Open Access
Add to BibSonomyAbstract
Bacteria biofilms in chronically infected wounds significantly increase the burden of healthcare costs and resources for patients and clinics. Because biofilms are such an effective barrier to standard antibiotic treatment, new methods of therapy need to be developed to combat these infections. Our group has demonstrated the potential of using Laser Generated Shockwaves as a potential therapy to mechanically disrupt the bacterial biofilms covering the wound. Previous studies have used rigid silica glass as the shockwave propagation medium, which is not compatible with the intended clinical application. This paper describes the exploration of five candidate flexible plastic films to replace the glass substrate. Each material measured 0.254 mm thick and was used to generate shockwaves of varying intensities. Shockwave characterization was performed using a high-speed Michelson displacement interferometer and peak stress values obtained in the flexible substrates were compared to glass using one-way nested Analysis of Variance and Tukey HSD post-hoc analysis. Results demonstrate statistically significant differences between substrate material and indicate that polycarbonate achieves the highest peak stress for a given laser fluence suggesting that it is optimal for clinical applications.
© 2015 Optical Society of America
1. Introduction
It has been well documented that 1) chronic wound infections significantly hinder the healing process, increasing healthcare costs and consumption of resources, and 2) bacteria biofilms play an integral role in the persistence of infected wounds [1–4]. When bacteria infect a wound and proliferate, they create a protective habitat to survive called a biofilm. This film is composed of an extracellular polysaccharide matrix, sometimes referred to as glycocalix or extracellular polymeric substance (EPS) [4–7]. The biofilm serves many functions that aid the survival of bacterial colonies. First, it is the beginning of the adhesion process for cell-cell or cell-surface interaction. Second, it is the avenue for collecting nutrients and minerals to sustain the bacteria [8], which in the case of open wounds are likely siphoned from the underlying tissue, leading to its necrosis. Third, and most pertinent to treatment, the biofilm is a protective mechanical and chemical shield from the environment around the bacteria. Studies have shown that standard antibiotic treatments are ineffective against biofilms, requiring up to 1000 times the normal dosage to generate significant therapeutic effects, which in turn becomes toxic to the patient [9,10].
Despite research in chronic infected wound care, there is still no treatment effective to replace standard sharp debridement, i.e. removing bacterial biofilms without damaging the underlying tissues. Recent advances in mechanically addressing infected wounds include negative-pressure wound therapy (NPWT), Low-Intensity Ultrasound (LIUS), and Laser-Generated Shockwaves (LGS). NPWT has demonstrated improved wound healing, however has had no effect on the reduction of biofilms [11]. LIUS, with frequencies ranging from 100 kHz to 3 MHz, has been shown to promote tissue growth and healing. However, the same permeabilizing effect for tissue growth leads to enhanced bacterial growth [12]. It is also postulated that given published values for bacterial adhesion strength, the ultrasound intensity necessary to delaminate the bacteria would cause significant underlying tissue damage.
LGS have recently been adapted as a potential solution to safe, efficient biofilm removal [13]. These shockwaves are generated using a short-pulsed laser incident on absorptive metal films deposited onto a carrier substrate. Absorption of laser energy by the metal film causes rapid thermal expansion and material ablation, creating a pressure wave, which propagates as a uniform plane wave through the substrate. To maximize conversion of electromagnetic energy to mechanical energy in a single direction, an optically transparent but acoustically rigid confining layer is placed over the thin metal film, focusing the shockwave into the carrier substrate. (Fig. 1)
Fig. 1 (left) Basic material setup for Laser-Generated Shockwaves. (right) Clinical setup of Laser-Generated Shockwave treatment of bacterial biofilms in infected open wounds.
A laser-generated shockwave is created by a plasma-generating/thermal-expansion process from a single laser pulse, propagating as a single, high frequency, compressive pressure wave without the tensile component found in most common shockwave profiles. This compressive pressure wave is a longitudinal wave following the same physical principles of acoustic wave mechanics and impedances. The shockwave must transfer through a continuous solid or semi-solid medium to the tissue. As in ultrasound, any air interfaces will cause almost total reflection and not transfer the mechanical energy to the biofilm. In the clinical application, the shockwave propagates through the carrier substrate and transferred through an acoustic medium, similar to ultrasound gel, to the biofilm and tissue. At the interface between the biofilm and underlying tissue, acoustic impedance differences will cause a portion of the shockwave energy to reflect back through the biofilm, with the reflected wave intensity based on the differences in physical properties between the biofilm and underlying tissue. This reflection converts the compressive wave into a tensile wave, which imparts significant axial loading onto the structural rigidity of the biofilm [14]. Delamination occurs when the peak tensile stress of the reflected wave is greater than the adhesion strength of the biofilm. It is also postulated that if the physical length of the shockwaves are on the same scale as the bacterial cells themselves, the same physical stresses occur against their plasma membranes, potentially leading to cellular death, but that analysis is outside the scope of this study.
Development of optimized LGS generation has been conducted using soda-lime glass or fused silica glass as the carrier substrate [15,16]. However, the rigid structure of glass is not compatible with clinical translation. Open-wound infections can occur on any contour of the body, requiring an LGS carrier substrate to be flexible and adapt to human anatomy (Fig. 1). Therefore, the present study is a comparative analysis of five flexible, plastic film substrates, as well as a glass control substrate, used to generate the shockwaves. The shockwave profiles created from five different plastic film materials: polycarbonate (PC), polyethylene terephthalate (polyester, PE), poly-ethyl-ethyl-ketone (PEEK), polyvinylchloride (PVC), and cellulose acetate (Acetate), were characterized across six laser energy fluences. These materials were compared to sodalime glass, and variations in shockwave peak stress and stress pulse decay times were noted. These two values most significantly affect the tensile wave reflection and, ultimately, one’s ability to delaminate bacterial biofilms. Our use of pulsed decay time as the temporal metric of choice over FWHM is motivated by two factors. First, with propagating pressure waves, the higher-pressure regions of the pulse travel at a faster speed than the lower pressure regions, creating wave steepening of the shock front. For the majority of the shockwave profiles, the rise times are very rapid (<2ns), so the FWHM value is predominantly dependent on the decay time, which is what is reported. Second we focus on decay time instead of FWHM because, in the intended clinical application, the intensity of the reflected tensile wave (engineered to illicit biofilm delamination) is strongly dependent on destructive interference with the tail end of the transmitting compressive wave. If the decay time is too long, the tensile wave will be negated by the compressive tail and will not cause delamination of biofilm. Since bacterial delamination is the ultimate aim, shockwave measurements are recorded from a surface after transmission through ultrasound gel, mimicking clinical setup.
2. Materials and methods
2.1 Plastics sample preparation
Stock of each plastic film was purchased in thickness of 0.254 mm from McMaster-Carr (Santa Fe Springs, CA, USA). Samples of each plastic film were cut into 25.4 x 76.2 mm2 to match the size of a standard microscope glass slide. The laser ablation metal used was titanium, due to its well-known biocompatibility and its high absorption of laser energy at near infrared (NIR) wavelengths. Titanium film was deposited on each plastic sample by RF sputtering using Denton Discovery 550 (Moorestown, NJ, USA) in a class 1000 cleanroom environment. Thickness of titanium was 500 nm ± 50 nm and was measured using Veeco Dektak 6M profilometer (Fullerton, CA, USA). To stay consistent with the clinical translation of this technology, coupling of the plastic samples to the interferometry measurement surface was completed using Ultra/Phonics ultrasound gel (Newark, NJ, USA), and we used a glass slide sputtered with 50 nm aluminum coating as the mirrored interferometry surface (Fig. 2). Thickness of ultrasound gel was kept constant at 0.040 inches using glass spacers between the samples and the mirrored surface. Waterglass (Rutland, Vermont, USA) was spread over the titanium surface and allowed to dry to create the confining layer (Fig. 2).
Fig. 2 Sample setup for characterization of different plastic materials. 1. Plastic sample coated with 500nm Ti and waterglass. 2. Ultrasound gel layer for coupling. 3. Glass slide coated with 50nm aluminum for interferometry measurement. 4. Glass spacers, 0.040 in, to keep consistent the distance traveled by shockwaves through ultrasound gel.
2.2 Interferometer setup and operation
Because these unique shockwaves are generated with a laser pulse width of <10 nanoseconds, their temporal width can range from 20 nanoseconds to 500 nanoseconds, with a rise time of 2 to 10 nanoseconds, depending on the coupling medium and dispersion characteristics of the material they are traveling through. In order to image and characterize these shockwaves (especially rise time), a sampling rate of at least 1 GHz is required. This is currently impossible using optical techniques and requires an indirect method of imaging. We are able to observe and record shockwave pulse profiles using a Michelson interferometer designed for high-speed displacement detection.
The displacement interferometer used to characterize the shockwave profiles is based off of a Michelson interferometer. A frequency-stabilized HeNe laser (Melles Griot, Carlsbad, CA) is split into two beams using a 50/50 beamsplitter, one beam (reference beam) traveling to a fixed mirror, and the second beam (sample beam) traveling to the mirrored surface of our sample. Beams are reflected back to the beamsplitter, recombined, and focused onto a high-speed (30ps rise time) photodiode. Two overlapping beams create a pattern of constructive and destructive interference, which the photodiode can detect, when properly aligned. Signal from the photodiode is amplified through two 20 dB low noise amplifiers and sent to a 2.25 GHz digital oscilloscope. (Fig. 3(a)) Waveforms generated on the oscilloscope were then transferred to a computer for data analysis using Matlab (Mathworks, Inc., Natick, MA) and OriginPro software (OriginLab Corp., Northhampton, MA).
Fig. 3 (a) Diagram of the experimental setup for characterization of plastic materials. (b) FWHM of Nd:YAG laser pulse for each energy density used in the experiment.
Under non-loading conditions, the signal generated is background noise within the system. Because the laser source for the interferometer is frequency stabilized the coherence length is much longer than the total path length of the interferometer thus relaxing the equal path length requirement typically imposed on both arms of the interferometer. During shockwave loading, the free surface of the sample mirror experiences a displacement correlating to the stress of the traveling shockwave. This surface displacement changes the path-length that the sample beam travels before arriving at the photodiode, causing that beam to arrive at the photodiode with a certain spatial offset. As the spatial offset changes, the orientation of constructive to destructive interference patterns change, therefore changing the voltage signal out of the photodiode, and the digital oscilloscope records that voltage signal change versus time. To achieve maximum signal intensity, and more importantly maximum extinction ratio, both beams of the interferometer must travel to the photodiode collinearly.
2.3 Shockwave generation and data collection
Shockwaves were generated using a Nd:YAG pulsed laser, operating at 1064 nm wavelength (Brilliant B, Quantel, France), focused onto a 3 mm 1/e spot size. Six energy levels were used to characterize shockwave profiles across the different plastic substrates, and each level was obtained by changing the delay between the flash lamp excitation and Q-switching within the laser. The laser pulse width, in FWHM, for each energy level is shown in Fig. 3(b). Energy densities were calculated based on the measured energy per pulse and ablated spot-size. Shockwaves were generated 6 times, at different spots, for each energy level, on each substrate (N = 216), and shockwaves for each energy level were tested on the same material sample to reduce manufacturing variance. The interference pattern of each shockwave measured (waveform) on the digital oscilloscope was triggered by an external line directly from the laser power supply and then transferred to a computer for analysis.
2.4 Data analysis
Waveforms recorded from the shockwave disturbance of the mirrored sample surface follow the pattern of a down-chirped waveform, with the fringes representative of changes in free surface acceleration. The waveform fringe frequency-change from high to low represent the surface deceleration as the shockwave passes through the free surface. Each transition from signal peak to signal valley represents a quarter-wavelength movement of the sample surface, corresponding to a complete transition from constructive interference to destructive interference of the superimposed HeNe laser beams on the photodiode. The waveform can be represented by the chirped signal equation, Eq. (1), as a function of free surface displacement, u0(t).
where A0 is the detected signal amplitude, Amax and Amin are the global maximum and minimum fringe amplitudes, respectively, λ is the HeNe wavelength (632.8 nm), and δ is phase angle in radians. The time values associated for each peak and valley allow us to graph the displacement, and calculate the velocity, of the shockwave-loaded surface. Because these are discrete time points, a fitting algorithm is incorporated to define the displacement and velocity profiles over the region of peak stress and decay. It has been shown previously [16,17] that the displacement and velocity of the free surfaces can be expressed as functions following the format: where t is time, u0 and v0 represent the free surface displacement and velocity, respectively, and α, β, and γ are scaling constants to fit the function to the measured data. Fitting of Eq. (2) to displacement vs. time plot is accomplished using a damped least-squares method in OriginPro software. Verification of fitting parameters α, β, and γ can be determined by inserting Eq. (2) into Eq. (1) and superimposing the result on the raw waveform. By adjusting the phase δ, correct fitting parameters lead to matched peaks and valleys between the raw waveform and Eq. (1). After fitting parameters are validated, the free surface velocity, Eq. (3), is calculated by taking the derivative of the displacement function, Eq. (2).Based off of one dimensional wave propagation theory, the mechanical stress caused by the shockwave can be calculated using the material’s surface velocity by the following equation:
where σ is the shockwave stress within the material, c is the speed of sound in the material and ρ is the material density [16,17]. Stress is shown negative here because during derivation and solution of the differential equations, tensile stress is considered in the positive direction, and our pulse is in the compressive region [16–19]. From the full stress profile, we record the peak stress value, and the decay time from 90% of max value to the 10% of max value, for statistical analysis. The workflow for this data analysis process is shown graphically in Fig. 4.Fig. 4 Process workflow for deriving the shockwave stress profile from the measured raw waveform. Stress is negative because it is in the compressive region.
2.5 Statistical analysis
To evaluate differences in both the maximum stress generated and decay times observed across all materials, the data was analyzed using a one-way, nested Analysis of Variance (ANOVA) where the calculated max stress and decay times are the dependent variables, and the independent variables are material type and laser fluence. Nesting allows materials to be statistically compared to each other within a given energy density. Multiple comparisons were made using Tukey HSD test to determine where the significance lies between each pair of groups.
3. Results
3.1 Maximum shockwave stress output
For each material type and energy density, six shockwaves were recorded (N = 6) and averaged to obtain an average stress pulse. In general, we see an expected trend of increasing peak stress with increasing laser energy density; however, within PVC, Acetate and PEEK the trend is not as consistent. Polycarbonate shows the highest peak stresses throughout the laser fluences, while PVC has the lowest peak stresses of all the materials, as shown in Fig. 5(a). Looking at the 110.14 mJ/mm2 laser drive density as an example, polycarbonate generated a peak stress of 258.75 (95% CI 194.5-322.7) MPa, and PVC generated a peak stress of 101.74 (95% CI 83.6-119.9) MPa, showing a difference of over 150 MPa.
Fig. 5 (a) Peak stress values as a function of laser energy density. Bars represent 95% confidence interval around mean stress value. *Polycarbonate significantly higher, p ≤ 0.05. **PVC significantly lower, p ≤ 0.05. (b) 90% - 10% decay time as a function of laser energy density. Bars represent 95% confidence interval around the mean decay time.
At the lowest energy fluence value, all materials created shockwaves with similar peak stress values minus PVC. With increasing energy values we see an increase in variability of peak stress values across all materials. Even with high variability, polycarbonate peak stress values prove to be significantly higher than the other materials (p ≤ 0.05), while PVC’s values show to be significantly lower than all other values (p ≤ 0.05).
3.2 Shockwave pulse decay times
Each shockwave pulse decay time was determined by calculating the time for the shockwave stress to decay from 90% to 10% of its maximum value. Results are shown in Fig. 5(b). Unlike the max stress values, there are no definitive trends in shockwave pulse decay times. Glass demonstrates the most consistent low pulse decay times, while PVC generally shows longer decay times except for the 93.49 mJ/mm2 energy density level. Again taking energy density of 110.14 mJ/mm2 as an example, decay times from shockwaves produced with glass is 137.45 (95% CI 105.0-169.9) nanoseconds, while shockwaves produced with PVC had an average decay time of 350.01 (95% CI 161.9-538.14) nanoseconds. Despite these two values being statistically distinct from each other, the glass substrate did not reveal a statistically different decay time than other materials. Individual pairing of material type and energy density show significance, and while there are no complete differences among materials across all energy density levels, there are general trends seen between glass, PVC, and Acetate. All significant findings with p<0.05 are shown in Table 1.
Table 1. Significant Pulse Decay Time Comparisons
4. Discussion
In comparing different flexible substrates to glass for shockwave generation, we found that polycarbonate film consistently produces a higher maximum peak stress from all other materials, and between 20% - 40% higher than soda-lime glass. We also found that PVC consistently generates shockwaves with lower peak stresses than other materials, and 30% - 50% lower than glass. The shockwave pulse decay times were consistently lowest for glass and, in general, highest for PVC. However, because of the variability between individual material trials, and the inconsistent order of lowest to highest decay times between different energy densities, we cannot definitively state which plastic material creates shockwaves with the fastest decay time. All average shockwave profiles for each material is shown in Fig. 6. It is seen qualitatively that glass, polycarbonate and polyester generate shockwaves with a more uniform profile and consistent trend of stress versus laser fluence.
Fig. 6 Average shockwave profiles for all materials at varying energy densities. Plastic thicknesses are 25.4μm; glass thickness is 1mm.
The data showed significant variation within each category of material and energy fluences, even among the standard material of glass. The sample setup is modeled after the clinical application of LGS therapy with standard ultrasound gel as the coupling medium. The heterogeneous composition of polymer-based gel network will cause non-uniform transmission of the shockwaves on the nanometer length scale. This effect shows as dispersion of the shockwave pulse, increasing the temporal width from tens of nanoseconds to hundreds of nanoseconds between the carrier substrate and the reflecting surface. The effect of dispersion for different materials may be influenced by each material’s mechanical properties, however comparison of shockwave pulses to different published material properties, specifically those influencing acoustic transmission, revealed minimal correlation. This is most likely due to the dynamic response of properties such as density, modulus, etc. to very high strain rates, such as those in shockwaves, rather than published values for quasi-static loading. It is well noted in shockwave research that the shock front changes the state of materials and therefore the material properties. The shockwave profiles measured off of the same mirrored glass surface allows for standardized calculation and relative comparisons, removing the need to include unknown properties of the varying plastics and any upstream effects on the shockwaves.
One could argue the same principle will apply to the polymer-based plastic substrates, that they themselves create shockwaves of larger temporal width than the original crystalline-based glass substrate, and that trend is seen in the data with glass have the shortest decay time between all the materials. Both material and ultrasound gel play a role in dispersion of shockwaves, as shown in the comparative graphs in Fig. 7. Direct measurement of shockwaves from polycarbonate showed a larger pulse width than glass by itself. Furthermore, the comparison of direct measurement from glass to the measurement of glass through ultrasound gel showed pulse width increase. These phenomena could come into effect when the shockwave hits the biofilm/tissue interface, and a portion of the compressive wave reflects back through the biofilm as a tensile wave. The beginning of the tensile wave will interfere with the tail of compressive wave as it continues to pass through the interface. The superposition of the tensile wave and the compressive wave must allow for enough tensile stress to delaminate the biofilm. Our current understanding indicates that peak stress and pulse width are the key parameters defining this scenario and efficient biofilm delamination. These effects will be further quantified in future ex vivo experiments.
Fig. 7 Shockwave profiles measured directly measured from polycarbonate film (left) and directly from glass substrate compared to shockwaves after traveling through ultrasound gel (right). Stress profiles have been normalized to reveal dispersion of shockwaves.
It has been noted in literature that while the Michelson interferometer is a simple setup, easy alignment, and straight-forward derivation of displacement from fringe count, there is a significant limitation to the displacement resolution in the nanometer region and indeterminate direction of displacement. Displacement measurements in this experiment ranged from 800 nm to 2 μm, Compensations are usually made on the reference arm of the interferometer to match both beam path-lengths at the photodiode. Background vibration (<1kHz) cause interference instability, however because of the high frequency nature of these shockwaves (200MHz-1GHz), the short time period of data acquisition minimizes effects of low frequency noise. To further increase the sensitivity of the interferometer and reduce effect of noise on signal acquisition, it would be beneficial to take measurements using a quadrature interferometer. Quadrature detection has the ability to discern direction of surface movement, has been shown to increase sensitivity to nanometer displacements, and therefore dynamic range of the system, plus can be less sensitive to noise [20].
While peak shockwave stress was found to be highest in polycarbonate, all materials generated significant shockwaves despite the differences in material properties. Peak stress values from each material are within the same order of magnitude as glass. This allows us the flexibility to use any of the plastic films in this experiment, should the clinical application require optimizing a different parameter besides peak stress.
5. Conclusion
In this work we have characterized five different plastic materials as potential substrates for producing laser-generated shockwaves, and compared those to the previously developed substrate of glass. From the results of peak stress and decay time, polycarbonate is determined to be the most suitable alternative to glass in future ex vivo and in vivo experiments. Experiments will be required to qualitatively verify shockwaves generated through polycarbonate can delaminate bacterial biofilms using the clinical setup.
References and links
1. S. E. Cosgrove, “The relationship between antimicrobial resistance and patient outcomes: mortality, length of hospital stay, and health care costs,” Clin. Infect. Dis. 42(Suppl 2), S82–S89 (2006). [CrossRef] [PubMed]
2. E. Rubinstein, M. Green, M. Modan, P. Amit, L. Bernstein, and A. Rubinstein, “The effects of nosocomial infections on the length and costs of hospital stay,” J. Antimicrob. Chemother. 9(Suppl A), 93–100 (1982). [CrossRef] [PubMed]
3. R. D. Scott II, “The Direct Medical Costs of Healthcare-Associated Infections in U.S. Hospitals and the Benefits of Prevention,” http://www.cdc.gov/hai/pdfs/hai/scott_costpaper.pdf.
4. G. A. James, E. Swogger, R. Wolcott, E. Pulcini, P. Secor, J. Sestrich, J. W. Costerton, and P. S. Stewart, “Biofilms in chronic wounds,” Wound Repair Regen. 16(1), 37–44 (2008). [CrossRef] [PubMed]
5. L. Hall-Stoodley, J. W. Costerton, and P. Stoodley, “Bacterial biofilms: from the Natural environment to infectious diseases,” Nat. Rev. Microbiol. 2(2), 95–108 (2004). [CrossRef] [PubMed]
6. J. W. Costerton, Z. Lewandowski, D. E. Caldwell, D. R. Korber, and H. M. Lappin-Scott, “Microbial Biofilms,” Annu. Rev. Microbiol. 49(1), 711–745 (1995). [CrossRef] [PubMed]
7. C. F. Schierle, M. De la Garza, T. A. Mustoe, and R. D. Galiano, “Staphylococcal biofilms impair wound healing by delaying reepithelialization in a murine cutaneous wound model,” Wound Repair Regen. 17(3), 354–359 (2009). [CrossRef] [PubMed]
8. J. W. Costerton, P. S. Stewart, and E. P. Greenberg, “Bacterial Biofilms: A Common Cause of Persistent Infections,” Science 284(5418), 1318–1322 (1999). [CrossRef] [PubMed]
9. T. F. Mah and G. A. O’Toole, “Mechanisms of biofilm resistance to antimicrobial agents,” Trends Microbiol. 9(1), 34–39 (2001). [CrossRef] [PubMed]
10. G. G. Anderson and G. A. O’Toole, “Innate and induced resistance mechanisms of bacterial biofilms,” Curr. Top. Microbiol. Immunol. 322, 85–105 (2008). [CrossRef] [PubMed]
11. L. C. Argenta and M. J. Morykwas, “Vacuum-assisted closure: a new method for wound control and treatment: clinical experience,” Ann. Plast. Surg. 38(6), 563–577 (1997). [CrossRef] [PubMed]
12. W. G. Pitt and S. A. Ross, “Ultrasound increases the rate of bacterial cell growth,” Biotechnol. Prog. 19(3), 1038–1044 (2003). [CrossRef] [PubMed]
13. Z. D. Taylor, A. Navarro, C. P. Kealey, D. Beenhouwer, D. A. Haake, W. S. Grundfest, and V. Gupta, “Bacterial biofilm disruption using laser generated shockwaves.” in Proceedings of Annual International Conference of the IEEE (Institute of Electrical and Electronics Engineers, Buenos Aires, 2010), pp. 1028–1032. [CrossRef]
14. A. Navarro, Z. D. Taylor, A. Z. Matolek, A. Weltman, V. Ramaprasad, S. Huang, D. O. Beenhouwer, D. A. Haake, V. Gupta, and W. S. Grundfest, “Bacterial biofilm disruption using laser-generated shockwaves,” Proc. SPIE 8214, 82141H (2012).
15. E. Hagerman, J. Shim, V. Gupta, and B. Wu, “Evaluation of laser spallation as a technique for measurement of cell adhesion strength,” J. Biomed. Mater. Res. A 82(4), 852–860 (2007). [CrossRef] [PubMed]
16. V. Gupta, V. Kireev, J. Tian, H. Yoshida, and H. Akahoshi, “Glass-modified stress waves for adhesion measurement of ultra thin films for device applications,” J. Mech. Phys. Solids 51(8), 1395–1412 (2003). [CrossRef]
17. J. Yuan and V. Gupta, “Measurement of interface strength by the modified laser spallation technique. I. Experiment and simulation of the spallation process,” J. Appl. Phys. 74(4), 2388–2396 (1993). [CrossRef]
18. V. Gupta and J. Yuan, “Measurement of interface strength by the modified laser spallation technique. II. Applications to metal/ceramic interfaces,” J. Appl. Phys. 74(4), 2397–2404 (1993). [CrossRef]
19. J. Yuan, V. Gupta, and A. Pronin, “Measurement of interface strength by the modified laser spallation technique. III. Experimental optimization of the stress pulse,” J. Appl. Phys. 74(4), 2405–2410 (1993). [CrossRef]
20. T. Požar, P. Gregorčič, and J. Možina, “A precise and wide-dymanic-range displacement-measuring homodyne quadrature laser interferometer,” Appl. Phys. B 105(3), 575–582 (2011). [CrossRef]
