Thickness Resonance Acoustic Microscopy for Nanomechanical

May 17, 2017 - Laser-generated Rayleigh wave for width gauging of subsurface lateral rectangular defects. Chuanyong Wang , Anyu Sun , Xiaoyu Yang ...
0 downloads 0 Views 7MB Size
Download PDF
Thickness Resonance Acoustic Microscopy for Nanomechanical Subsurface Imaging Gajendra S. Shekhawat,*,† Arvind K. Srivastava,§ Vinayak P. Dravid,*,† and Oluwaseyi Balogun*,‡ †

Department of Material Science and Engineering and NUANCE Center and ‡Department of Mechanical Engineering, Northwestern University, Evanston, Illinois 60208, United States § Systron Donner Inertial, 2700 Systron Drive, Concord, California 94518, United States ABSTRACT: A nondestructive scanning near-field thickness resonance acoustic microscopy (SNTRAM) has been developed that provides high-resolution mechanical depth sensitivity and sharp phase contrast of subsurface features. In SNTRAM technology, we excited the sample at its thickness resonance, at which a sharp change in phase is observed and mapped with a scanning probe microscopy stage in near field to provide nanometer-scale nanomechanical contrast of subsurface features/defects. We reported here the remarkable subsubsurface phase contrast and sensitivity of SNTRAM by exciting the sample with a sinusoidal elastic wave at a frequency equal to the thickness resonance of the sample. This results in a large shift in phase component associated with the bulk longitudinal wave propagating through the sample thickness, thus suggesting the usefulness of this method for (a) generating better image contrast due to high S/N of the transmitted ultrasound wave to the other side of the sample and (b) sensitive detection of local variation in material properties at much better resolution due to the sharp change in phase. We demonstrated that the sample excited at the thickness resonance has a more substantial phase contrast and depth sensitivity than that excited at off-resonance and related acoustic techniques. Subsurface features down to 5−8 nm lateral resolution have been demonstrated using a standard sample. KEYWORDS: acoustic microscopy, subsurface imaging, nanomechanical imaging, sample thickness resonance, contact resonance, scanning near-field detection

T

At present, voids in TSVs and other opaque components are identified using a combination of focused ion beam (FIB) milling and scanning electron microscopy (SEM), where the part is etched down to reveal the defect. This process is timeconsuming and destructive. X-ray tomography can be used for nondestructive void inspection, but it may not be amenable to online process control. Acoustic waves can be used to measure several properties, such as the thickness, sound speed, acoustic impedance, density, bulk modulus, and attenuation, for a wide range of materials, soft as well as hard. Since these are elastic strain waves that can travel through different materials, without any damage to them, they can be used for imaging subsurface structures noninvasively, a concept used widely in medical imaging. The need for improving resolution of detection using acoustics started getting addressed when several groups1−3 conceived of scanning acoustic microscopy (SAM). SAM has been used for mapping the nano- and microscale features in

hree-dimensional (3D) integration is enabling the manufacture of increasingly smaller and faster electronic components with enhanced data transfer speeds, low transmission losses, optimized performance, and low power consumption. This will, for example, lead to longer battery life in a portable electronic device and greater input−output functionality. Nondestructive inspection and metrology are important components of electronic manufacturing and allow identification of defects during component fabrication, feedback control of process parameters without halting the production process, and failure analysis of component performance. Defect metrology techniques with micro- and nanoscale spatial resolution are particularly important, as the shrinking dimensions of optically opaque components like through-silicon vias (TSVs) in integrated circuits increase. Typically, TSVs are narrow metal (copper or tungsten)-filled channels that provide electronic signaling pathways between multiple die layers and the external package. Voiding in TSVs can impede the flow of charges and compromise the performance of the electrical connection. Voiding in TSVs also results in a gradual buildup of mechanical stresses in between TSVs and the adjacent silicon layers due to the outgassing effect. © 2017 American Chemical Society

Received: March 29, 2017 Accepted: May 17, 2017 Published: May 17, 2017 6139

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145

Article

www.acsnano.org

Article

ACS Nano

Figure 1. Schematic of the thickness resonance ultrasound holography detection method. The sample is resonating at its natural resonance frequency via the piezotransducer underneath it.

Figure 3. (a) 1.5 μm size holes created with FIB etching, (b) crosssectional SEM image at 50° tilt, and (c) holes with reduced diameter deep inside the bulk from both the edge and top surface.

mechanism for images. The extreme sensitivity of the probe to the tip−sample contact stiffness facilitates the detection of local changes in the elastic stiffness within the compressed volume. Furthermore, the motion of the cantilever supporting the AFM probe tip depends on the sample surface displacement. As such, the AFM probe serves as a local elastic wave detector, and cantilever vibrations at ultrasonic frequencies are known to enhance the sensitivity to local elastic properties. The combination of ultrasonics with AFM platform is still a developing area of research, where understanding the physics of the mechanism of detection and determining the sensitivity of detection of surface or subsurface properties are still in progress. Despite the progress in ultrasonic-based methods, some of the major challenges are (a) how we can generate better image contrast due to high signal-to-noise of the transmitted ultrasound wave to the other side of the sample and (b) the sensitive detection of the local variation in material properties at much better resolution due to a sharp change in phase (especially when two materials have similar mechanical properties). Heterodyne ultrasonic detection methods rely on the mixing of two different ultrasound waves to generate a beat frequency, the phase of which provides elastic information on buried features with nanoscale resolution. The selection of tip/sample frequency to generate beat frequency is tricky. So far it has been based on a random search around the resonance frequency of matched piezoelectric transducers. The selection of tip/sample excitation frequency and beat frequency is very critical in generating high-fidelity images of subsurface features. Random selection of ultrasound waves is not only time-consuming but also prone to reproducibility problems if the sample is remounted on the sample piezo-transducer. To avoid this problem, we have come up with the idea of exciting the sample at its thickness resonance at which a sharp change in phase is observed. In this work, the sample is actuated at frequency f1 corresponding to the lowest longitudinal thickness resonance mode or resonance modes in the actuating transducer, and the AFM cantilever is operated in contact mode and driven with a sinusoidal voltage source at frequency f 2, such that, f1 − f 2 = fc,

Figure 2. Two possible contrast mechanisms for subsurface feature detection in ultrasonic AFM methods. (a) Bulk ultrasonic plane waves experience diffraction on coming across a feature in their path. If the sample surface is in the near field of this diffraction region and the diffraction leads to surface perturbation of the order of at least the z-resolution of AFM, then it can be detected by the AFM tip. (b) AFM stress field perturbation on encountering a subsurface feature. The reaction force on the AFM tip is different in the presence and absence of a feature.

construction materials, semiconductor devices, and biological specimens. However, the resolution of detection is still limited to ∼500 nm. The need for scales of observation is constantly being pushed to even smaller scales as the features in semiconductor devices shrink down to 10’s of nanometers. The International Technology Roadmap for Semiconductors4 highlights defect detection as a critical challenge for the future of the semiconductor industry as it continues to push for smaller nanoscale. The spatial resolution, i.e., the smallest defect size or resolvable separation between close defects, is limited by elastic wave diffraction to a fraction of the wavelength in the far field, a distance from the defect that is several times the wavelength. High attenuation at higher frequencies in the transmitting medium limits the penetration depth. Nanoscale spatial resolution can be achieved by combining atomic force microscopy (AFM) and ultrasonic microscopy, using an AFM probe as a local mechanical detector of elastic waves. Some of these methods, such as resonant difference atomic force microscopy (RD-AFM),5 ultrasonic force microscopy, and related methods based on heterodyne detection,6−18 propose a nonlinear force curve operation where ultrasonic scattering by subsurface features is detected and imaged at the sample surface, while others, such as atomic force acoustic microscopy (AFAM)19 or contact resonance frequency atomic force microscopy,10,20−23 propose linear force curve operation where a differential of stress field generated by an AFM tip20 owing to the presence of subsurface defects is proposed as the contrast 6140

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145

Article

ACS Nano

Figure 4. (a) AFM topography of the FIB-etched holes showing a very flat topography and (b) an SNTRAM image at sample thickness resonance (8.78 MHz). The remarkable phase contrast demonstrates the high sensitivity and sharp phase contrast from buried etched holes. Sample thickness resonance (fsample = 8.78 MHz) is calculated as vSi/2tSi where vSi = 8430 m/s and tSi = 500(±20) μm. (c and d) Topography and phase image of the same location, respectively, when the sample was at off resonance (piezotransduced oscillation frequency = 10.981 MHz). (e) Schematic of the sample orientation vis-à-vis the AFM probe over conical holes and (f) the corresponding cross-sectional phase profile of part (b). (g and h) Topography and phase image of FIB-etched holes, respectively, when the sample was at off resonance (piezotransduced oscillation frequency = 2.426 MHz). The phase image does not show any contrast coming from buried etched holes.

where fc is the contact resonance frequency of the cantilever. With this arrangement, we resolved a buried air channel with a

diameter of approximately 100 nm at a depth of 760 nm below the sample surface. On the basis of numerical simulation of elastic 6141

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145

Article

ACS Nano wave scattering from the buried circular cavity, we confirm that elastic wave contribution to the image contrast is negligible, leaving the tip−sample contact mechanics as the dominating factor. This technique called scanning near-field thickness resonance acoustic microscopy (SNTRAM) demonstrated a significant increase in subsurface contrast (Figure 1), large signal-to-noise ratio, and sharp change in acoustic phase as compared to when the sample is excited off-resonance as well as at piezotransducer resonance. In addition, we evaluate two possible contributions to image contrast from subsurface features for ultrasonic AFM methods where both the AFM probe tip and sample are excited at ultrasonic frequencies in contact mode. The two contrast mechanisms are ultrasonic scattering from subsurface features (Figure 2a) and the differential stress field caused by a subsurface feature when an AFM tip in contact generates a stress field in the sample (Figure 2b). The ultrasonic plane waves generated at the base of the sample travel through the sample bulk. Diffraction of these waves by a subsurface feature can be detected at the surface if the surface lies within the near field of this feature. Therefore, it is expected that the AFM tip generates a stress field in the sample. This stress field can get modified if there is a subsurface defect in its range. The dominant mechanism of detection will depend on the relative magnitudes of the signals and signal perturbations.

RESULTS AND DISCUSSION A series of cone-shaped buried air channels are etched into a (100) silicon wafer with a nominal thickness of 500 μm for the elastic wave imaging experiment. The channels are etched through the silicon wafer by focused ion beam (FIB) milling perpendicular to a cleaved edge of the Si wafer. A set of SEM images (at a tilt angle of 50°) of transverse sections etched through the thickness of the wafer by FIB milling are presented in Figure 3 to show the circular cross-section and dimensions of the channels at different depths from the sample surface. A thin film of palladium (Pd) was deposited along a line on the sample to serve as an etch stop, and FIB milling is performed from the sample edge to the Pd line to an appropriate depth where the buried channel can be visualized. In Figure 3a, two surface channels used as alignment marks for AFM imaging are observed, and the diameter of each channel is 1.5 μm at the entry point. The channel diameter decreases linearly to approximately 77 nm over a length of 8.3 μm, where it is buried at a depth of 780 nm below the sample surface. We remark that the transverse-sectioned images of the buried channel were obtained after the SNTRAM experiments were completed. The sample was mounted on a longitudinal thickness mode resonance piezoelectric transducer with a nominal center frequency of 10 MHz. A thin film of phenyl salycilate powder heated to a transition temperature of 40 °C24 is used as an adhesive layer between the sample and transducer and to enhance the transmission of elastic waves into the silicon sample. The velocity of the acoustic wave within the material in the longitudinal direction is 8430 m/s, and the wavelength depends on the frequency of excitation of the sample by a piezotransducer.25 Similar model samples have been fabricated by other groups such as Parlak et al.20 and Streigler et al.10 for use in AFAM experiments where holes are made underneath the top surface by pointing the ion beam on one of the edges of a Si(100) piece. Such kind of test structures not only avoid any effects from surface features or material heterogeneity but also help to understand the components of image contrast given

Figure 5. AFM topography (a) of the FIB etched holes shown in cross-section. (b) SNTRAH image at sample thickness resonance (8.98 MHz). The remarkable phase contrast demonstrates the high lateral sensitivity (5−8 nm) and sharp phase contrast from buried etched holes (125−150 nm). (c) Cross-sectional line profile demonstrating a very high subsurface lateral resolution of 5−8 nm, indicating a very sharp change in phase when the cantilever probe passes over buried FIB holes.

prior work on one of the mechanisms of detection, i.e., stress field. Buried Circular Air Channel. We have used a 500(±20 μm) thick silicon wafer, the thickness resonance of which, as per our calculation, was expected in the range of 8.7812 to 8.1058 MHz. The exact value of the sample resonance was measured as 8.78 MHz using laser interferometry. For a quick setup, we also devised a method to measure the thickness resonance by exciting the sample with a bottom transducer and analyzing the optical signal using a spectrum analyzer. As we sweep the frequency across the bottom transducer, a sharp peak is observed the moment the excitation frequency hits the sample thickness resonance, which was found to be exactly the same as shown by laser interferometry. Figure 4a shows a conventional topography image of the silicon surface below which holes are drilled, while Figure 4b is a corresponding (simultaneously recorded) SNTRAM phase image taken at the sample resonance frequency of 8.78 MHz using a 10 MHz ceramic crystal. The surface topography image is uniform and featureless. However, the corresponding SNTRAM phase image shown in Figure 4b clearly reveals the 6142

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145

Article

ACS Nano phase contrast reminiscent of buried conical holes created by FIB in silicon. We have demonstrated here a very sharp phase contrast and high mechanical sensitivity with this imaging method. For comparison, we also investigated the sample at off resonance, i.e., at 10.981 MHz, resonance frequency of the transducer (Figure 4d) and at 2.426 MHz (Figure 4h) (achieved using a 2 MHz crystal), which is much lower than the sample resonance (8.78 MHz). At both 10.981 and 2.426 MHz signals the phase image does not show any contrast coming from conical FIB holes. Figure 4e shows the schematic of the sample orientation vis-à-vis the AFM cantilever over conical holes, and Figure 4f shows the cross-sectional phase profile of the SNTRAM image shown in Figure 4b. To validate the SNTRAM image, the sample was subsequently physically cross-sectioned using FIB. The highresolution SEM image (at tilt angle of 50°) of FIB milled sections confirms the presence of conical holes (Figure 3a, b, and c). These results demonstrate that by exciting the sample at thickness resonance it is possible to achieve significantly enhanced phase contrast and depth sensitivity of the buried defects. Figure 5 shows the very high subsurface lateral resolution SNTRAM image of FIB-etched holes in silicon dioxide. The cross-sections of FIB-etched holes are shown in AFM topography in Figure 4a, while Figure 4b demonstrates the remarkable contrast coming from SNTRAM of etched holes buried around 125−150 nm. The cross-sectional profile in Figure 5c clearly demonstrates the lateral resolution down to 5−8 nm (full width at half-maxima). We applied the SNTRAM technique for direct application to semiconductor interconnect structures. Some of the challenges

in the industry with low-K dielectric interconnects are the effect of thermal treatment on polymeric regions, voids and cracks that appear across the trench wall/interface. Conventional methods of characterization include wet chemical etching and electrical testing, which is spatially insensitive and requires contact with the wafer. We demonstrated the effect of mechanical and thermal effects on polymer−metal trenches with sharp phase contrast and higher sensitivity. The trenches were etched in silicon dioxide and are around 1.5 μm deep. A 500 nm thin polymer was spin-coated followed by thermal annealing for curing. The topmost region of the trench structure was further patterned to deposit metal to study the effect of mechanical and thermal effects that these trenches undergo during several mechanical polishing and thermal cycles during semiconductor processing. This sample validated the direct application of SNTRAM in semiconductor trenches/ interfaces. Figure 6a shows a schematic of isolated shallow trench structures. Figure 6b depicts the conventional topography image, while Figure 6c the corresponding (simultaneously recorded) acoustic phase image obtained when the piezotransdcuer was in resonance. The resonance frequency of the sample and cantilever piezotransducers are 10.05 and 10.23 MHz, respectively. The acoustic phase image identifies the patterned metal−polymer interface and mechanical polishing effects on the trench after undergoing thermal treatments. It lacks both phase contrast and mechanical sensitivity. Subsurface defects are not very clearly identified here even after increasing the oscillation amplitude of the piezotransducers. However, when the thickness resonance mode of the sample was excited at 7.24 MHz, a remarkable phase contrast and mechanical sensitivity was observed in these interconnect structures and is

Figure 6. (a) Schematic of a metal−polymer trench structure with the topmost region further patterned into polymer and metal. The structures were patterned to study the phase contrast and mechanical sensitivity with the SNTRAM imaging method. (b) Typical topographical image showing the trench structure. (c) Ultrasound holography phase imaging when exciting the resonance of the transducer on which this sample was placed (10.05 MHz). The image identifies the metal−polymeric structures and shows the effect of thermal annealing and mechanical polishing effects on both polymer and metal. (d) SNTRAM image at sample thickness resonance (7.24 MHz). The remarkable phase contrast demonstrated the high mechanical sensitivity and sharp phase contrast with this method. The effect of thermal annealing, which results in hardening of the polymer and voids in the metal, is demonstrated here. (e) Cross-sectional profile across the image in Figure 5d, demonstrating sharp phase contrast and high mechanical sensitivity. 6143

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145

Article

ACS Nano shown in Figure 6d. The polymer coating shows a significant increase in mechanical rigidity when it went through a series of thermal treatments and mechanical polishing defects. The aluminum trenches show significant voids, which were created during mechanical polishing. Figure 6e is the corresponding phase profile of the SNTRAM phase image depicting a sharp phase change. It demonstrates the high sensitivity and sharp phase contrast of the SNTRAM technology. The thickness resonance of the sample was determined by interferometry. The dark contrast in the SNTRAM phase image corresponds to voids both in the metal and in the polymer, i.e., voids underneath the contact and trench wall. The SNTRAM image reveals polymer stiffness in the trenches and sidewall that may have happened when it underwent thermal treatments and/or due to poor adhesion with the oxide.

from Soft and Hybrid Nanotechnology Experimental (SHyNE) Resource (NSF-ECCS-1542205), the MRSEC program (NSF DMR-1121262) at the Materials Research Center; the International Institute for Nanotechnology (IIN); the Keck Foundation; and the State of Illinois, through the IIN. The authors are thankful to Dr. Yuan Li for generating the schematic of the SNTRAM technique. This work was supported by grants from the National Science Foundation Award Number 1256188, IDBR: Development of Higher Eigen mode Ultrasound Bioprobe for Sub-Cellular Biological Imaging. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

CONCLUSIONS In conclusion, development of SNTRAM imaging technology with sharp phase contrast and mechanical sensitivity provides a wide range of applications in nanomechanical imaging of semiconductor structures and a wide range of other materials. The representative examples of imaging buried conical holes created by FIB and low-K dielectrics patterned trenches demonstrate the versatility of this technique. We believe our development will fill a critical void in the subnanometer spatial range for nondestructive subsurface imaging in physical sciences.

REFERENCES (1) Sokolov, S. Y. Ultraacoustic Methods of Studying Properties of Hardened Steel and Detecting Intrinsic Flaws of Metal Articles. Zh. Tekh. Fiz. 1941, XI (1−2), 160−169. (2) Sokolov, S. Y. Ultraacoustic Microscope. Zh. Tekh. Fiz. 1949, 19, 271−273. (3) Briggs, G. A. D. Acoustic Microscopy; Clarendon Press: Oxford, 1992. (4) International Technology Roadmap for Semiconductors. http:// www.itrs.net/Links/2011ITRS/Home2011.htm (accessed April 2017). (5) Cantrell, S. A.; Cantrell, J. H.; Lillehei, P. T. Nanoscale Subsurface Imaging via Resonant Difference Frequency Atomic Force Ultrasonic Microscopy. J. Appl. Phys. 2007, 101, 1−8. (6) Garcia, R.; Herruzo, E. T. The Emergence of Multifrequency Force Microscopy. Nat. Nanotechnol. 2012, 7, 217−226. (7) Raman, A.; Trigueros, S.; Cartagena, A.; Stevenson, A.; Susilo, M.; Nauman, E.; Contera, S. A. Mapping Nanomechanical Properties of Live Cells using Multi-harmonic Atomic Force Microscopy. Nat. Nanotechnol. 2011, 6, 809−814. (8) Kolosov, O. V.; Castell, M. R.; Marsh, C. D.; Briggs, G. A. D. Imaging Elastic Nanostructures of Ge Islands using Ultrasonic Force Microscopy. Phys. Rev. Lett. 1998, 81, 1046−1049. (9) Kolosov, O. UFM Shakes out the Details at the Nanoscopic Scale. Mater. World. 1998, 6, 753−754. (10) Striegler, A.; Koehler, B.; Bendjus, B.; Roellig, M.; Mueller, M. K.; Meyendorf, N. Detection of Buried Reference Structures by use of Atomic Force Acoustic Microscopy. Ultramicroscopy 2011, 111, 1405− 1416. (11) Yamanaka, K.; Ogiso, H.; Kolosov, O. Ultrasonic Force Microscopy for Nanometer Resolution Subsurface Imaging. Appl. Phys. Lett. 1994, 64, 178−180. (12) Yamanaka, K. Ultrasonic Force Microscopy. MRS Bull. 1996, 21, 36−41. (13) Shekhawat, G. S.; Srivastava, A.; Avasthy, S.; Dravid, V. P. Ultrasound Holography for Non-Invasive Imaging of Buried Defects in Advanced Interconnect Architectures. Appl. Phys. Lett. 2009, 95, 263101. (14) Shekhawat, G.; Srivastava, A.; Avasthy, S.; Tark, S. H.; Dravid, V. P. Probing Buried Defects in Extreme Ultraviolet Multilayer Blanks using Ultrasound Holography. IEEE Trans. Nanotechnol. 2010, 9, 67. (15) Cuberes, M. T.; Alexander, H. E.; Briggs, G. A. D.; Kolosov, O. V. Heterodyne Force Microscopy of PMMA/Rubber Nanocomposites: Nanomapping of Viscoelastic Response at Ultrasonic Frequencies. J. Phys. D: Appl. Phys. 2000, 33, 2347. (16) Shekhawat, G. S.; Dravid, V. P. Nanoscale Imaging of Buried Nanostructures using Scanning Near-Field Ultrasound Holography. Science 2005, 310, 89. (17) Tetard, L.; Passian, A.; Lynch, R.; Shekhawat, G.; Dravid, V.; Thundat, T. Elastic Phase Response of Silica Nanoparticles Buried in Soft Matter. Appl. Phys. Lett. 2008, 93, 133113.

METHODS Scanning Probe Microscopy. The system incorporates a commercial contact mode AFM system (JEOL JSPM-5200). The piezoelectric transducer−sample assembly is mounted on a three-axis scanning translation stage in the AFM system. A silicon AFM cantilever with a spring constant of ∼40.0 N/m, a free resonance frequency of 300 kHz, and a nominal tip radius of 10 nm was used. A custom AFM cantilever holder was fabricated to mount a second thickness mode piezoelectric transducer on the cantilever for excitation of flexural vibrations. A portion of the photodetector output voltage is fed into a radio frequency lock-in amplifier to demodulate the defection of the cantilever at the difference frequencies; f = f1 − f 2, where f1 and f 2 are the sinusoidal actuation frequencies of the sample and cantilever. The reference input to the lock-in-amplifier is obtained by mixing a fraction of the drive voltages to two piezoelectric transducers in an electronic mixer and passing the output through a low-pass filter with a cutoff frequency of 1 MHz. A path-stabilized Michelson optical interferometer is incorporated into the setup to directly monitor the normal sample surface displacement without the probe-tip in place. Several references are available in the literature detailing the detection of surface displacements using the Michelson interferometer.

AUTHOR INFORMATION Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. *E-mail: [email protected]. ORCID

Gajendra S. Shekhawat: 0000-0003-3497-288X Vinayak P. Dravid: 0000-0002-6007-3063 Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work made use of the SPID facilities of the NUANCE Center at Northwestern University, which has received support 6144

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145

Article

ACS Nano (18) Tetard, L.; Passian, A.; Venmar, K.; Lynch, R.; Shekhawat, G.; Dravid, V.; Thundat, T. Imaging Nanoparticles in Cells by Nanomechanical Holography. Nat. Nanotechnol. 2008, 3, 501. (19) Rabe, U.; Arnold, W. Acoustic Microscopy by Atomic-Force Microscopy. Appl. Phys. Lett. 1994, 64, 1493−1495. (20) Parlak, Z.; Degertekin, F. L. Contact Stiffness of Finite Size Subsurface Defects for Atomic Force Microscopy: Three-Dimensional Finite Element Modeling and Experimental Verification. J. Appl. Phys. 2008, 103, 114910. (21) Hu, S. Q.; Su, C. M.; Arnold, W. Imaging of Subsurface Structures using Atomic Force Acoustic Microscopy at GHz frequencies. J. Appl. Phys. 2011, 109, 084324. (22) Hurley, D. C. Contact Resonance Force Microscopy Techniques for Nanomechanical Measurements. Applied Scanning Probe Methods XI, Scanning Probe Techniques. 2009, 97−138. (23) Hurley, D. C.; Muller, M. K.; Langlois, E. D.; Kos, A. B.; Barbosa, N., III Mapping Substrate/Film Adhesion with Contact Resonance Frequency Atomic Force Microscopy. Appl. Phys. Lett. 2009, 89, 021911. (24) Gliozzi, A. S.; Scalerandi, M.; Antonaci, P.; Bruno, C. L. E. Conditioning-Induced Elastic Nonlinearity in Hysteretic Media. Appl. Phys. A: Mater. Sci. Process. 2010, 100, 421−424. (25) Taubert, R.; Hudert, F.; Martels, A.; Merkt, F.; Habenicht, A.; Leiderer, P.; Dekorsky, T. Coherent Acoustic Oscillations of Nanoscale Au Triangles and Pyramids: Influence of Size and S ubstrate. New J. Phys. 2007, 9, 376.

6145

DOI: 10.1021/acsnano.7b02170 ACS Nano 2017, 11, 6139−6145