Polymer Testing 29 (2010) 95–99

Contents lists available at ScienceDirect

Polymer Testing
journal homepage: www.elsevier.com/locate/polytest

Test Method

Comparative analysis on the nanoindentation of polymers using atomic

force microscopy
Ah-Young Jee, Minyung Lee*
Division of Chemistry and Nanoscience, Ewha Womans University, Seoul 120-750, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Using atomic force microscopy (AFM) nanoindentation, we have measured the mechanical
Received 6 August 2009 properties of various polymers: low density polyethylene, polyvinyl alcohol, high density
Accepted 15 September 2009 polyethylene, ultrahigh molecular weight polyethylene, polyvinyl chloride, polycarbonate,
Nylon 6, poly(methyl methacrylate), polystyrene and polyacrylic acid. The hardness and
Keywords: Young’s modulus of the polymers were obtained by AFM through both the force–inden-
Nanoindentation
tation and area–depth curves. Our comparative analysis shows that the two methods give
Atomic force microscopy
almost identical results with self-consistency.
Hardness
Elastic modulus Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction advantage of AFM over conventional indentation is that it

measures the contact area and depth from the imaging data
Nanoindentation testing is a method of measuring the directly with the same tip used to obtain the force–
mechanical properties of a material such as hardness and displacement curve. This means that two sets of data are
Young’s modulus on the microscopic scale [1]. In conven- available from AFM nanoindentation [10,13,20]. With the
tional nanoindentation, a small tip is pressed into a sample force–displacement curve, the mechanical properties can
with a known load (force) and retracted sequentially, which be obtained by applying the O&P procedure. In addition, the
generates a force–displacement curve. The indentation hardness and Young’s modulus can be directly obtained
procedure developed by Oliver and Pharr (O&P) has been from the imaging data. By AFM nanoindentation, we have
widely used for hard materials such as metals and ceramics measured the mechanical properties of various polymers:
[2,3]. The O&P procedure is based on Sneddon’s approach low density polyethylene (LDPE), polyvinyl alcohol (PVA),
in which force F is given as a power low of displacement h: high density polyethylene (HDPE), ultrahigh molecular
FðhÞ ¼ ahm , where a and m are constants that depend on weight polyethylene (UHMWPE), polyvinyl chloride (PVC),
the mechanical properties of the sample and tip geometry. polycarbonate (PC), polycaprolactam (Nylon 6), poly(-
The model assumes: the sample deformation is elastic as methyl methacrylate) (PMMA), polystyrene (PS) and poly-
well as plastic; deformation is time-independent; the acrylic acid (PAA). The results, processed by using O&P and
indenter is a rigid punch; no sink-in or pile-up exists. image analysis, show that hardness and Young’s modulus
Commercial nanoindenters that utilize the O&P proce- obtained from the two methods give almost identical
dure do not offer a wide range of loads necessary for soft results.
materials. As an alternative, atomic force microscopy (AFM)
nanoindentation has been used to measure the mechanical
properties of polymers and biomaterials [4–21]. An 2. Experimental details

LDPE (MW 35,000), HDPE, (MW 125,000), UHMWPE

* Corresponding author. Tel.: þ82 2 3277 2383; fax: þ82 2 3277 2384. (MW 4,500,000), PS (MW 192,000), PVA (MW 115,000), PC
E-mail address: [email protected] (M. Lee). (MW 64,000), PMMA (MW 120,000), and PAA (MW

0142-9418/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymertesting.2009.09.009
96 A.-Y. Jee, M. Lee / Polymer Testing 29 (2010) 95–99

240,000) were purchased from Aldrich. Nylon 6 (MW 3. Results and discussion
18,000) was bought from Polyscience.
Solvents to prepare polymer solutions were: THF for A schematic illustration of the O&P procedure is shown
LDPE, PVC, PMMA, and PC; decalin for HDPE and UHMWPE; in Fig. 2. The hardness is defined by
6:1 mixture of DMSO and H2O for PVA. cyclohexane for PS;
water for PAA; 2, 2, 2-trifluoroethanol for Nylon 6. The Fmax ðhmax Þ
H ¼ (1)
samples were prepared by compression molding, annealing Aðhc Þ
at 110–200  C depending on polymer, and cooling to room
and the reduced elastic modulus is given by
temperature. The resulting sample is a disk having the
thickness of ca. 1 mm and a diameter of 1.5 cm with a flat rffiffiffiffiffiffiffiffiffiffiffi
S p
surface. Er ¼ (2)
2b Aðhc Þ
Nanoindentation measurements were carried out at
room temperature with an AFM (XE-70, PSIA) in an acoustic where S is the slope of the unloading curve and b ¼ 1.034
box. The indenter was a Berkovich tip (sapphire cantilever for a Berkovich indenter. The elastic modulus (E) can be
with a diamond tip) having a resonance frequency of calculated from the reduced modulus and the indenter
45.9 kHz and a spring constant of 145 N/m. As shown in modulus Ei by
Fig. 1, the tip geometry consists of a sharp three-sided    
pyramid, the base of which is an equilateral triangle with 1 1  n2 1  n2i
¼ þ (3)
a half angle of 30 . The radius of curvature at the tip apex Er E Ei
was nominally less than 25 nm. Load on the surface layers
where n and ni are the Poisson ratio of the sample and
of polymers ranged from 0.1 to 2.25 mN. A speed of 300 nm/
indenter, respectively. The diamond tip is much stiffer than
s was used for all the experiments. The AFM images were
polymer (E < < Ei), and thus the Eq. (3) is reduced to
obtained with the same Berkovich tip by a non-contact
1=Er zð1  n2 Þ=E. The contact depth (hc) is given by
mode at 0.5 Hz scan rate.
The measurements of AFM nanoindentation have been 3Fmax
performed following some general rules [2]. First, the hc ¼ hmax  (4)
S
indentation depth should be deep enough to minimize the
surface effect. However, the indentation depth should be The parameter 3 depends on the tip shape, spanning from
less than 10% of the film thickness when the sample is 0.72 (conical) to 1 (flat). For the Berkovich indenter, 3 ¼ 0.75
mounted on a hard substance. Otherwise, the measured is recommended. If hc is known, then we can calculate the
value is usually larger than it should be, due to the effect of area function
the support. The second is sink-in or pile-up effect
Aðhc Þ ¼ 24:56 h2c (5)
appearing in the unloading curve as a bulge (‘‘nose’’). It can
be overcome by performing the nanoindentation at a high It should be noted that AFM nanoindentation is different
unloading rate with long holding time. Lastly, to obtain from the conventional nanoindentation because it utilizes
more accurate results, it is necessary to calibrate AFM the cantilever deflection.
components such as cantilever spring constant, tip shape,
and piezoelectric scanner movement. FðdÞ ¼ kd (6)

Fig. 1. The schematic illustration of a Berkovich tip with the definition of

a half angle. A is the indentation area, h is the indentation depth, and a is the Fig. 2. The schematic illustration of the force–displacement curve. The
half angle. meanings of parameters are described in the text.
 A.-Y. Jee, M. Lee / Polymer Testing 29 (2010) 95–99 97

2500
0

2000
-1000

Depth (nm)
PAA
1500 PAA PMMA
Load (nN)

PMMA PS
PS -2000 Nylon 6
Nylon 6 PC
1000 PC PVC
PVC UHMWPE
UHMWPE -3000 HDPE
500 HDPE PVA
PVA LDPE
LDPE -4000
0 0.2 0.4 0.6 0.8 1.0 1.2
0 100 200 300 400 500 600 700 Width (µm)
Piezo displacement (nm)
Fig. 5. The indentation–depth profiles of polymers used in this work.
Fig. 3. The experimentally measured force–displacement curve for ten
polymers.
polymers. The contact area images of ten polymers, recor-
ded by AFM, are shown in Fig. 4. Smaller contact area
where k is the cantilever stiffness and d is the cantilever indicates higher hardness. The hardness was calculated
deflection. That is, the raw data, which is the force– from H ¼ F/A, where F is the applied force and A is the area
displacement (F–z) curve, should be converted to the force– obtained from the image.
indentation (F–h) curve. The indentation depth is given by The indentation depth profiles are shown in Fig. 5. The
curves were obtained by sweeping the indentation image
h ¼ zd (7)
parallel to the baseline of the triangle that crosses the
where z is the piezo displacement and d is the cantilever deepest point. Any sink-in or pile-up effects were not visibly
deflection. The conversion of the F–z curve to the F–h curve seen in the data. The indentation depth is given in the
is depicted in detail in Fig. 2. following order: LDPE (3.43 mm), PVA (2.16 mm), HDPE
Ten different polymers were chosen to cover a wide (1.60 mm), UHMWPE (1.27 mm), PVC (1.19 mm), PC (0.96 mm),
range of Young’s modulus (0.1–5 GPa). Fig. 3 shows the Nylon 6 (0.81 mm), PS (0.73 mm), PMMA (0.70 mm), and PAA
force–displacement curve for these polymers. The (0.63 mm).
maximum load applied was 2.25 mN. The data show that In order to obtain Young’s modulus from the AFM F–h
LDPE has the lowest slope and PAA is the highest with the curve, various models have been suggested such as Hertz,
order LDPE < PVA < HDPE < UHMWPE < PVC < PC < Nylon Johnson–Kendall–Roberts (JKR), Derjaguin–Muller–Top-
6 < PS < PMMA < PAA. After the conversion of F–z to F–h, orov (DMT) and Dimitriadis. The Hertz model assumes
the hardness and Young’s modulus were obtained through infinite thickness, elastic deformation and no adhesion
Eq. (1) to Eq. (5). between tip and sample [22]. For this reason, usually the
The AFM images of the contact area and depth also give loading curve is used to apply the Hertz model. JKR and
information on hardness and elastic modulus of the DMT accommodate adhesion forces to the Hertz model

0.5 µm

LDPE PVA HDPE UHMWPE PVPC

PC Nylon 6 PS PMMA PAA

Fig. 4. The indentation images of polymers used in this work.

98 A.-Y. Jee, M. Lee / Polymer Testing 29 (2010) 95–99

Table 1 The Hertz model states that the Young’s modulus of

The mechanical properties of ten polymers measured by AFM a material is inversely proportional to the square of
nanoindentation.
indentation depth at a given force.
Polymer Poisson Olive & Pharr Image analysis We present all the data obtained by O&P and image
ratio analysis in Table 1. Hardness and Young’s modulus of the
H (MPa) E (GPa) H (MPa) E (GPa)
LDPE 0.40 22.4 0.19 21.1 0.16 ten polymers are plotted together in Fig. 6. The correlation
PVA 0.49 39.2 0.40 37.8 0.37 is linear with the correlation coefficient of R2 ¼ 0.99 for
HDPE 0.45 61.8 0.73 61.5 0.75 both cases. It seems that both values analyzed by F–h
UHMWPE 0.46 73.1 1.39 71.2 1.13
curves and A–d images agree well within experimental
PVC 0.38 86.9 1.58 85.5 1.40
PC 0.40 127.2 2.27 125.8 1.98
error.
Nylon 6 0.35 157.9 3.15 157.3 3.11 It can, therefore, be stated that AFM nanoindentation is
PS 0.33 167.5 3.72 166.3 3.56 a versatile tool to measure the mechanical properties of
PMMA 0.35 187.1 3.83 185.4 3.97 polymers, and both of O&P and image analysis are self-
PAA 0.46 244.8 4.67 243.4 4.50
consistent.

4. Conclusions
during contact. JKR is useful for a large tip and soft mate-
rials with high adhesion [23]. DMT is good for a small tip Using AFM nanoindentation, we have measured hard-
and stiff materials with low adhesion [24]. For a thin film ness and Young’s modulus of ten different polymers: LDPE,
supported on a hard substance, a modified Hertz model PVA, HDPE, UHMWPE, PVC, PC, Nylon 6, PS, PMMA and
was suggested by Dimitriadis and coworkers [25]. To apply PAA. Compared with conventional indentation, AFM
the Dimitriadis model, the film thickness should be known measures the contact area and depth from the imaging data
accurately [26]. directly with the same tip used to obtain the load–
From the indentation depth profile, the Young’s displacement curve. With the force–indentation curve, the
modulus can be calculated by applying the Hertz model [8]. mechanical properties of the polymers were obtained by
For a Berkovich tip, it is given by applying the O&P procedure. In addition, the hardness and
Young’s modulus were directly obtained from the imaging
3Etan a 2 data. The results, processed by using O&P and image anal-
F ¼ d (8)
4ð1  n2 Þ ysis, show that the mechanical properties obtained from
where d is the indentation depth identical to h in the O&P the two methods give almost identical results. It can,
procedure. The half angle a as shown in Fig. 1 is 30 . The therefore, be stated that AFM nanoindentation is a versatile
values of Poisson ratio n for ten polymers were taken from tool to measure the mechanical properties of polymers.
the literature [27–33], and given in Table 1. One can
calculate the elastic modulus by fitting the F–h curve to Eq.
References
(8). In this work, the elastic modulus was simply calculated
from the maximum depth at the maximum load. That is, [1] W.C. Oliver, G.M. Pharr, An improved technique for determining
  hardness and elastic modulus using load and displacement sensing
4 1  n2 Fmax indentation experiments. J. Mater. Res. 7 (1992) 1564.
E ¼ (9) [2] W.C. Oliver, G.M. Pharr, Measurement of hardness and elastic
3tan a d2 modulus by instrumented indentation: advances in understanding
max
and refinements to methodology. J. Mater. Res. 19 (2004) 3.
[3] D. Tranchida, S. Piccarolo, J. Loos, A. Alexeev, Accurately evaluating
Young’s modulus of polymers through nanoindentations:
a phenomenological correction factor to the Oliver and Pharr
300 procedure. Appl. Phys. Lett. 89 (2006) 171905.
[4] R.E. Mahaffy, S. Park, E. Gerde, J. Kas, C.K. Shih, Quantitative analysis
Olive & Pharr of the viscoelastic properties of thin regions of fibroblasts using
250 Image analysis atomic force microscopy, 86 (2004) 1777.
[5] T.-H. Fang, W.-J. Chang, S.-L. Tsai, Nanomechanical characterization
of polymer using atomic force microscopy and nanoindentation.
Hardness (MPa)

200 Microelectron. J. 36 (2005) 55.

[6] D. Tranchida, S. Piccarolo, On the use of the nanoindentation
unloading curve to measure the Young’s modulus of polymers on
150
a nanometer scale. Macromol. Rapid Commun. 26 (2005) 1800.
[7] C.A. Clifford, M.P. Seah, Quantification issues in the identification of
100 nanoscale regions of homopolymers using modulus measurement
via AFM nanoindentation. Appl. Surf. Sci. 252 (2005) 1915.
[8] M.J. Rosenbluth, W.A. Lam, D.A. Fletcher, Force microscopy of non-
50 adherent cells: a comparison of Leukemia. Biophys. J. 90 (2006)
2994.
[9] A.C. Fischer-Cripps, Critical review of analysis and interpretation of
0 nanoindentation test data. Surf. Coat. Technol. 200 (2006) 4153.
0 1 2 3 4 5 [10] D. Tranchida, S. Piccarolo, M. Soliman, Nanoscale mechanical char-
acterization of polymers by AFM nanoindentation: critical approach
Young's modulus (GPa) to the elastic characterization. Macromolecules 39 (2006) 4547.
[11] K. Miyake, N. Satomi, S. Sasaki, Elastic modulus of polystyrene film
Fig. 6. The correlation between hardness and elastic modulus calculated by from near surface to bulk measured by nanoindentation using
the O&P procedure and image analysis, respectively. atomic force microscopy. Appl. Phys. Lett. 89 (2006) 031925.
 A.-Y. Jee, M. Lee / Polymer Testing 29 (2010) 95–99 99

[12] M.P.E. Wenger, L. Bozec, M.A. Horton, P. Mesquida, Mechanical [22] H.J. Hertz, On the contact of rigid elastic solids and hardness. Reine
properties of collagen fibrils. Biophys. J. 93 (2007) 1255. Angew. Math. 92 (1881) 156.
[13] D. Tranchida, S. Piccarolo, J. Loos, A. Alexeev, Mechanical charac- [23] K.L. Johnson, K. Kendall, A.D. Roberts, Surface energy and the contact
terization of polymers on a nanometer scale through nano- of elastic solids. Proc. R. Soc. London Ser. A. 324 (1971) 301.
indentation. A study on pile-up and viscoelasticity. Macromolecules [24] B.V. Derjaguin, V.M. Muller, Y.u.P. Toporov, Effect of contact defor-
40 (2007) 1259. mations on the adhesion of particles. J. Colloid Interface Sci. 53
[14] D. Tranchida, Z. Kiflie, S. Piccarolo, Viscoelastic recovery behavior (1975) 2.
following atomic force microscope nanoindentation of semi- [25] E.K. Dimitriadis, F. Horkay, J. Maresca, B. Kachar, R.S. Chadwick,
crystalline polyethylene. Macromolecules 40 (2007) 7366. Determination of elastic moduli of thin layers of soft material using
[15] B. Tang, A.H. Ngan, J.B. Pethica, A method to quantitatively measure the atomic force microscope. Biophys. J. 82 (2002) 2798.
the elastic modulus of materials in nanometer scale using atomic [26] H. Brochu, P. Vermette, Young’s moduli of surface-bound liposomes
force microscopy. Nanotechnology 19 (2008) 495713. by atomic force microscopy force measurements. Langmuir 24
[16] F. Bedoui, F. Sansoz, N.S. Murthy, Incidence of nanoscale heteroge- (2008) 2009.
neity on the nanoindentation of a semicrystalline polymer: experi- [27] D. Briassoulis, E. Schettini, Measuring strains of LDPE films: the
ments and modeling. Acta Mater. 56 (2008) 2296. strain gauge problems. Polym. Test. 21 (2002) 507.
[17] D. Passeri, A. Bettucci, A. Biagioni, M. Rossi, A. Alippi, M. Lucci, [28] Christele Bartholome, Alain Derre, Olivier Roubeau, Cecile Zakri,
I. Davoli, S. Berezina, Quantitative measurement of indentation Philippe Poulin, Electromechanical properties of nanotube-PVA
hardness and modulus of compliant materials by atomic force composite actuator bimorphs. Nanotechnology 19 (2008) 325501.
microscopy. Rev. Sci. Instrum. 79 (2008) 066105. [29] J.G.J. Beijer, J.L. Spoormaker, Modeling of creep behaviour in injec-
[18] B. Cappella, D. Silbernagl, Nanomechanical properties of polymer tion-moulded HDPE. Polymer 41 (2000) 5443.
thin films measured by force-distance curves. Thin Solid Films 516 [30] S.A. Chizhik, Z. Huang, V.V. Gorbunov, N.K. Myshkin, V.V. Tsukruk,
(2008) 1952. Micromechanical properties of elastic polymeric materials as pro-
[19] W. Ngwa, K. Chen, A. Sahgal, E.V. Stepanov, W. Luo, Nanoscale bed by scanning force microscopy. Langmuir 14 (1998) 2606.
mechanics of solid-supported multilayered lipid films by force [31] T.D. Fornes, D.R. Paul, Modeling properties of nylon 6/clay
measurement. Thin Solid Films 516 (2008) 5039. nanocomposites using composite theories. Polymer 44 (2003)
[20] D. Tranchida, Z. Kiflie, S. Piccarolo, Viscoelastic recovery behavior 4993.
following atomic force microscope nanoindentation of semi- [32] W. Xu, X.F. Yao, H.Y. Yeh, G.C. Jin, Fracture investigation of PMMA
crystalline polyethylene. Macromolecules 40 (2007) 7366. specimen using coherent gradient sensing (CGS) technology. Poly-
[21] T. Thome, S. Fouchez, S. Delalande, Determination of silicone coating mer Test 24 (2005) 900.
Young’s modulus using atomic force microscopy. Physica B 404 [33] Toshikazu Takigawa, Yoshiro Morino, Kenji Urayama, Toshiro
(2009) 22. Masuda. Polym. Gels Networks 4 (1996) l.