Materials Sciences and Applicatio ns, 2011, 2, 1452-1464
doi:10.4236/msa.2011.210196 Published Online October 2011 (http://www.SciRP.org/journal/msa)
Determination of Anodized Aluminum Material
Characteristics by Means of Nanoindentation
Measurements
Maria Datcheva1, Sabina Cherneva1, Maria Stoycheva2, Roumen Iankov1*, Dimitar Stoychev3
1Institute of Mechanics, Bulgarian Academy of Sciences, Sofia, Bulgaria; 2Institute of Electrochemistry and Energy Systems, Bul-
garian Academy of Sciences, Sofia, Bulgaria; 3Institute of Physical Chemistry, Bulgarian Academy of Sciences, Sofia, Bulgaria.
Received June 27th, 2011; revised July 26th, 2011; accepted August 11th, 2011.
ABSTRACT
An aluminium AD-3 has been anodized under four different conditions, namely at low temperature (−5˚C), room tem-
perature (25˚C), with and without sealing the anodized coating in boiling distilled water. The solution used for forma-
tion of alumina layer in all cases was an electrolyte containing 180 g/l sulphuric acid at a constant forming voltage
(voltastatic anodizing). In order to assess the mechanical properties of the obtained anodic alumina layers a series of
nanoindentation tests was performed employing different indentation procedures. The two mechanical characteristics of
the alumina films, the indentation hardness (HIT) and the indentation modulus (EIT), were determined by means of the
instrumented indentation and the Oliver & Pharr approximation method. All measurements were done on Agilent G200
Nanoindenter fitted with a diamond Berkovich type tip. Time dependent effects were investigated by tests with different
peak hold time and different loading rate. The change of the mechanical properties with indentation depth was also
examined. The effect of the working temperature during the growth of the alumina layers and the influence of the pore
sealing on the mechanical properties are evaluated via comparison of the average load-displacement curves. The role
of the temperature of the electrolyte and the sealing process during the formation of the alumina films, with respect to
possible changes of their chemical composition and structure, are discussed in order to explain the observed differences
in the measured load-displacement curves and the determined HIT and EIT.
Keywords: Thin Films, Alumina, Mechanical Properties, Nanoindentation
1. Introduction
Alumina is the most wide used oxide ceramic material.
Basic applications of alumina are for/as a protective and
wear-resistant films, filler for plastics, sunscreens, carrier
layers in converters for gas purification, CD/DVD pol-
ishing, etc. Alumina is used in dentistry as alternative to
the sodium bicarbonate for patients that have high blood
pressure, as well as in medicine for hip replacement. The
technology utilizing aluminum oxide detector material
for radiation dose measurement is at the core of many
dosimeter systems and services. Other applications of
alumina coatings are for protection against corrosion, in
optoelectronics and etc. The basic characteristics of alu-
mina, which are important for these applications, are the
high compression and electrical strength, high hardness,
resistance to abrasion and to chemical attacks by a wide
range of chemicals, high thermal conductivity, resistance
to thermal shocks, high degree of refractoriness, etc.
Because of the wide field of application of alumina
and because of the fact, that usually the mechanical
properties of the thin films are very different from the
mechanical properties of the bulk materials, we selected
anodic alumina films as a subject of our research on as-
sessing the mechanical properties by means of instru-
mented nanoindentation.
The alumina layers are also an interesting model sys-
tem for investigation of physical characteristics and me-
chanical properties because the anodically formed Al2O3
layers, depending of the temperature of formation, are
characterized with quite different thickness, structure,
porosity, micro-hardness and wear resistance. In the
same time their chemical composition practically re-
mains unchanged.

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements1453
Mechanical properties of pure aluminium are well
known [1-3], but there are very few data in literature
about mechanical properties of anodic alumina films,
[4-7] and this was an additional motivation for perform-
ing nanoindentation tests in order to determine the alu-
mina mechanical properties. Since Oliver and Pharr
promoted in 1992 the method for determining mechani-
cal properties of materials by instrumented indentation
techniques [8], this method has been widely adopted for
characterization of the mechanical behaviour of materials
at small scales. Its attractiveness stems largely from the
fact that mechanical properties can be determined di-
rectly from indentation load and displacement measure-
ments without the need to image the indentation impress-
sion. With high-resolution testing equipment, this facili-
tates the determination of properties at the micrometer
and nanometre scales [9-11]. For this reason the method
has become a primary technique for determining the
mechanical properties of thin films without removing the
film from the substrate and as well as for capturing small
structural features [12-29]. Nanoindentation technique
nowadays is applied for characterisation of thin films
prepared from metals, polymers, rubber-like materials
[30] and soft materials.
The aim of the presented here work is to assess the ef-
fect of the temperature of the working electrolyte during
the anodic formation of porous alumina and the influence
of the pore sealing on the mechanical properties of the
alumina layers. In the present work different indentation
programs are applied in order to determine the indenta-
tion hardness (HIT) and the indentation modulus (EIT) of
the alumina layers.
2. Theoretical Part
Indentation experiments had been traditionally used to
measure hardness of materials. The method of Oliver and
Pharr (1992) is used to determine the indentation hard-
ness (HIT) and indentation modulus (EIT) of materials
from indentation load-displacement data obtained during
one cycle of loading and unloading. This technique in-
volves a number of simplifying assumptions: 1) the
specimen is an infinite deformable half-space; 2) the in-
denter has an ideal geometry; 3) the material is liner-
elastic; 4) pile-up is negligible, and 5) there are no inter-
action surface forces during contact such as adhesion or
friction forces [31-35].
A schematic representation of a typical data set ob-
tained with a Berkovich indenter is presented in Figure 1,
where the parameter P designates the load and h—the
indentation depth relative to the initial undeformed sam-
ple surface.
There are three important quantities that can be obtain-
ed from the P-h curves: 1) the maximum load Pma x ; 2)
Figure 1. Schematic illustration of indentation load–dis-
placement data showing important measured parameters
[36].
the maximum displacement hmax and 3) the elastic un-
loading stiffness. The unloading stiffness or the so called
the contact stiffness is defined as the slope S = dP/dh of
the upper portion of the unloading curve during the initial
stages of unloading.
The exact procedure used to determine HIT and EIT is
based on the unloading processes shown schematically in
Figure 2, in which it is assumed that the behaviour of the
Berkovich type indenter can be modelled by a conical
indenter with a half-included angle that gives
the same depth-to-area relationship as the Berkovich in-
denter.
70.3  
Letting


c
A
h be an “area function” that describes
the projected area of the indenter at a distance hc = hmax –
hs. Once the contact area is determined, the indentation
hardness is calculated from the maximum force divided
by the projected area:

max
IT
c
P
H
A
h
. (1)
The indentation modulus can be determined by:


2
2
1
21
π
IT
ci
i
E
Ah
E
S















, (2)
where

is a correction factor, whose value depends on
the indenter geometry (for Berkovich indenter β = 1.03 is
adopted),

is the Poisson’s ratio of the probe, and
i
E

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements 1454
magnification × 250 and × 1000, respectively.

h
max
We decided to realize series of 25 indentations on each
sample probe in order to have better statistics (see Figure
3). There are several pre-existing indentation methods
provided by the Agilent Technologies and for the pur-
poses of our study we chose the following three methods
described in more details below: Method A (fixed maxi-
mum displacement), Method B (fixed maximum load)
and Method C (loading with force control).
3.2.1. Method A
Figure 2. Schematic illustration of the unloading process
showing parameters characterizing the contact geometry
[36].
This method prescribes a single load/unload cycle to a
specified depth. Hardness and modulus are determined
using the stiffness as calculated from the slope of the
load-displacement curve during unloading.
i

are the indenter’s elastic parameters [36].
In the frame of this method the indenter tip begins ap-
proaching the surface from a distance (Surface Approach
Distance) above the surface of approximately 1000 nm.
Because of the high roughness of the samples, used in
this study, we had to increase the Surface Approach Dis-
tance from the default to 5000 nm.
3. Experimental Part
3.1. Deposition of Alumina Films
Four different alumina films of 9.5 μm thickness were
deposited on 2000 μm thick AD-3 aluminium substrate.
The chemical composition of the AD-3 substrate is:
99.67% Al and 0.33% Fe. The deposition process was
performed in anodizing bath of 180 g/l H2SO4 Merck
electrolyte at a constant forming voltage of 20 V and it
was lasting 40 minutes (voltastatic anodizing). The elec-
trolyte’s temperature for samples 27 and 28 was −5˚C,
while for samples 31 and 14 it was 25˚C. Samples 28 and
14 were kept after anodizing 1 hour in a bath of distilled
water at temperature 100˚C aiming this way to seal the
alumina pores. Samples description is given in Table 1.
The approach velocity is determined by Surface Ap-
proach Velocity parameter. When the device determines
3.2. Nanoindentation Experiments
Nanoindentation experiments reported hereafter were
realized by Agilent G200 Nano-indenter. The nano-tester
is fitted with a Berkovich three-sided diamond pyramid
with centerline-to-face angle of 65.3˚ and 20 nm radius at
the tip of the indenter. The minimum load possible to be
applied is 10 mN, and the maximum load is 500 mN.
Displacement recording resolution is 0.01 nm and the
load recording resolution is 50 nN. The device is equip-
ped with an optical microscope with 2 objectives of Figure 3. Residual imprints of sample 29 (×250).
Table 1. Investigated materials.
Sample No. Materials Thickness [µm] Electrolyte type Electrolyte T [˚C]Anodization regime Sealing
27 Al2O3/Al 9.5/1990.5 180 g/l H2SO4 (Merck) −5 20 V for 40 min NO
28 Al2O3/Al 9.5/1990.5 180 g/l H2SO4 (Merck) −5 20 V for 40 min YES
29 Al AD-3 2000 N/A N/A N/A N/A
31 Al2O3/Al 9.5/1990.5 180 g/l H2SO4 (Merck) 25 20 V for 40 min NO
14 Al2O3/Al 9.5/1990.5 180 g/l H2SO4 (Merck) 25 20 V for 40 min YES

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements1455
that it has contacted the test surface, according to the
criteria Surface Approach Sensitivity (Table 2), the in-
denter penetrates the surface at a rate determined by
Strain Rate Target (Table 2). When the surface penetra-
tion reaches the Depth Limit (Table 2), the load on the
indenter is held constant for Peak Hold Time (Table 2).
The load on the indenter is then reduced by an amount
defined by Percent to Unload (Table 2) at a rate equal to
the maximum loading rate. Then the indenter is held in
contact with the sample under constant force for 75 sec-
onds. Finally, the indenter is withdrawn from the sample
completely and the sample is moved into position for the
next test [37].
Input parameters for method A are given in Table 2.
We realized series of nanoindentation experiments with
1500 nm maximum displacement and 1 s peak hold time.
Moreover we realized nanoindentation experiments with
3000 nm maximum displacement at 1 s, 10 s and 20 s
peak hold time.
As a result we obtained the load-displacement curves,
indentation hardness and modulus for each of the inves-
tigated alumina films at two different depths and for
various peak hold time.
3.2.2. Method B (G-Series Load, Displacement and
Time)
This method prescribes a single load-unload cycle. No
properties are calculated from the load-displacement-
time information. We used this method to compare load-
displacement curves of the films at fixed maximum load.
During realization of this method the indenter tip be-
gins approaching the surface from a distance above the
surface of approximately Surface Approach Distance
(Table 2). The velocity is determined by Surface Ap-
proach Velocity parameter (Table 2). When the indenter
contacts the test surface, according to the criteria Surface
Approach Sensitivity (Table 2), the single load/unload
cycle begins. The indenter penetrates the surface at a rate
defined by Maximum Load/Time to Load (Table 2).
Loading is terminated when the Load on Sample reaches
Maximum Load (Table 2). The load on sample is then
held constant for ten seconds. Then the indenter is with-
drawn completely at a rate that is twice as fast as the
loading rate.
Input parameters for method B are given in Table 2.
We realized series of nanoindentation experiments by
Table 2. Input parameters.
Parameter Unit Method A Method B Method C
Percent to Unload % 90 90 90
Surface Approach Velocity nm/s 10 25 10
Depth Limit nm 1500/3000 N/A N/A
Delta X for Finding Surface µm −50 −50 −50
Delta Y for Finding Surface µm −50 −50 −50
Strain Rate Target 1/s 0.05 0.05 N/A
Allowable Drift Rate nm/s 0.05 0.05 0.05
Perform Drift Test Segment - 1 1 N/A
Approach Distance to Store nm 1000 1000 1000
Peak Hold Time s 1/10/20 N/A 20/200
Surface Approach Distance nm 5000 5000 5000
Surface Approach Sensitivity % 40 40 40
Poisson’s Ratio - 0.3 0.3 0.3
Maximum load gf N/A 20 50
Time to load s N/A 30 15
Number of times to load - 1 1 10

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements 1456
method B at fixed maximum load of 20 gf (≈196 mN)
and 10 seconds peak hold time. As a result we obtained
load-displacement curves for each of the investigated
alumina films.
3.2.3. Method C (G-Series Basic Hardness, Modulus,
Tip Cal, Load Control)
This method prescribes a series of load/unload cycles in a
single indentation experiment. Indentation hardness and
modulus are determined using the stiffness as calculated
from the slope of the load-displacement curve during
each unloading cycle. The indenter tip approaches the
surface at a rate of Surface Approach Velocity (Table 2)
starting from a distance above the surface of about Sur-
face Approach Distance (Table 2). When the indenter
senses the surface, according to the criteria Surface Ap-
proach Sensitivity (Table 2), the cyclical loading/un-
loading algorithm begins. For each cycle i, the indenter
penetrates the surface at a rate defined by (Maximum
Load/Time to Load )*(2^i/2^Number of Times to Load ).
Loading for the cycle ends when the Load on Sample
reaches Maximum Load*(2^i/2^Number of Times to
Load). At the peak load for the cycle, the Load on Sam-
ple is held constant for a period equal to Peak Hold Time
(Table 2). Then, the indenter is withdrawn at a rate de-
fined by Load Rate Multiple for Unload * Loading Rate
until the Load on Sample reaches Percent to Unload *
Load Limit (Table 2). This load/unload process is re-
peated, incrementing i for each cycle, until i reaches
Number of Times to Load (Table 2). After the last load/
unload cycle, the Load on Sample is held constant for 75
seconds. The indenter is then withdrawn completely and
the sample is moved into position for the next test.
Input parameters for method C are given in Table 2.
We realized series of nanoindentation experiments by
method C at fixed maximum load of 50 gf and with 20
and 200 seconds peak hold times and 10 cycles.
3.3. SEM and EDS Analysis
Scanning electron microscopy (SEM) investigation was
performed on JEOL JSM 6390 apparatus equipped with
INCA energy-dispersive X-ray spectrometer (EDS). It
has been done in order to better understand the structure
changing of the alumina layers and to better visualize the
imprints and the surrounding area because in some cases
the resolution of the optical device of the nanotester was
not sufficient to recognize the imprint images. EDS ana-
lysis/spectrum of the investigated specimens only gives a
rough indication (in atomic%) since the electron beam
does not have enough high spatial resolution (Ø 1 μm
and few μm depth) to analyze each particle individually.
The SEM pictures were performed in the SEI regime.
3.4. Determination of the Surface Roughness
The roughness of investigated alumina films was meas-
ured by means of Perthometer C3A (“Mahr Perthen”,
Germany), equipped with recorder Perthograph C40
(“Mahr Perthen”, Germany). Test section t
I
(the sec-
tion which pin of the perthometer pass during one meas-
urement) for all measurements was 5 mm. We used ver-
tical magnification 500:1 and horizontal magnification
20:1. As a result we determined the average roughness
a and the mean roughness depth R
z
R of samples 28
and 14. The definitions of and
a
R
z
R read:
0
1d
m
I
a
m
R
I
yx

(3)

12345
1
5
z
R ZZZZZ, (4)
with y(x)-profile ordinates of the roughness profile; m
I
-
measured section length (this part of test section length
t
I
, which we evaluate); i
Z
, (i = 1, ···, 5) is the vertical
distance between the highest peak and the deepest valley
within i-th sampling length of five consecutive single
measured sections. The results are given in Table 3 and
they show that sample 28 has higher roughness compared
to sample 14.
4. Results
As a result of nanoindentation experiments, we obtained
load-displacement curves for each of the alumina sam-
ples and after that by means of Oliver & Pharr method
the indentation hardness HIT and indentation modulus EIT
were calculated by the software available as part of the
Agilent G200 Nanoindenter. The comparison between
load-displacement curves obtained by means of method
A at 1500 nm indentation depth and 1 s peak hold time is
shown in Figure 4 for samples 27 and 28, and in Figure
5 for samples 14 and 31. The outcome of this comparison
is that the treatment of the sample anodized at −5˚C with
boiling water does not much influence its mechanical
properties. On the contrary, the sealing of pores by
means of a bath at 100˚C distilled water for the alumina
film obtained at 25˚C electrolyte’s temperature influ-
ences essentially its mechanical response; e.g. the HIT of
sample 14 is higher than the indentation hardness of
sample 31.
Table 3. Roughness measures of samples 14 and 28.
Roughness measure Sample 14 Sample 28
Ra [µm] 0.32 0.48
Rz [µm] 2.5 3.9

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements1457
0
40
80
120
160
200
02004006008001000 12001400 1600
Displacement into Surface (nm)
Load on Sample (mN)
28(9.5)/Al AD-3
27(9.5)/Al AD-3
Figure 4. Sample average sheets for samples 28 and 27, test
method A1500/1.
0
50
100
150
200
250
0200 400 600 8001000120014001600
Displacement into Surface (nm )
Load on Sample (mN)
31(9.5)/Al AD-3
14(9.5)/Al AD-3
Figure 5. Sample average sheets for samples 31 - 14, test
method A1500/1.
A comparison between HIT and EIT, obtained as a result
of 25 indentations via method А with depth limit 1500
nm (A1500/1) and 3000 nm (A3000/1) and pick hold
time 1 s for samples 31, 14, 27, 28 and 29 is shown re-
spectively in Figures 6 and 7. The numbers given aside
the symbols of each of the experimental series show the
average maximum load in mN for this series (larger val-
ues of the maximum load correspond to the data obtained
via method A3000/1).
The results show that the apparent indentation hard-
ness of the film-substrate system is over 4 times larger
than the hardness of the bulk sample from pure alumi-
num (see Figure 6, sample 29).
It is evident that pore sealing leads to increasing the
HIT and EIT of the alumina, obtained at 25˚C electrolyte’s
temperature and does not influence the characteristics of
the film, obtained at −5˚C electrolyte’s temperature.
Figures 8(a) and (b) show a comparison between the
load-displacement curves of sample 28, for two different
maximum indentation depths, namely 1500 nm and 3000
nm (method A). At larger indentation depths there is a
well pronounced pop-in effect and it may be stated it
occurs at h > 1500 nm, (Figure 8(b)). Figure 8(a) shows
averaged curves and therefore the pop-in effect is smeared
0
2
4
6
8
HIT [GPa]
555
210
401
160
30.2
377
188
426
190
31 14 2927 28
Figure 6. Calculated hardness H, based on test methods
A1500/1 and A3000/1.
50
70
90
110
130
150
E
IT
[GPa]
555
210
401
160
30.2
377
188
426
190
31142927 28
Figure 7. Calculated Young modulus E, based on test meth-
ods A1500/1 and A3000/1.
and manifested in the decrease of the slope of the loading
branch of the load-displacement curve. However there
may be a different reason for such decrease of the slope
of load to sample—displacement into surface curve, e.g.
the influence of the substrate as far as below 1500 nm
depth the penetration exceeds 15% of the film thickness.
One significant problem with the method of Oliver and
Pharr is that it does not consider a pile-up of a material
around the contact impression. When pile-up occurs, the
contact area is underestimated by the method and both
HIT and EIT may be overestimated sometimes up to 50%.
Bolshakov and Pharr proposed a convenient, experimen-
tally determined parameter that can be used to identify
whether pile up is coming into the picture [38]. This pa-
rameter is the ratio of final indentation depth
f
h to the
depth of the indentation at peak load, hmax. When
max 0.7 f
hh it is most possible we have a pile-up of the
material around the imprint. That is why, for each sample,
we calculated maxf hh for all 25 nanoindentations. The
results are shown in Figures 9, 10, 11 and 12. It can be
concluded that for samples 27 and 31 we have predomi-
nantly max 0.7
f
hh  and most probably a pile-up. For
sample 31 the existence of a pile-up was proven by SEM

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements 1458
28(9.5)/Al AD-3
0
100
200
300
400
500
05001000 1500 2000 2500 3000 3500
Displacement Into Surface (nm)
Load on Sample (mN)
h
max
=3000nm
h
max
=1500nm
pop-in
(a)
0
50
100
150
200
250
300
350
400
450
500
05001000 1500 2000 25003000 3500
Displacement Into Surface (nm)
Load On Sample (mN)
pop-in
h
max
=1500nm
h
max
=3000nm
(b)
Figure 8. (a) Comparison between average load-displace-
ment curves, obtained by methods А1500/1 and А3000/1 for
sample 28; (b) an example of single load-displacement
curves, obtained by methods А1500/1 and А3000/1 for sam-
ple 28.
0.6
0.7
0.8
0510 15 20 25
Te st number
h
f
/h
max
pile-up Sample 27
Figure 9. Determined hf/hmax for sample 27; test method
B20/10.
micrograph of a residual imprint as can be seen in Figure
13.
The creep effects at 20 gf maximum load with 10 s
peak hold time (Method B) are shown in Figure 14 and
Figure 15, and average maximum displacements for
each of the samples obtained by means of method B are
given in Figure 16. Sample 31 shows the larger relative
0.6
0.7
0.8
051015 20 25
Te st nu mber
hf/hmax
pile-up Sample 28
Figure 10. Determined hf/hmax for sample 28; test method
B20/10.
0.5
0.6
0.7
0.8
051015 20 25
Test number
h
f
/h
max
pile-up
Sample 31
Figure 11. Determined hf/hmax for sample 31; test method
B20/10.
0.5
0.6
0.7
0.8
0510 15 2025
Test number
h
f
/h
max
pile-up
Sample 14
Figure 12. Determined hf/hmax for sample 14; test method
B20/10.
creep displacement (3.4%), followed by sample 14
(3.1%), sample 27 (3.1%), sample 28 (2.7%) and sample
29 (2.1%), while sample 29 has the larger absolute creep
displacement (83.8 nm), followed by sample 31 (56 nm),
sample 27 (51 nm), sample 14 (46 nm) and sample 28
(43 nm).
The variation of HIT and EIT depending on depth of in-
dentation is given in Figures 17 and 18. These two fig-
ures depict the results obtained by means of method C at

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements1459
Figure 13. Residual imprint with pile up (×20000, sample
31).
40
60
80
average h
creep
[nm]
31 142927 28
Figure 14. Comparison between average hcreep for all sam-
ples (test B20/10).
50 gf maximum load and 20 s peak hold time prior each
of unloading step.
It is seen that sample 14 has higher HIT compared to
the hardness of sample 28 at the same applied maximum
load (Figure 17). The same behaviour is observed for the
EIT but the difference here is moderate (Figure 18). At
the same time samples 14 and 28 have higher indentation
hardness than the substrate (sample 29) and almost the
same indentation modulus. Samples 14 and 28 have been
sealed in boiling water and the only difference in their
formation is the electrolyte’s temperature during the
Figure 15. Comparison between average h and hcreep for all
samples (test B20/10).
1300
2300
3300
4300
average hma x [nm]
31 1429 27 28
Figure 16. Average maximum displacement in test B20/10.
able 4 the two samples have identical chemical compo-
arison of HIT and EIT of sample 28 at differ-
en
ntation method has been used to com-
pa
T
sition, that is why we suppose that the reason for the dif-
ference in HIT may be due to a difference in the micro-
structure. The micro and nano-structute of the two alu-
mina layers have been investigated by means of SEM
image analysis. The SEI clearly shows the amorphous
structure of the alumina layers. The average size of
grains and pores for sample 14 are 20 - 30 nm, while for
sample 28 the average size of grains is 60 - 80 nm, and
for the pores it is 40 - 60 nm. These grain and pore size
values were determined in SEM regime at magnification
100000×.
The comp
t indentation depths and different peak hold time is
shown in Figures 19 and 20. The results shown in these
figures are obtained using Method C with 50 gf maxi-
mum load and two different values of the peak hold time
–20 s and 200 s.
The same inde
re the HIT and EIT of sample 14 at different indentation
depths and different peak hold time. The results are given
in Figures 21 and 22. From the results presented in Fig-
ures 19-22 it is evident that the HIT and EIT values at one
and the same maximum load are higher for the case when
the peak hold time is 20 s. anodization of the AD-3 substrate. However as shown in

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements 1460
ble 4. EDS analysis of the surface of Al2O3 (specimens Nos. 14 and 28).
Element Sample Weight percent [%] Atomic percent [%]
Ta
14 54.49 ± 0.47 67.30
O on line Kα
Al on line Kα
S on line Kα
28 53.64 ± 0.47 66.54
14 40.11 ± 0.44 29.38
28 40.79 ± 0.43 30.01
14 5.40 ± 0.19 3.32
28 5.57 ± 0.19 3.45
0
1
2
3
4
5
6
01000 2000 3000 4000 5000 6000 7000
Displacement Into Surface (nm)
Hardness (GPa)
Sample 14
Sample 28
Sample 29
Figure 17. Comparison between indentation hardness of
samples 14, 28 and 29, obtained with 20 s peak hold time by
method C.
0
50
100
150
200
250
01000 2000 3000 4000 5000 6000 7000
Displacement Into Surface (nm)
Modulus (GPa)
Sample 14
Sample 28
Sample 29
Figure 18. Comparison between indentation modulus of
An explanation of this observation may be the effect of
th
This is the case with the pores that seems to be closed
n of this study was to investigate the
es of anodized AD-3 per se and for
mple 14 (ano-
di
samples 14, 28 and 29, obtained with 20 s peak hold time by
method C.
e creep that seems to be an inherent property of the
alumina-substrate system. The SEM micrographs in Fig-
ures 23(d) and (e) show that the grain size inside the
imprint, at the imprint boundary and outside the imprint
is almost the same. It may be a proof that there is no
grain crushing during the indentation. Figure 23(c)
shows the pore structure of the alumina film (sample 28).
and the volume inside the imprint may become com-
pacted.
5. Conclusions
The primary intentio
mechanical properti
this reason no treatment of the surface was applied. Even
measurements were done on various penetration depths,
the analysis presented here is using the obtained me-
chanical characteristics for depths exceeding 500 nm
because of the high roughness of the alumina surface. On
the other side we tried to minimize the influence of the
substrate on the results and this is the reason for consid-
ering indentation depths up to 1500 nm (up to 15% of the
aluminum oxide layer). Nevertheless the sample rough-
ness may play significant role in our measurements. The
outcome of our observation within these constraints is
that the determined by means of instrumented nanoin-
dentation test indentation hardness of anodized alumi-
num AD-3 varies with anodization conditions and it is
over 4 times higher than the hardness of the AD-3 sam-
ple. The elastic characteristics of anodized AD-3 and the
non-anodized AD-3 are almost the same, they vary in the
same interval of 70 to 130 GPa depending on the loading
regime. Therefore we did not observe significant differ-
ence in the EIT of the different samples and we accept
that the electrochemically produced Al2O3 layers are
having almost the same EIT as the AD-3.
The reactive sealing shows better results against im-
proving the hardness when applied to sa
zed AD-3 at 25˚C) over against its application to the
“hard anodized” sample 28. As far as our investigation of
chemical composition of samples 14 and 28 shows that
they have identical chemical composition, the difference
in HIT modulus is considered to be due to the difference
in the film microstructure. This assumption was proven
by SEM micrographs where we found that sample 14 has
smaller grain size and pore diameter. It may be con-
cluded that for the alumina film formed in electrolyte at

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements1461
0
0.5
1
1.5
2
2.5
3
3.5
4
Sam ple 28, 200s peak hold time
Sample 28, 20s peak hold time
01000 2000 3000 4000 5000 6000 7000
Displacement Into Surface (nm)
Hardness (G P a)
Figure 19. Comparison of HIT of sample 28, obtained with different peak hold time.
0
20
40
60
80
100
120
140
160
01000 2000 3000 4000 5000
Displacement Into Surface (nm )
Modulus (GPa)
Sam ple 28, 200s peak hold time
Sam ple 28, 20s peak hold time
Figure 20. Comparison of EIT of sample 28, obtained with different peak hold time.
Hardness vs Displacement Into Surface
(sample 14, side 1, P=50gf)
0
1
2
3
4
5
6
7
05001000 1500 2000 2500 3000 3500 4000
Displacement Into Surface (nm)
Hardness (GPa)
20s peak hold time
200s peak hold time
Figure 21. Comparison of HIT of sample 14, obtained with different peak hold time.

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements 1462
Modulus vs D isplacement Into Surface
(sample 14, side 1, P=50gf)
0
20
40
60
80
100
120
140
160
180
200
05001000 15002000 25003000 35004000
Displacement Into Surface (nm)
Modulus (GPa)
20 s peak hold time
200s peak ho ld time
Figure 22. Comparison of EIT of sample 14, obtained with different peak hold time.
(a) (b)
(c) (d) (e)
Figure 23. SEM images of the surface of sample 28, near and far from the imprint.
−5˚C the process of pore sealing for 1 hour in boiling
water has no essential impact on the HIT modulus. How-
ever, for the alumina film formed at room temperature
the influence of the pore sealing on the mechanical prop-
erties is noticeable. Most likely, this effect is connected
with different degrees of amorphisation of the Al2O3 lay-
ers depending on the temperature of the electrolyte dur-
ing their anodic formation.
The comparison between HIT and EIT of samples 28
and 14 derived for indentation with 20s and 200 s peak
hold time shows that HIT and EIT of these two samples are
higher for the series with 20s peak hold time, and this is
most probably due to the creep of the alumina-substrate
system.
At larger indentation depth (tests A with maximum
depth of 3000 nm) there is well pronounced pop-in effect
and it firstly occurs at h of approximately 1500 nm. The
analysis of the ratio hf/hmax shows that for some of the
samples it exceeds the required value for the method of
Oliver and Pharr to be applicable. Therefore the obtained
based on the Oliver and Pharr method HIT and EIT for
samples 27 and 31 with hf/hmax > 0.7 should be further
approved against work hardening property as suggested
in [38].

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements1463
In order to further verify the applicability of the Oliver
and Pharr method for determining HIT and EIT it is fore-
seen to perform a simulation of the experimental data via
FE-analysis of the nano-indantation tests.
6. Acknowledgements
Authors gratefully acknowledge the financial support of
the Bulgarian National Science Fund under grant No.
TK01/0185 and of the ESF OP “Human Resources De-
velopment” under the contract BG051PO001/07/3.3-02.
REFERENCES
[1] W. Frank, G. Koch and J. Mills, “Properties of Pure Alu-
minum,” In: J. Hatch, Ed., Aluminum: Properties and
Physical Metallurgy, Vol. 1, American Society for Metals,
2005, pp. 1-24.
[2] J. Kissel and R. Ferry, “Aluminum Structures,” John
Wiley & Sons, New York, 2002.
[3] T. Tsui, W. Oliver and G. Pharr, “Influences of Stress on
the Measuremen
indentation: Part I. Experimental Studies in an Aluminum
t of Mechanical Properties Using Nano-
Alloy,” Journal of Materials Research, Vol. 11, No. 3,
1996, pp. 752-759. doi:10.1557/JMR.1996.0091
[4] F. Zhang, H. Luo and S. Roberts, “Mechanical Properties
and Microstructure of Al2O3/Mullite Composite,” Journal
of Materials Science, Vol. 42, No. 16, 2007, pp. 6798-
6802. doi:10.1007/s10853-006-1402-z
[5] T. Sekino and K. Niihara, “Microstructural Characteris-
tics and Mechanical Properties for Al2O3/Metal Nano-
composites,” NanoStructured Materials, No. 5-8,
1995, pp. 663-666. doi:10.1016/0965-9773(95)00145-X
, Vol. 6
[6] T. Nieh, J. Wadsworth and O. Sherby, “Superplasticity in
Metals and Ceramics,” Cambridge University Press, Cam-
bridge, 1997. doi:10.1017/CBO9780511525230
[7] D. Munz and T. Fett, “Ceramics: Mechanical Properties,
Failure Behavior, Materials Selection,” Springer Series in
Materials Science, Vol. 36, Springer-Verlag, Berlin, 1999.
[8] W. Oliver and G. Pharr, “An Improved Technique for
Determining Hardness and Elastic Modulus Using Load
and Displacement Sensing Indentation Experiments,”
Journal of Materials Research, Vol. 7, No. 6, 1992, pp.
1564-1583.
[9] J. Pethica, R. Hutchings and W. Oliver, “Hardness Meas-
urement at Penetration Depths as Small as 20 nm,” Phi-
losophical Magazine A, Vol. 48, No. 4, 1983, pp. 593-606.
doi:10.1080/01418618308234914
[10] F. Fröhlich, P. Grau and W. Grellmann, “Performance and
Analysis of Recording Microhardness Tests,” Physica
Status Solidi (A), Vol. 42, No. 1, 1977, pp. 79-89.
[11] D. Newey, M. A. Wilkins and H. M. Pollock, “An Ul-
tra-Low-Load Penetration Hardness Tester,” Journal of
uments, Vol. 15, No. 1, 1982
022-3735/15/1/023
Physics E: Scientific Instr
pp. 119-122. doi:10.1088/0
,
Vol. 17, No. 7, 1992, pp. 28-33.
[13] G. Pharr, “Measurement of Mechanical Properties by
Ultra-Low Load Indentation,” Materials Science and En-
gineering A, Vol. 253, No. 1-2, 1998, pp. 151-159.
doi:10.1016/S0921-5093(98)00724-2
[14] T. Shen, C. Koch, T. Tsui and G. Pharr, “On the Elastic
Moduli of Nanocrystalline Fe, Cu, Ni, and Cu-Ni Alloys
Prepared by Mechanical Milling/Alloying,” Journal of
Materials Research, Vol. 10, No. 11, 1995, pp. 2892-
2896. doi:10.1557/JMR.1995.2892
[15] D. Stone, W. LaFontaine, O. Alexopoulos, T. Wu and
C.-Y. Li, “An Investigation of Hardness and Adhesion of
Sputter-Deposited Aluminum on Silicon by Utilizing a
Continuous Indentation Test,” Journal of Materials Re-
search, Vol. 3, No. 1, 1988, pp. 141-147.
doi:10.1557/JMR.1988.0141
[16] W. Nix, “Mechanical Properties of Thin Films,” Metal-
lurgical Transactions A, Vol. 20, No. 11, 1989, pp.
2217- 2245. doi:10.1007/BF02666659
[17] S. Hainsworth, H. Chandler and T. Page“Analysis of
t Loading Curves,”
, 1996, pp.
[12] G. Pharr and W. Oliver, “Measurement of Thin-Film Me-
chanical-Properties Using Nanoindentation,” MRS Bulletin,
,
Nanoindentation Load-Displacemen
Journal of Materials Research, Vol. 11, No. 8
1987-1995. doi:10.1557/JMR.1996.0250
[18] M. Tabbal, P. Merel, M. Chaker, M. El Khakani, E. Her-
bert, B. Lucas and M. O’Hern, “Effect of Laser Intensity
on the Microical Properties of
Pulsed LaserLike-Carbon Thin
structural and Mechan
Deposited Diamond-
Films,” Journal of Applied Physics, Vol. 85, No. 7, 1999,
pp. 3860-3865. doi:10.1063/1.369757
[19] S. Hainsworth, M. McGurk and T. Page, “The Effect of
Coating Cracking on the Indentation Response of Thin
Hard-Coated Systems,” Surface and Coatings Technology,
Vol. 102, No. 1-2, 1998, pp. 97-107.
doi:10.1016/S0257-8972(97)00683-X
[20] T. Tsui and G. Pharr, “Substrate Effects on Nanoindenta-
tion Mechanical Property Measurement of Soft Films on
Hard Substrates,” Journal of Materials Research, Vol. 14,
No. 1, 1999, pp. 292-301. doi:10.1557/JMR.1999.0042
[21] T. Tsui, J. Vlassak and W. Nix, “Indentation Plastic Dis-
placement Field: Part II. The Case of Hard Films on Soft
Substrates,” Journal of Materials Research, Vol. 14, No.
6, 1999, pp. 2196-2203. doi:10.1557/JMR.1999.0295
[22] S. Bec, A. Tonck, J.-M. Georges, E. Georges and J.-L.
Loubet, “Improvements in the Indentation Method with a
Surface Force Apparatus,” Philosophical Magazine A,
Vol. 74, No. 5, 1996, pp. 1061-1072.
doi:10.1080/01418619608239707
[23] N. Randall, “Direct Measurement of Residual Contact
Area and Volume during the Nanoindentation of Coated
Materials as an Alternative Method of Calculating Hard-
ness,” Philosophical Magazine A, Vol. 82, No. 10, 2002,
pp. 1883-1892. doi:10.1080/01418610208235700
[24] Y. Lim, M. Chaudhri and Y. Enomoto, “Accurate Deter-
mination of the Mechanical Properties of Thin Aluminum
Films Deposited on Sapphire Flats Using Nanoindenta-
tion,” Journal of Materials Research, Vol. 14, No. 6,
1999, pp. 2314-2327. doi:10.1557/JMR.1999.0308

Determination of Anodized Aluminum Material Characteristics by Means of Nanoindentation Measurements 1464
, N. Nagashima and S. Matsuoka, “Devel-[25] K. Miyahara
opment and Application of a Combined Atomic Force
Microscopy-Nanoindentation System with a Silicon Tip
and a Diamond Indenter,” Philosophical Magazine A, Vol.
82, No. 10, 2002, pp. 2149-2160.
doi:10.1080/01418610208235724
[26] R. Saha and W. Nix, “Effects of the Substrate on the De-
termination of Thin Film Mechanical Properties by Nano-
indentation,” Acta Materialia, Vol. 50, No. 1, 2002, pp.
23- 38. doi:10.1016/S1359-6454(01)00328-7
[27] J. Knapp, D. Follstaedt, S. Myers, J. Barbour and T. Fried-
man, “Finite-Element Modeling of Nanoindentation,” Jour-
nal of Applied Physics, Vol. 85, No. 1, 1999, pp. 1460-
1474.
[28] T. F. Page, G. M. Pharr, J. C. Hay, W. C. Oliver, B. N.
Lucas, E. Herbert and L. Riester, “Nanoindentation Chara-
cterization of Coated systems: P/S2—A New Approach
Using the Continuous Stiffness Technique,” Proceedings
of Materials Research Society Symposium, Vol.
1998, pp. 53-64.
522,
[29] J. Mencik, D. Munz, E. Quandt, E. Weppelmann and M.
Swain, “Determination of Elastic Modulus of Thin Layers
Using Nanoindentation,” Journal of Materials Research,
Vol. 12, No. 9, 1997, pp. 2475-2484.
doi:10.1557/JMR.1997.0327
[30] Z. Chen and S. Diebels, “Numerical Investigation of
Nanoindentation of Rubber-Like Hyperelastic Layers and
Parameter Re-Identification,” Mechanics of Materials,
2010, In Print.
[31] I. Sneddon, “The Relation between Load a
in the Axisymmetric Boussin
nd Penetration
esq Problem for a Punch of
Arbitrary Profile,” International Journal of Engineering
Science, Vol. 3, No. 1, 1965, pp. 47-57.
doi:10.1016/0020-7225(65)90019-4
[32] A. Love, “Boussinesq’s Problem for a R
terly Journal of Mathematics, Vol
igid Cone,” Quar-
. 10, No. 1, 1939, pp.
161-175. doi:10.1093/qmath/os-10.1.161
[33] A. Love, “The Stress Produced in a Semi-Infinite Solid by
Pressure on Part of the Boundary,” Philosophical Transac-
tions of the Royal Society A, Vol. 228, No. 659-669, 1929,
pp. 377-420. doi:10.1098/rsta.1929.0009
[34] J. Harding and I. Sneddon, “The Elastic Stresses Pro-
duced by the Indentation of the Plane Surface of a
Semi-Infinite Elastic Solid by a Rigid Punch,” Proceed-
ings of the Cambridge Philosophical Society, Vol. 41, No.
1, 1945, pp. 16-26.
[35] I. Sneddon, “Fourier Transforms,” McGraw-Hill, New
York, 1951.
[36] W. Oliver and G. Pharr, “Measurement of Hardness and
Elastic Modulus by Instrumented I
in Understanding and Refinemen
ndentation: Advances
ts to Methodology,”
Journal of Materials Research, Vol. 19, No. 1, 2004, pp.
3-20. doi:10.1557/jmr.2004.19.1.3
[37] Agilent Techologies G200 Nano Indenter User’s Guide,
Agilent Technologies, Inc., USA, 2009.
[38] A. Bolshakov and G. Pharr, “Influences of Pileup on the
Measurement of Mechanical Properties by Load and
Depth Sensing Indentation Techniques,” Journal of Ma-
terials Research, Vol. 13, No. 4, 1998, pp. 1049-1058.
doi:10.1557/JMR.1998.0146