible for determining the nanocrystalline grain size. This is different from amorphization
(20,21) or disordering (22) induced by ball miIling where it appears the energy and frequency of
ball-powder-ball collisions determine the final structures formed in “driven systems”. It is,
however, consistent with observations of nanocrystallites formed by high strain values using
other non-cyclic deformation methods (23-32).

VARIABLES IN THE FORMATION OF NANOCRYSTALLINE

MICROSTRUCTURES BY MECHANICAL ATTRITION

Mill Energy

As pointed out above, in elemental powders, similar ultimate nanocrystalline grain sizes
have been achieved in high energy shaker mills (lOJl.12) and conventional low energy ball
mills (17). These results suggest mill energy per se is not critical to the final microstructure
16 CC KOCH

but of course the kinetics of the process are dependent on the energy, and times for attaining the
same microstructure can be several orders of magnitude longer in the low energy mills than in
high energy mills.

MillingTemperature

Milling temperature has been observed to affect the rate at which the nanocrystalline
structure develops. The milling time at which a given grain size was attained in a TiNi
intermetallic compound was a function of milling tempemnue (19). In this case amorphization
occnrmd at a “critical”grain size of about 4-5 nm so the final nanocrystalline grain size at each
milling temperature could not be determined. Shen and Koch (33) also observed smaller nc
grain sizes in both Cu and Ni milled at -85°C compared to samples milled at room
temperature. For example for Cu d = 26 f 3 nm for room temperature milling and d = 17 f 2
nm for milling at -85°C. Evidence for smaller nanocrystalline grain size at low milling
temperatures was also obtained on the intermetalliccompound CoZr (34).

Alloy Eflects

It has been suggested that the ultimate grain size achievable by milling is determined by
the minimum grain size that can sustain a dislocation pile-up within a grain and by the rate of
recovery during milling (12). To estimate the composition dependence of grain size after
milling one may use the formula suggested by Nieh and Wadsworth (35). The minimum
distance L between dislocations is L = 3Gb/n(l-v)h with shear modulus G. Burgers vector b.
Poisson ratio v, and hardness h. If this is indeed related to the formation of a minimum nc
grain size, then this nc grain size is inversely proportional to hardness. In fact, a decreasing nc
grain size with solute concentration is observed in nc alloy systems which exhibit solid
solution hardening such as Cu(Fe) (36,37). Ti(Cu), Nb(Cu), Cu(Ni), and Cu(Co) (37). Also
consistent with this is the essentially constant nc grain size in Ni(Co) where hardness does not
change significantly (33) and the increased grain size in nc Ni(Cu), Fe(Cu), and Cr(Cu) which
exhibit a solid solution softening effect (36).

MRCHANISM(S) FOR FORMATION OF NANOCRYSTALLINR

MICROSTRUCTURES BY MECHANICAL ATTRITION

While there is some agreement on the phenomenology of the development of a nc grain

structure by ball milling as described above, a detailed microscopic model is still lacking,
related at least in part to our lack of understanding of the deformation mechanisms in nc
materials.
Fecht (9) has summarized the phenomenology of the development of a nanocrystalline
microstructnre by mechanical attrition into three stages, namely;

stage 1. deformation localizationin shear bands containing a high dislocation density.

 SVNTHESISOF NANOSTRUCTURED
MATERIALS
BYMECHANICAL
MILLING 17

stage 2. annihilation/recombiition/ rearrangement to form a cell/subgrain

dislocation
structure with nanoscale dimensions - further milling extends this structure
thrrughout the sample.
stage 3. the orientation of the grains becomes ran- i.e. low angle grain boundaries
+ high angle grain boundaria - grain boundary sliding, rotation is likely.

From evidence of the atomic level lattice strain and the stored enthalpy as a function of
reciprocal grain size (or milling time) it was concluded that two different regimes can be
distinguished i.e. dislocation versus grain boundary deformation mechanisms (9). The lattice
strain in the fee elements studied by Eckert et al (12). (milled in a high energy shaker mill) was
found to increase continuously with decreasing grain size and reached a maximum value at the
smallest grain size. This is in contrast to the earlier observation on Ru and AlRu which
indicated a maximum in strain vs. l/d (10) and the recent study of Olesxakand Shingu (17) in a
low energy mill which also shows a broad maximum in strain vs. l/d for a number of elements
including several fee elements. The lattice strain values availablefrom the literature are plotted
against reciprocal grain size, l/d, in Figure 2. With the exception of Ru, the data for

l.l_

l.O_

0.9_

0.6_

k
3 0.7,

= 0.6_
2
z 0.6_

0.4 _

0.3 -

0.2 -

0.1 -

0.00 0.06 0.10 0.15

l/d (nm-1)

Figure 2. Lattice strain vs. reciprocal grain size. (Al, Ni, Fe, W (17); Ru (10); Pd (12); Ti
(56)).

increasing lattice strain with l/d appear to fall on a common relatively narrow band before
decreasing from the maximum strain values. However, these data are from several groups
using mills with various energy levels and possible differences in milling temperature. It has
been demonstrated (33) that lower milling temperature resulted in larger values for lattice strain
for Cu and Ni. However, the data from the low energy mill (17) should be self-consistent and
 18 CC Kow

exhibits interesting behavior. That is, the strain rises with decrea&g grain sire, reaches a
maximum, and then decreases to low values for the smallest nanocrystalline grain sizes. A
maximum in strain with l/d was explained previously by either a change in deformation
mechanism from dislocation generation and movement to grain botmdary sliding, grain rotation
(lo), or, in brittle intermetallics to fmctnre after the strain maximum is reached (10,38). The
f~tmechanismisunlikelyinthiscasesincethencgrainsizecontinuestodecreaseaftertbe
strain maximum is reached and grahr boundary sliding, grain rotation presumably cauuot result
in grain size reduction. lhere is no clear evidence at present in the elements for the fracture
mechauism which would also presumably result in a finer particulate distribution.
Additional information to help explain the mechankm of nanocrystallite formation comes
from measurements of stored enthalpy. Maxima in stored enthalpy vs. l/d are typically
observed (10,12,17). However, the maximum in stored entbalpy is usually found at smaller
grain sixes than the strain maximum, as illustrated for W (17) in Figure 3. Here the maximum
insaainoccursatd=8.3nmwhilethemaximuminstoredenthalpyisatd=5.5nm. Several
suggestions have been offered to explain the maxima in stored enthalpy with l/d including
decreasing strain (10,17) or impurity pick-up during milling (10). The latter is unlikely for the
samples from low energy milling (17) where metallic impurity contmnination is negligible. It
is stated that the stored enthalpy comes mainly from grain boundaries (12.17) and grain
boundary strains. Stress relaxation may be responsible for the maxima (17) but, as above, the
ti and stored enthalpy maxhua don’tnecessarilycoincide.

0.1

l/d (nm-1)

Figure 3. Stored enthalpy (AI-I)and lattice strain (c) vs. reciprocal gmin size for w (17).

Themajoropenquestion related to the above is: what is the deformation mechanism in

materials with nc grain size? In this regard, very interesting in situ transmission electron
 SYNTHESIS
OF NANOSTRUCIURED
MATERIALS
BYMECHANICAL
MILLING 19

microscopy (TEM) deformation studies have been carried out on UC thin film Au or Ag
specimens (39,40,41). Samples with 100 mu grain sixes exhibited the usual deformation
mechanisms including extensive dislocation activity, ligament formation, and ductile fracture
within a given grain. However, 10 nm grain size samples did not exhibit dislocations within
the grains. Instead, evidence was seen for grain boundary rotation and sliding (which can result
in 3096 strains at the crack tip) as well as opening of small pores at grain boundary triple
junctions. Such studies and their analysis should help us better understaud the formation of
nanocrystalline structures by mechanical attrition.

NANOCRYSTALLINE MICROSTRUCTURES BY MECHANICAL

ALLOYING OF ELEMENT AL POWDERS

Most of the studies of the microstructural evolution of a nc structure during ball milling
have been carried out on single phase mate&Is such as elements or compounds. However,
nanocrystalline grains are also observed during the mechanical alloying (MA) of multi-
component powders. Klassen et al (42) followed the phase formation and microstructural
development during MA of Ti and Al powder blends of overall composition Ti+lz. TEM
revealed mmocrystalli.ne grains of the partiahy otdered L12 phase with a crystalhte size of 10-30
nm in the alloy layers at the interface between Ti and Al lamellae at very early stages of the
milling process. The alloy phase which develops between the pure powder components
consists of nc grains presumably because of the multiple nucleation events and the slow
growth which occur at relatively low temperahues (lOO-200°C above ambient) during milling.
Trudeau et al (38) prepared nc FeTi by both MA of elemental Fe and Ti powders and MM of
FeTi compound powders. The grain size of MM FeTi steadily decreases with milling time
while that for MA FeJTi fust increases and then decreases to values essentially identical to
those for the MM samples.

MECHANICAL CRYSTALLIZATION OF AMORPHOUS ALLOYS

Crystallixation has been observed by ball milling certain amorphous materials and
nanocrystalline phases often result. In some cases, the crystallization is either clearly due to
impurity pick-up changing the composition and therefore crysWixation tempeaature Tx of the
alloy (43) or to large local temperature increases to above Tx in energetic mills (eg. 44).
However, there are other examples, first cited by Trudeau et al (45). which are not simply
explained by the above two effects. Trudeau and co-workers (45,46,47) and Huang et al (48)
have observed crystallization induced by ball milling in several Fe-base amorphous alloys in
this regard. Explanations offered for this crys&lhzation include: enhancement of diffusion by
the milling deformation which allows decomposition of the metastable amorphous phase (49);
local temperature rises associated with shear bands due to bending as well as wear processes
(48). This phenomenon of crystallization by ball milling in certain amorphous alloys (it is
composition dependent) is still not well understood, but may be related to the observation of
nanocrystal formation within shear bands of melt spun Al-based amorphous alloys induced by
bending the ribbon samples (50). It was suggested this nanocrystallixation is due to local
20 CC KOCH

atomic rearrangements within the shear bands which exhibit enormous plastic strain. Such
atom rearrangement depends on the topological order and the chemical bonding of the
amorphous alloy.

PROBLEMS: CONTAMINATION AND POWDER CONSOLIDATION

A serious problem with the milling of fine powders is the potential for significant
contamination from the milling media (balls and vial) or atmosphere. If steel balls and
containers are used, iron contamination can be a problem. It is most serious for the highly
energetic mills, eg. the Spex shaker mill, and depends on the mechanical behavior of the
powder being milled as well as its chemical affinity for the milling media. For example,
milling Ni to attain the minimum nc graht size in a Spex mill resulted in Fe contamination of
13 at.%, while the Fe contamination in nc Cu similarly milled is only 5 1 at.% (37). Lower
energy mills result is much less, often negligible, Fe contamination. Other milling media,
such as tungsten carbide or ceramics, can be used but contamination Tom such media is also
possible. Surfactants (process control agents) may also be used to minimize contamination.
Mishurda et al (51) studied the effect of several process control agents on Fe and interstitial
element contamhnttion during ball milling of prealloyed Ti&lssWz powders. Boric acid and
borox were quite effective in reducing Fe contamination (from 1.04 wt.96 to 0.44 wt.% and
0.26 wt.%, respectively). Using a “seasoned” vial - i.e. media coated with the alloy powder
- resulted in very low values for Fe contamination of 0.06 wt.% (with boric acid) and 0.09
wt.% (with borax). Interstitial contamination can be controlled by milling and subsequent
powder handling in a pure inert gas atmosphere with care taken that the milling vial is leak-fme
during processing. Goodwin and Ward-Close (52) have recently reported on a new production
scale mill which cau process reactive materials such as Ti-Al alloys with essentially no
metallic or interstitial element contamination. Such devices should generate increased interest
in the use of mechanical attrition for synthesis of nc materials.
Powder consolidation to theoretical density of nc materials prepared by mechanical
attrition without signikant coarsening is necessary for many properties eg. mechanical
behavior. There is not room in this paper to adequately review the increasing effort in this
area. However, a number of successes have been documented by both conventional, eg. hot
isostatic pressing (HIP) (53) and dynamic, eg., explosive compaction (54). While theoretical
densities have been achieved in many studies, tensile deformation has revealed that interparticle
debonding can still be a problem (55). Additional work is needed on this important problem.

ACKNOWLRDGRMRNTS

The author wishes to thank Prof. C. Suryanarayana for his unpublished data on the effect
of surfactauts on contamination in T&Al alloys. Thanks to T. Malow and T. D. Shen for
useful discussions. The author’s work described in this paper was supported by the U. S.
National Science Foundation under grant No. DMR-9508797.
 SYNTHESIS
OF NANOSTRUCTURED
MATERIALS
BYMECHANGAL
MILLING 21

1. R. Birringer, H. Gleiter, H. P. Klein, and P. Marquardt, Phys. Lett. A. lQ& 356 (1984).
2. A. Inoue, Mater. Sci. Engg. Bu9/AllU, 57 (1994).
3. R. 0. Hughes, S. D. Smith, C. S. Pande, H. R. Johnson, and R. W. Armstrong, Scripta
Metall. 2a 93 (1986).
4. Z. G. Li and D. J. Smith, Appl. Phys. Lett. fi 919 (1989).
5. K. Lu, W. D. Wei, and J. T. Wang, Ser. Met&I. Mater. a2319 (1995).
6. B. H. Kear and P. R Strutt, Nano-Structured Mater. 6,227 (1995).
7. C. C. Koch, Nano-Structured Mater. 2,109 (1993).
8. H. J. Fe&t in “Nanophase Materials”, G. C. Hadjipanayis and R W. Siegel eds., KIuwer
Acad. Pub. 125 (1994).
9. H. J. Fecht, Nano-Structured Materials, d 33 (1995).
10. E. Hellstem, H. J. Fecht, C. Garland, and W. L. Johnson, J. Appl. Phys. 65, 305
(1989).
11. E. HeIIstem. H. J. Fe&t, C. Garland, and W. L. Johnson, Mat. Res. Sot. Symp. Proc.
132,137 (1989).
12. J. Eckert, J. C. Holzer, C. E. Krill, III, and W. L. Johnson, J. Mater. Res. Z 1751
(1992).
13. E. Gaffet and M. Hamtelin, J. Less. Comm. Met. m 201(1990).
14. T. D. Shen, C. C. Koch, T. L. McCormick, R. J. Nemanich, J. Y. Huang, and J. G.
Huang. J. Mater. Res. u1 139 (1995).
15. T. D. Shen, W. Q. Ge, K. Y. Wang, M. X. Quan, J. T. Wang, W. D. Wei, and C. C.
Koch, Nano-Structured Mater. 2,393 (19%).
16. G. W. Nieman and J. R. Weertman, in “Morris E. Fine Symposium”, ed. P. K. Liaw et
al, TMS (1991) pp. 243-250.
17. D. Oleszak and P. H. Shingu, J. Appl. Phys. 23 2975 (1996).
18. G. K. Williamson and W. H. Hall, Acta Metall. 1,22 (1953).
19. K. Yamada and C. C. Koch, J. Mater. Res. S, 1317 (1993).
20. G. Martin and E. Gaffet, J. Phys. France (CoIkques) ti Cl-71 (1990).
21. N. Burgio, A. Iasonna, M. Magini, S. Martelli, F. Padella, Nuovo Cimento, fi, 459
(1991).
22. P. Pochet, L. Chaffron. and G. Martin, Mater. Sci. Forum, 179_181.91(1995).
23. V. A. Pavlov, 0. V. Antonova, A. P. Adakhovskiy, A. A. Kuranov, V. M. AIya’ev, and
A. I. Dergagin, Phys. Met. MetaIl. a 158 (1984).
24. V. A. Pavlov, Phys. Met. MetaII. s 1 (1985).
25. M. Hatherly and A. S. M&t, Scripta Met. & 449 (1984).
26. D. A. Rigney, L. H. Chen, M. G. S. Naylor, and A. R. Rosenfield, Wear, m 195
(1984).
27. D. Turley, J. Inst. Metals, B, 271 (1971).
28. R. Z. Valiev, A. V. Komikov, and R. R. Mulyukov, Mater. Sci. Eng. A16& 141
(1993).
29. R. Z. Valiev. Nano-Structured Mater. 6.73 (1995).
30. J. Languillaume, F. Chmelik, G. Kapelski, F. Bordeaux, A. A. Nazarov, G. Canova, C.
Esling, R. Z. Valiev, and B. Bandelef Acta Metall. Mater. 4L 2953 (1993).
22 CC Koai

31. A. Belyakov, and R. Kai~yshev, Nano-Structured Mater. 6 893 (1995).

32. R. W. Cahn, “Recovery and Recrystallization”, Chapt. 25 in Physical Metallurgy, 3rd
edition, ed. R. W. Cahn and P. Haasen, Elsevier Science Pub. (1983) p. 1613.
33. T. D. Shen and C. C. Koch, MateriaIs Science Forum, I=,17 (1995).
34. D. K. Pathak, Ph.D. Dissertation, North Can&a State University, 1995.
35. T. G. Nieh and J. Wadsworth, Scripta Metall. Mater. Z. 955 (1991).
36. J. Eckert, J. C. Holzer, C. E. KriIl, III, and W. L. Johnson, J. Mater. Res. l, 1980
(1992).
37. T. D. Shen and C. C. Koch, Acta Mater. 44 753 (1996).
38. M. L. Trudeau, R. Schulz, L. Zaluski, S. Hosatte, D. H. Ryan, C. B. Doner, P. Tessier,
J. D. Strom-Olsen, and A. Van Neste, Materials Science Forum, &?-9Q 537 (1992).
39. W. W. MiIIigan, S. A. Hackney, M. Ke, and E. C. Aifantis, Nano-Structured Mater. 2,
269 (1993).
40. M. Ke, S. A. Hackney, W. W. MiIIigan, and E. C. Aifantis, Nano-Structured Mater. 3,
689 (1995).
41. S. A. Hackney, M. Ke, W. W. MilIigan, and E. C. Aifantis, in ‘Processing and
Properties of NanocrystaIIine Materials”, ed. C. Suryanarayana., J. Singh, and F. H.
Froes, TMS, 1996, pp. 421-426.
42. T. Klasxn, M. Oehring, and R. Bormann, J. Mater. Res. 2 47 (1994).
43. U. Mizutani and C. H. Lee, J. Mat. Sci. Z., 399 (1990).
44. J. Eckert, L. Schultz, and E. Hellstem, J. Appl. Phys. 64. 3224 (1988).
45. M. L. Trudeau, R. Schulz, D. Dussault, and A. Van Neste, Phys. Rev. Lett. f&99
(1990).
46. M. L. Trudeau, J. Y. Huot, R. Schulz, D. Dussault, A. Van Neste, and G. L’Esp&ance,
Phys. Rev. B & 4626 (1992).
47. M. L. Trudeau, L. Dignard-Bailey. R. Schulz, D. Dussauh, and A. Van Neste, Nano-
Struct. Mater. 2,361 (1993).
48. B. Huang, R. J. Perez, P. J. Crawford, A. A. Sharif, S. R. Nutt, and E. J. Lavernia,
Nano-Struct. Mater. j, 545 (1995).
49. M. L. Trudeau, Appl. whys. Lett. 6pI 3661 (1994).
50. H. Chen, Y. He, G. J. ShifIet, and S. J. Peon, Nature, %!, 541 (1994).
5:. J. Mishurda, C. Suryanarayaua, aud F. H. Froes, University of Idaho, 1994, unpublished
results.
52. P. Goodwin and M. Ward-Close, “Mechanical Alloying of Ti-Based Materials”,
ISMANAM-%, Rome, Italy, May 20-24.1996, to be published in the proc=d&.
53. C. Suryanarayana, G. E. Korth, and F. H. Froes, iu “Processing and Properties of
Nanocrystalline Materials”, ed. C. Suryanarayana, J. Singh. and F. H. Froes, TMS,
Warn&ale. Penn. (1996) pp. 291-302.
54. G. E. Korth and R. L. Williamson, MetalI. Mater. Trans. A, 26& 2571 (1995).
55. M. Jain and T. Christman, Acta Metah. Mater. 42.1901(1994).
56. S. Enzo, M. SampoIi, G. Cocco, L. Schiffini, and L. Battezzati, Phil. Mag. B Se 169
(1989).