![Characteristics plot of surface roughness vs. WF for aluminum as per reference [17].](https://www.researchgate.net/profile/Gustavo-Molina-3/publication/229036890/figure/fig3/AS:300851173314566@1448739917219/Characteristics-plot-of-surface-roughness-vs-WF-for-aluminum-as-per-reference-17_Q320.jpg)
Content uploaded by Gustavo J Molina
Author content
All content in this area was uploaded by Gustavo J Molina
Content may be subject to copyright.
THE ANNALS OF UNIVERSITY “DUNĂREA DE JOS “ OF GALAŢI
FASCICLE VIII, 2006 (XII), ISSN 1221-4590
TRIBOLOGY
22
A PRELIMINARY STUDY ON THE DYNAMICS OF TRIBOEMISSION,
WORK FUNCTION AND SURFACE CHARGING OF CERAMICS
Gustavo J. MOLINA
Department of Mechanical and Electrical Engineering Technology, Virginia Polytechnic Institute and State
University (Virginia Tech), GA 30460-8045, USA
gmolina@georgiasouthern.edu
ABSTRACT
This paper discusses the tribological dynamics of electron triboemission and
after-contact emission during scratching of ceramics. The origin of such
triboemission is still unclear, but the author and colleagues consistently observed
two features that can explain the emission and its observed time evolution: that a
large fraction of electron is emitted for energies lower that the electron work
function of the bulk material, and that there is evidence of surface charge that
relaxes over a long period. These experimental findings and their analysis are
reviewed. Literature data on the reduction of electron work function during dry
sliding is correlated to the experimental observations. Surface charging is discussed
as a factor in triboemission: trapping and release of charges from the activation
and cancellation of surface lattice defects are connected to the relaxations of
mechanical energy and of polarization during wear. These phenomena can help
explaining the observed time evolution of electron triboemission from ceramics.
This preliminary study indicates that understanding of the electron triboemission
origin and evolution requires a mechanism that integrates the evolution of defects
during sliding and its effects on the dynamics of electron work function change and
of energy storage and release by charge trapping.
KEYWORDS: Triboemission, electron work function, tribocharging, alumina.
1. INTRODUCTION
The emission of electrons, ions, neutral parti-
cles, photons, and acoustic emission under conditions
of tribological contact and damage is called tribo-
emission [1]. Figure 1 shows a conceptual view of
triboemission.
Triboemitted electrons, which make the majori-
ty of charge-triboparticle emission from insulators
and semiconductors, are known to be important fac-
tors in, for instance, the initiation and control of che-
mical reactions during lubrication processes [2].
These triboemitted electrons are detected from the
scratching of ceramics and semiconductors in
vacuum, and they are characteristically of low-energy
(e.g., 0 to about 10eV) [3-5]. Because of the very low
load and velocities employed in experiments,
thermionic emission can be ruled out as the origin of
this triboemission. For the tested contact conditions of
Molina et al. [3], metals are not known to produce
significant electron triboemission. Photon tribo-
emission also has been measured, but no correlation
has been explored to the charged-emission [4].
Fig. 1. Conceptual view of triboemission.
Work on charged-particle triboemission was
independently carried out by (a) Nakayama et al. [4,
6] of the Mechanical Engineering Laboratory, Japan,
(b) Dickinson et al. [5, 7] of Washington State
University, and (c) Molina et al [3, 8-11], initially of
Virginia Tech and presently of Georgia Southern
University.
Molina et al. [4, 10] extensively characterized
burst-type low-energy triboelectrons for scratching (at
THE ANNALS OF UNIVERSITY “DUNĂREA DE JOS “ OF GALAŢI
FASCICLE VIII, 2006 (XII), ISSN 1221-4590
TRIBOLOGY
23
constant load and speed) of alumina, sapphire and and
Si3N4, and of the semiconductors Si and Ge.
Triboemission rates and statistical significance of
these experiments are presented by Molina et al. [9,
10]. They detected much lower levels of positively-
charged emission from insulators [4] and reported the
absence of semiconductor triboemission after the
contact ceased, when compared to significant post-
contact triboemission from all tested insulators [10].
2. EXPERIMENTAL
Molina et al. [11] also developed triboelectron
measurements from an alumina-ball sliding on an
alumina-disk. The triboemission instrument
developed by Molina et al. [4] was used for electron
intensity measurements from scratching of alumina
disks in a vacuum of 10-6 Pascal or better. A channel
electron multiplier (CEM) detector in the pulse-
counting mode was operated at 2750V and a +200V
input bias. A grounded-grid was placed between
CEM and wear track. Individual charged-particles are
detected as counts in 10 msec windows.
The contact geometry consisted of rotating
25.4mm-diameter-disks of amorphous alumina
(99.5% isostatically pressed polycrystalline alumina)
scratched by a stationary 0.125-inch diameter
alumina-ball (99.5% alumina, grade 25). For each
measurement an initial 30-second reference the
background-noise was less than 0.1count/sec.
An example of data is presented in figure 2. The
most important feature of the triboemission evolution
in figure 2 is that a considerable delay consistently
occurs until the appearance of large bursts of
triboemission for a ball-on-flat contact, when
compared to scratching by a diamond-cone.
Description and discussion of this type of
experiments have been presented by Molina et al.
[11].
For the contact conditions of Figure 2, samples
of alumina-ball on alumina disk contacts, each
obtained after different number of passes on the same
wear track, were studied by wear-track profilometry
and microscopy [11]. That study showed that from
start to about 14 passes the wear track barely outlined
in isolated patches and no wear was measurable by
the used profilometry: surface contact was isolated
around contact asperities with some wear debris
production by asperity crushing and fracture, and
corresponding electron triboemission was low and
seemingly diminishing. After such initial period (e.g.,
from about 14 passes) the wear-track became fully
defined with surface plastic deformation and
measurable increasing wear by particle detachment,
while large bursts of triboemission are simultaneous.
A decreasing wear rate was observed after this onset
of wear and the large bursts of emission diminished to
a lower level in following passes.
Fig. 2. Negatively-charged triboemission for alumina-ball on alumina disk.
Acquisition window: 10msec. Load: 10N. Speed: 0.48cm/s.
THE ANNALS OF UNIVERSITY “DUNĂREA DE JOS “ OF GALAŢI
FASCICLE VIII, 2006 (XII), ISSN 1221-4590
TRIBOLOGY
24
Following figures 3 to 5 show examples of alu-
mina-disk SEM photographs for the virgin surface,
and for the center of the wear tracks for different
number of passes of an alumina-ball on an alumina-
disk. For these SEM pictures, disk surfaces were
gold-sputtered for 1 minute; employed film was
Polaroid 55.
Fig. 3. SEM photograph of virgin alumina for
experiments as of Figure 2. Magnification: 1,000X.
Fig. 4. SEM photograph of wear track center for
alumina-ball on an alumina-disk in experiments as of
Figure 2. Magnification: 1,000X. Contact period: 20
passes on the same wear track.
The sequence of figures 3 to 5 shows that the
center of the disk wear-track for this ball-on-disk
contact presents increasing plastic deformation
leading to an amorphous state for large number of
passes on the same wear track. Figure 4, for 20 passes
on the same wear track, shows plastically deformed
patches, which are likely corresponding to asperity
crushing and sliding. A large fraction of the generated
wear debris was removed to the edges of the wear
track (not shown in the photographs) or by cleaning
previous to SEM; however, abundant wear debris is
observed.
Figure 5, for 114 passes on the same wear track,
shows an amorphous transfer film which presents
some cracking and would detach by flaking; wear
debris is present but not as abundant as in figure 4.
This wear evolution is consistent with the work of
Ajayi and Ludema [12]; they found that for an
alumina pin sliding on an alumina disk a wear
transition from severe-to-mild regime occurred
because of the transfer film. This film builds up by
including wear debris. It would decrease contact
pressure (because the nominal contact area increases)
and the wear rate should decrease with time, as
observed in this paper experiments; further sliding
would result in surface flaking.
Fig. 5. SEM photograph of wear track center for
alumina-ball on an alumina-disk in experiments as of
Figure 2. Magnification: 1,000X. Contact period: 114
passes on the same wear track
The author believes that description of
triboemission phenomena requires studying of the
emission evolution while considering the correspond-
ding surface states. Electron triboemission origin is
still unclear, but it is known that a large fraction of
emitted electrons is produced for energies lower than
the electron work function (WF) of bulk materials.
For insulators, triboemission also is observed after the
contact ceases. This paper discusses possible relations
between experimentally measured reduction of WF
and the wear states on surfaces, and the possibility
that surface charging can be a factor on triboemission.
3. DISCUSSION
3.1. Electron Work Function Change
During Sliding Contact
Experimental evidence shows that wear
evolution relates to changes of electron work function
(WF). The WF of a metal is the energy required to
THE ANNALS OF UNIVERSITY “DUNĂREA DE JOS “ OF GALAŢI
FASCICLE VIII, 2006 (XII), ISSN 1221-4590
TRIBOLOGY
25
extract an electron from the solid to a point outside
the surface where there are no interactions between
the electron and the solid. The work function is
dependent on the atomic structure of the solid and it
can show local variations. The WF is the difference
between the electrostatic surface potential and the
Fermi energy. For a metal, the Fermi energy is the
highest filled energy level in the conduction band.
The seminal work of Zharin et al. [13] measured
WF during tribological damage by the non-contacting
Kelvin probe (KP). Bhushan et al. [14] used KP
signal to detect wear precursors at ultra-low loads for
Au, Al and Si (but the technique was difficult for
alumina). Li et al. [15] correlated elastic deformation
during sliding to decreasing WF for Cu, Al and steel,
and plasticity to a stable WF value, of 0.3 to 0.4eV
lower than those of unworn surfaces. Voevodin et al
[16] used a KP for the characteri-zation of DLC
nanocomposite coatings during wear.
For metals, the contact potential difference
(CPD) can be readily measured, and it is the
difference in surface potential when the two metals
are electrically connected. The CPD method and a
Kelvin probe can be used to estimate WF, which is
proportional to the reciprocal of CPD, measured with
respect to a reference metal. This method was exten-
sively used by Shpenkov et al. [17] to correlate CPD
and WF to selected metal surface states (e.g., for Al,
Zn, bronze and brass). They found that the surface
roughness (obtained by mild abrasive grinding and
measured by surface Ra) relates to the WF in a
characteristics curve. That curve is sketched in figure
6 for aluminum.
The plot of surface roughness vs. WF in figure 6
shows characteristics minima and maxima when
roughness is reduced from the typical value for virgin
surface at (a) to a polishing state at (e). The micro-
structures observed in the Shpenkov et al.’s
experiment are matched in the following table 1 to the
wear states and triboemission features that were
observed in the Molina et al’s experiment of figure 2.
Five cha-racteristics maxima or minima are discussed
in table 1.
Table1 shows that when surface states (and
pertaining WF values) are matched to corresponding
sliding-wear states of the triboemission experiments,
minima of WF values correspond to large intensity of
electron triboemission. Similarly, maxima of WF
correspond to lower or diminishing emission
intensity. Discussion of Table 1 strongly suggests that
reduction WF by defect creation during wear is an
important factor for the occurrence of electron
triboemission.
Fig. 6. Characteristics plot of surface roughness vs.
WF for aluminum as per reference [17].
Table 1. Evolution features and microstructure of figure 6 as matched to
corresponding sliding-wear states of the triboemission experiments of figure 2.
Evolution
feature in
Fig. 6
Ra
(micro-
meters)
Microstructure wear state for surfaces
of figure 6
Corresponding wear state and
triboemission for experiment of
Figure 2
(a) >2 Virgin surface: Surface defects but mainly
inactive centers for relatively high WF.
No contact or triboemission
(a) to (b) 2 to 0.29 Formation of new defects (e.g., dislocation
piling, etc.), new surface and active
centers with reduced WF.
Elastic-plastic deformation and brea-
king for asperities tops. Non-measu-
rable wear. Low level triboemission.
(b) to (c) 0.29 to
0.1
Submicrocrack formation in near surface
that reduces the number of active centers
and increases WF.
Submicrocrack and plastic flow
develop. Low level diminishing
triboemission.
(c) to (d) 0.1 to
0.05
Failure of surface structures and grain
refinement. Intensive local increase of
dislocation density reduces WF.
Intensive particle detachment,
measurable increasing wear. Large
bursts of triboemission.
(d) to (e) <0.05 Fine abrasion and polishing brings surface
to amorphous state, leading to maximum
WF.
Wear rate reduces and stays stable.
Diminishing burst of triboemission in
consecutive passes.
THE ANNALS OF UNIVERSITY “DUNĂREA DE JOS “ OF GALAŢI
FASCICLE VIII, 2006 (XII), ISSN 1221-4590
TRIBOLOGY
26
There are, however, important features of the
electron triboemission phenomena that are not
explained by WF change alone. The significant post-
contact triboemission from insulators (i.e., after the
contact ceases) suggests that for triboemission to
occur a factor is involved that relaxes over a long
period of time. The possibility is following discussed
that insulator surface charge can be such factor.
3.2. Surface Charging During Sliding
There is extensive data on surface (i.e.,
electrostatic) charge measured during and after wear
of insulators: Molina et al [18] obtained evidence of
surface and debris charging from diamond-on-disk
triboemission tests, which were carried out without
the customary grounded-grid between the CEM
charged-particle detector and the wear track. In those
experiments, significant wear debris was found in the
CEM-input as it was picked-up by this detector
cathode-voltage; that charged-debris also showed as
augmented emission rates when compared to usual
grounded-grid measurements [18].
Insulator surface-charging that is observed at the
macroscopical scale (e.g., when measured in a
continuum as insulator permittivity) can be
understood as the aggregate of microscopic surface
states that are isolated from the bulk and originate
from surface changes (i.e., from changes of
polarizability) [19]. A distinct characteristic of
insulators is that charges can be trapped in sites of the
lattice because of the existence or creation of lattice
defects (e.g., vacancies, interstitials, dislocations,
grain boundaries, etc). Mechanical energy is stored as
potential energy around the trapped charges by
displacements of the lattice from their equilibrium
sites [20]. The potential energy of these traps can be
estimated in between a few meV to 5 eV. Because of
the electrostatics nature of insulators, energy release
and storage can result from electrostatic interactions.
In the case of sliding contact the input
mechanical energy can be stored as potential energy
upon the creation of a new defect, to be released
when such defect is canceled. The dynamics of
insulator sliding includes the trapping of charge
(called polarization) in the presence of defects or
upon the creation of them by surface modification,
followed by a detrapping of charge (by relaxation of
the polarization energy) when defects are canceled
[21].
Detrapping of charges has a twofold effect:
charges are released and the lattice must reach a local
equilibrium while releasing energy. Therefore, energy
is output and it may cause material breakdown (and
phonon emission) while detrapped charges are
simultaneously produced for the low energy levels of
the insulators traps. This mechanism is consistent
with the findings of Molina et al. [11], who
experimentally showed that the energy relaxation at
the onset of wear correlates to large bursts of low-
energy electron triboemission.
There is abundant experimental evidence that
energy storage and relaxation and the respective
polarization and depolarization are factors in insulator
mechanical behavior. Medevielle et al. [22] showed
that the relaxation of mechanical energy that occurs
from microfracture and wear is simultaneous to the
relaxation of the polarization energy. Vallayer et al.
[23] hypothesized that insulator polarization and
depolarization could allow a dynamics of surface
linear defects to explain the plastic behavior of brittle
ceramics on the surface of wear tracks.
Fast relaxation of mechanical energy that occurs
during tensile or bending fracture of brittle materials
produces low-energy electron emission. The
phenomenon is known as fractoemission. In a
relevant research work Dickinson et al. [24]
demonstrated a relationship between crack initiation
and propagation, and the simultaneous fracto-
emission. They found for tensile-strained aluminum
in atmosphere that the onset of plastic deformation
and of surface-oxide cracking, which was detected by
acoustic emission, was simul-taneous with a series of
bursts of electrons. Electron fractoemission began and
was most intense during fracture. Such emission
slowly decayed after surface cracking if no added
strain was applied, indicating that the fractured
surfaces remained in an activated state. In that, an
exponential decay was observed for the electron
emission, with a relaxation mechanism slower than
that of most electron emission processes [25].
This study discussed two factors that seem to be
of importance for triboemission under vacuum. Study
of triboemission in atmospheres (e.g., in other than
vacuum) is experimentally more complex, and other
factors can have influence, namely gas desorption
from microfracture crack tip, gas adsorption on
freshly exposed metal surfaces, species ionization and
surface chemical reaction.
4. CONCLUSIONS
This paper discussed the possibility that two
factors, the reduction of WF and the surface charging,
can explain the energy levels and the observed
evolution of the electron triboemission outputs from
insulators.
Comparison of surface states from literature data
and the corresponding sliding-wear states of the
author’s triboemission experiments indicates that
minima of WF values correspond to large intensity of
electron triboemission, while maxima of WF
correspond to lower or diminishing emission
intensity.
Insulators can store mechanical energy by
trapping of charges in lattice defects. Insulator
THE ANNALS OF UNIVERSITY “DUNĂREA DE JOS “ OF GALAŢI
FASCICLE VIII, 2006 (XII), ISSN 1221-4590
TRIBOLOGY
27
mechanism of energy storage and release by,
respectively, charge trapping and detrapping can be
driving factors in the dynamics of electron triboemi-
ssion. In the case of sliding wear of ceramics, the
input mechanical energy can be stored as potential
energy by trapping upon the creation of new defects,
and both energy and charge are released when such
defect is canceled. Surface charging, which relaxes
over a longer period of time, also can explain the
significant post-contact triboemission from insulators
after the contact ceases.
The two discussed factors for the occurrence of
triboemission, the reduction of WF and the surface
charging, are consistent with different features of the
triboemission measurements, and this study suggests
that both are complementary needed for triboemission
from insulators. It also indicates that understanding of
the electron triboemission phenomena requires
experi-menttal work and modeling that integrate the
evolution of defects during sliding and their effects on
the dyna-mics of electron work function and of
energy storage and release by charge trapping. The
author plans to develop simultaneous measurements
of electron tribo-emission and electron work function
for these required in-situ real-time characterizations
of surface changes.
ACKNOWLEDGEMENTS
The author gratefully acknowledges a
Faculty Research Committee grant from Georgia
Southern University and the generous encouragement
from Dr. M.J.Furey and Dr. A. Ritter (Virginia Tech)
and from Dr. C.Kajdas (Central Petroleum Institute,
Warsaw, Poland).
REFERENCES
1. Kajdas C., Furey, M.J., Ritter A.L., Molina G.J., 2002,
Triboemission as a basic part of the boundary friction regime: a
review, Lubrication Science, 14, pp. 223-254.
2. Furey, M.J., Kajdas C., Kempinski R, Tripathy B., 1997,
Action Mechanism of Selected Vinyl Monomers Under Boundary
Lubrication of an Alumina-on-Alumina System, Lubrication
Science, 10, pp. 3-25.
3. Molina G.J., Furey M.J., Ritter A.L., Kajdas, C., 2001,
Triboemission from Alumina, Single Crystal Sapphire, and
Aluminum, Wear, 249, 3-4, pp. 214-219.
4. Nakayama K., Fujimoto T., 2004, The energy of electrons
emitted from wearing solid surfaces, Tribology Letters, 17, 1, pp.
75-81.
5. Kim M., Langford S.C., Dickinson J.T., 1995, Electron and
photon emission accompanying the abrasion of MgO with diamond,
Tribology Letters, 1, pp. 147-157.
6. Nakayama K., Hashimoto H., Susuki T., 1992, Triboemission
of Charged Particles and Photons from Solid Surfaces During
Frictional Damage, J. of Physics D, Vol.25, pp. 303-308.
7. Dickinson J.T., Scudiero L., Yasuda K., Kim K. W., Langford
S.C., 1997, Dynamic tribological probes: particle emission and
transient electrical measurements, Tribology Letters, 3, pp. 53-67.
8. Molina G.J., Furey, M.J., Ritter, A.L., Kajdas, C., 2003,
Frequency analysis and modeling of charged-particle triboemission
from ceramics, Wear, Vol.255, 1-6, pp. 686-694.
9. Molina, G.J., Furey, M.J., Ritter, A.L. Kajdas, C., 2003, The
Role of Triboemission in Ceramic Tribology: A Review,
Tribologia-Finnish J. of Tribology, Vol.21, 4, pp. 3-11.
10. Molina, G.J., Mazilu, D.A, Furey, M.J., Kajdas, C., On
triboemission from the sliding contact of Si and Ge, The Annals of
University “Dunarea De Jos“ of Galaţi, Romania, Fascicle VIII,
Tribology, 2004, pp. 22-29.
11. Molina G.J., Furey M.J., Kajdas C., Steika N., On the
Mechanism of Electron Triboemission from Ceramics, The Annals
of University “Dunarea de Jos” of Galati, Romania, Fascicle VIII,
Tribology, 2005, pp. 17-23.
12. Ajayi, O.O., Ludema, K.C., 1990, Mechanism of transfer film
formation during repeat pass sliding of ceramic materials, Wear,
140, pp. 191–206.
13. Zanoria, E.S., Hamall, K., Daniluk S., Zharin, A., 1997, The
non-vibrating Kelvin probe and its application for monitoring
surface wear, J. of Testing and Evaluation, 25, 2, pp. 233-238.
14. Bhushan B., Goldade A.V., 2000, Kelvin probe microscopy
measurements at surface potential change under wear at low loads,
Wear, 244, 1, pp. 104-117.
15. Li W., Li D.Y., 2000, Exploring the application of the Kelvin
method in studying the history prior to wear and the onset of wear,
Wear, 253, 7, pp. 746-751.
16. Wu J.-H., Sanghavi M., Sanders J.H., Voevodin A.A.,
Zabinski J.S., Rigney D.A., 2001, Sliding behavior of
multifunctional composite coatings based on diamond-like carbon,
Wear, 255, pp. 859-868.
17. Shpenkov G.P., Friction Surface Phenomena, Elsevier Science
Publishers B.V., Tribology Series 29, Amsterdam, Holland, 1995,
p.p. 35-51
18. Molina G.J., 2000, Triboemission of Charged Particles from
Ceramics: Integrated Charge and Energy Distribution
Characterizations, PhD dissertation, Dept. of Mechanical
Engineering, Virginia Tech, Blacksburg, VA, USA.
19. Saint Jean M., Hudlet S., Guthmann C. Berger J., 1999,
Local triboelectricity on oxide surfaces, The European Physical
Journal B, 12, pp. 471-477.
20. Vallayer, B, Bigarre, J., Berrou, A., Fayeulle, S., Treheux,
D., Le Gressus, C., Blaise, G., 1994, Friction of dielectric
materials: how is energy dissipated?, Dissipative Processes in
Tribology, Editor: D. Dowson, Elsevier Science Publishers B.V.,
Amsterdam, Holland, pp.185-192.
21. Fayeulle S., Berroug H., Hamzaoui B., Treheux,D., 1993,
Role of dielectric properties in the tribological behavior of
insulators, Wear, 162-164, pp. 906-912.
22. Medevielle A., Thevenot F., Treheux D., 1997, Wear
resistance of zirconias. Dielectric approach, Wear, 213, pp. 13-20.
23. Vallayer J., Paulhe O., Juve D. Vernet J.M., Treheux D.,
2000, Endommagement dans les materiaux isolants par effet des
charges electriques, Materiaux et techniques, 7-8, pp. 15-19.
24. Dickinson J.T., Donaldson E., Snyder D.B., 1980, Acoustic
emission and electron emission during deformation of anodized
aluminum, J. Vacuum Sci. Technol. 17 (1), pp. 429–432.
25. Dickinson J.T., Jensen L.C., Jahan-Latibari A., 1983,
Fracto-Emission: The Role of Charge Separation, J. Vacuum
Science Technology, A2(2), pp. 1112-1116..
