Photophysical mechanisms of collagen photomodification (CFP) by the use of a 80 MHz, 780 nm femtosecond titanium-sapphire laser were investigated. Our observation that the decrease in collagen second harmonic generation and increase in two-photon autofluorescence intensity occurred primarily at sites where photoproducts were present suggested that the photoproducts may act to facilitate the CFP process. Laser power study of CFP indicated that the efficiency of the process depended on the sixth power of the laser intensity. Furthermore, it was demonstrated that CFP can be used for bending and cutting of collagen fibers and creating 3D patterns within collagen matrix with high precision (~2 μm).
© 2010 OSA
Fabrication of artificial scaffolds with the appropriate biological, mechanical, and structural properties is an important consideration in designing tissues for biomedical applications. For example, tissue engineering scaffolds need to be porous with interconnected channels to allow penetration of cells and nutrients as well as the efficient removal of waste products . Furthermore, factors such as pore shape, size and surface area to volume ratio can affect cell proliferation within the scaffold . Experimentally, it was demonstrated that depending on the tissue to be regenerated, optimum pore sizes can be different, ranging from 5 μm in neovascularization to 100–350 μm in bone regeneration .
A number of methods for preparing porous three-dimensional (3D) tissue engineering scaffolds using natural and synthetic polymers, ceramics, and metals have been described in the literature [2–4]. The main drawbacks of artificial materials are their relatively poor compatibility for cell growth. In contrast, natural materials, such as collagen, glycosaminoglycan, chitin, and chitosan, have the desirable microstructure and molecular properties for cell proliferation, and have been successfully used for repairing nerves, skin, cartilage, and bone . Among these biomaterials, collagen has been widely used for surface modification or synthesis of composite scaffolds with the required biochemical properties . However, it is not trivial to fabricate collagen fibers with the desired cross-section or other structural features, and creating porous 3D scaffolds with the optimal microarchitecture from collagen or other natural materials is a challenge in tissue engineering.
To address this issue, femtosecond (fs) laser ablation emerged as a promising method for spatially-specific processing of biomaterials. Specifically, intense ultrafast laser pulses can induce nonlinear photophysical processes, such as laser induced breakdown and can be used in the fabrication of 3D scaffolds. The underlying mechanism in this phenomenon is the multiphoton-induced plasma formation, which generates cavitation bubbles and destructive shock waves . Ultrafast laser ablation has been successfully applied to micromachining of transparent materials and microsurgery of biological objects such as living cells, neural and other tissues. For example, minimally invasive and precise processing of cornea tissue has been achieved by focusing fs laser sources through a high numerical objective. In the case of cornea, the energy threshold of laser ablation was estimated to be 1 nJ [8,9]. Recently, femtosecond laser ablation has also been applied in fabricating collagen scaffolds for tissue engineering applications . Specifically, patterns of holes, lines and grids were created on the surface of collagen gel. The sustained proliferation of mesenchymal stem cells from rat bone marrow and human fibroblasts seeded within the ablated patterns was also demonstrated . However, plasma mediated tissue ablation has several drawbacks. First, due to the threshold mechanism and strong nonlinear dependence of the ablation effect on excitation intensity, variations in absorption, and the self-focusing of the excitation source can lead to uncontrollable ablation of the target resulting in structures of uncertain dimensions [10–13]. In addition, during laser-induced plasma formation, the destruction of organic molecules at the focus can occur. Photoproducts from of such decomposition may hinder cellular growth. Finally, there are hazards associated with particulate emissions from tissues during plasma formation. Specifically, it was observed that nanometer-scale materials emitted during fs laser ablation of human dermal tissue are considerably smaller (<50 nm) than particles generated with conventional laser sources such as CO2 or Nd:YAG lasers. Therefore, due to the toxic or virulent potential of these highly airborne particles, medical personnel have to use protective equipment to avoid possible health risks .
However, recent advances in femtosecond laser technology have given rise to new approaches that can be utilized for biomaterial processing. Specifically, multiphoton polymerization, modification, or deposition of materials have been demonstrated [14,15]. Another method is the nonlinear, low-density plasma formation . According to this model, there is a wide intensity range (~5 × 1011-5 × 1012 W/cm2) when fs pulses with energies below the optical breakdown threshold are capable of generating free electrons at sufficiently high density to induce photochemical reactions, but insufficient energy to cause thermo-mechanical effects such as particle emission or the production of ionizing radiation. Both experimental data and numerical simulations show that fs laser nanosurgery of cells or tissues is possible by pulse trains of sub-threshold energy at MHz repetition rates, and the underlying mechanism of this technique is the chemical decomposition of molecular bonds by fs laser induced free-electrons which is not accompanied by thermal effects . Collagen photomodification of using illumination intensity below the optical breakdown threshold has been studied  and applied for tissue microprocessing . The effect was induced and simultaneously monitored by multiphoton imaging (MPI). By the use of 780 nm fs, 80 MHz laser with pulse width of around 120 fs, it was shown that degradation of collagen fiber structure was possible with pulse energy in the 0.1-0.5 nJ range.
In the present study, the dependence of collagen fiber photomodification (CFP) rate on laser intensity was investigated in an effort to elucidate the initial photophysical mechanisms of this process. We also applied this approach in spatially specific microprocessing of different collagen-containing tissues of tendon, cartilage, and skin. Specifically, cutting, bending of collagen fibers as well as for engraving 3D patterns inside the collagen matrix were performed.
2. Materials and method
Simultaneous fs laser photomodification and MPI were performed using a laser scanning microscope system (LSM 510 META, Zeiss, Jena, Germany) coupled to a fs, titanium:sapphire laser operating at 780 nm with a pulse repetition frequency of 80 MHz (Tsunami, Spectra Physics, Mountain View, CA). Average power of the fs laser on the sample surface (P) was between 6 and 60 mW and the laser pulse width was estimated to be 120 fs. The detection bandwidths of the second harmonic generation (SHG) and two-photon autofluorescence (TPA) signals were 380-400 nm and 435-700 nm, respectively. Effects of fs laser photomodification were induced and observed using long working distance (WD) air objective (Plan-Neofluar 20x/NA 0.5, WD 1.3 mm) and an oil immersion objective (Fluar 40x/NA 1.3 Oil, WD 230 μm). Pixel dwell time in imaging/modification process was 6.4 μs and the frame (256 × 256 pixels) rate was 2.1 Hz. The collagen fiber specimens included dry, type I collagen from rat tail tendon (Fluka Biochemika, Chemie GmbH CH-9471 Buchs, 27666, UK) and from bovine Achilles’ tendon (Fluka, 27662, Switzerland). In addition, collagen-containing air-dried samples from bovine leg tendon and chicken leg tendon, as well as wet samples from chicken skin and leg bone cartilage, were used. The experiments on fs photomodification and MPI were conducted at the ambient temperature of 18° C. MPI results were analyzed with Origin 7.5 (OriginLab Corporation, Northhampton, MA) and ImageJ.
3.1. Mechanisms of femtosecond photomodification in collagen fiber
Shown in Fig. 1 are time-lapsed multiphoton images performed on a single collagen fiber from rat tail tendon (RTT) and collagen network from bovine Achilles’ tendon (BAT). The CFP effect was induced by fs laser with moderate mean powers (P = 16 and 30 mW) and the simultaneous MPI allowed recording of the CFP process. As can be seen in Fig. 1A, SHG signal was high in the initial phase of scanning (1-10 sec). However, with increased illumination, one or more centers of damaged sites became noticeable at 10 and 25 seconds (indicated by arrows in Fig. 1A). During the CFP process, a reduction of the SHG signal and simultaneous increase in the TPA intensity were observed at the damaged sites.
The decrease in SHG intensity indicates the disruption of the non-centrosymmetric structure of collagen fiber needed for the effective production of the SHG signal. Increase in the longer wavelength, TPA signal can be associated with formation of new photoproducts, most likely due to tyrosine dimerization [17,19]. Upon further illumination, fs laser induced structural photomodification started at initial CFP sites and then extended along the collagen fiber resulting in increase of the sizes of the photomodification regions. Finally, the TPA signal of the photoproduct vanished (36 sec image, Fig. 1A) which may be interpreted as photobleaching of the photoproducts. Similar observations were found in the BAT specimen (Fig. 1B).
In addition to qualitative imaging, we also studied the kinetics of the CFP process. Shown in Fig. 2 are the time-lapsed MPI of bovine leg tendon specimens and the corresponding kinetic measurements of SHG and TPA signals at different excitation laser intensities. In this case, we used air-dried samples to avoid an increase in SHG intensity in the initial phase of the CFP due to dehydration . In addition, we calculated the temporal derivative of SHG intensity and the analysis of these curves showed several interesting features. First, the maximum rate of CFP was observed near the moment when SHG intensity decreased to 50% of its initial level. Therefore, this time was designated as the half photomodification time t ½. We also noted that increase in illumination intensity by 1.44 (from P = 24 mW to P = 34.5 mW) dramatically reduced t ½ by 8.5 times (from 6.8 min to 0.8 min) (see Fig. 2B). These results suggested that the CFP process was not linear to the laser intensity and that multiple photon absorption was the responsible photomodification mechanism. In addition, we performed an experiment to investigate the kinetics of CFP in two nearby regions in dried bovine leg tendon, where the scanned areas differed by a factor of two (11.5 × 11.5 μm2 and 16.5 × 16.5 μm2) (Fig. 3 ). We found that t ½ increased from 2 to 4.05 mins when the illumination area was doubled. This observation means that the CFP effect was proportional to the laser pulse number which had fallen onto the scanning area.
Further insight of the nature of CFP can be gained from an experiment in which time lapses were introduced during the photomodification process. Specifically, fs laser illumination was initially applied to dried chicken leg tendon. However, at time points of t1 = 276 sec and t2 = 296 sec, laser illumination was respectively blocked for 15 and 10 seconds (Fig. 4A ). As Fig. 4B shows no significant change in SHG and TPA intensities was observed after laser illumination restarted. This and similar experiments with longer pauses in illumination or with illumination by continuous wave laser  showed that collagen structural modification was irreversible and that the CFP only occurred when fs laser was used. Furthermore, the photoproducts were stable in the absence of fs pulses.
These observations suggest that within the temporal resolution of our experiments, thermal effects to collagen modifications can be neglected. Furthermore, in an earlier study, it was found that during collagen thermal denaturation, uniform decrease in both SHG and TPA intensities was observed across the entire frame, and decrease in TPA intensity with the temperature rise was associated with the thermal decomposition of fluorophores . As opposed to thermal treatment, in fs laser CFP, an increase in TPA intensity was only observed at the photomodified sites. These results indicated that during fs laser CFP, one can ignore heat diffusion and analyze SHG intensity decay without consideration thermal effects.
Since SHG microscopy can be used in the quantitative analysis of collagen [20,21], we attempted to use SHG and TPA changes in constructing a photophysical model for the CFP process. In this effort, we define the two quantities ΔI(t)SHG = ISHG(max)-ISHG(t) and ΔITPA(t) = ITPA(t)-ITPA(min). In these definitions, ITPA(min) refers to the initial level of TPA in native collagen and ITPA(t) is the TPA intensity at time t. Accordingly, ΔITPA (t) is the TPA intensity of photoproducts as a function of time and is related to the photoproduct concentration. Similarly, ISHG(max) and ISHG(t) correspond to the SHG intensity from the collagen fibers at the initial time point and a later time t. Therefore, ΔISHG(t) is a measure of the amount of photomodified collagen. To see how these definitions can be used to analyze SHG and TPA signal changes during the CFP process, shown in Fig. 5A are plots of SHG and TPA signal changes from fs laser illumination of dry sample from bovine leg tendon. The SHG and TPA plots in Fig. 5A show, respectively, the characteristic decrease and increase from illumination time. Furthermore, we found that during the initial phase of the CFP process, ΔITPA(t) can be fitted with an exponential function of time exp(Kt) where K is the rate of photoproduct formation (Fig. 5A). In addition, the plot of ΔISHG(t) versus ΔITPA(t) for the period before ITPA(t) saturation (1.5-2.2 minutes) shows that ΔISHG(t) was proportional to ΔITPA(t) (Fig. 5B), and this observation can be expressed as ISHG(t) + αΔITPA(t) = constant, or C(t) + A(t) = N, where C(t) = ISHG(t) is a measure of the amount of unmodified collagen fibers, A(t) = αΔITPA(t) is a measure of the CFP photoproduct present, and N = ISHG(max) is a measure of the initial amount of collagen before the CFP process.
Since the initial change of ΔITPA(t) is exponential in form, A(t) = αΔITPA(t) = A0exp(Kt) and A0 is the proportionality factor. Therefore, we have dA/dt = -dC/dt = KA0exp(Kt) = KA(t). This result indicated that CFP rate was proportional to photoproduct concentration in the initial phase. Similar conclusion can be made from qualitative analysis of the CFP process in RTT, BAT and bovine leg tendon (Figs. 1, 2). Specifically, it was found that collagen photomodification and photoproduct formation progressed rapidly at the sites, where the TPA signal is present. These observations suggest that the CFP rate is proportional to the amount of autofluorescent photoproduct already present at the photomodification site. Therefore, CFP rate depends on both the concentration of native collagen and the amount of photoproduct present. This conclusion can be described using the following equations:
For the simplicity, we set N as 1 and ln(A0/(N-A0)) as D, and used Eq. (2) to fit for ISHG(t) of dried samples from bovine leg tendon, and chicken leg tendon. As a result, k and k/D were determined and the dependences of ln(k) and ln(k/D) vs. lnP for bovine leg tendon are plotted in Fig. 6 . Linear fitting of the plots resulted in the slopes of 5.9 ± 0.3 and 5.7 ± 0.3 for ln(k) and ln(k/D) series respectively. For chicken leg tendon the corresponding slopes were 5.6 ± 0.3 and 5.5 ± 0.4. These results indicated that the efficiency of the CFP process depends is approximately to the sixth power of the laser intensity.
3.2. Application of collagen femtosecond photomodification
In addition to studying the CFP mechanisms, we also investigated the different modes of collagen fiber manipulation possible with this approach. Examples of controlled photomodification and manipulation of collagen fibers are shown in Figs. 7 and 8 . Figure 7 shows the cutting of single collagen fibers in papillary dermis of wet chicken skin.
An interesting feature was observed when the illuminated collagen fiber started to be modified. It was found that the collagen fiber became flexible and can be bent at the illuminated site. The bending process can be observed through the microscope eyepiece as this process can be controlled in real time. Shown in Fig. 8 are collagen fibers which were bent at the CFP sites due to short-term illumination (6-10 sec). Therefore, short-term exposure to fs laser with moderate power allowed softening and bending of collagen fibers with the thickness up to the 10 μm range.
3.3. 3D photomodification of collagen scaffolds
In addition to cutting and bending of collagen fibers under moderate fs laser illumination, longer illumination can be used for the patterning of 3D collagen matrix. To demonstrate this possibility, we repeatedly scanned selected regions of interest inside a wet sample from chicken leg bone cartilage tissues (Fig. 9 -11 ). Due to the nonlinear mechanism of photomodification, only the collagen fiber in the laser focus was exposed to sufficient light intensity for photomodification. Figure 9 demonstrates the capability of the nonlinear approach to fabricate interconnected channels with widths less than 2 μm in the collagen tissue. In the case of short term illumination, traces of photoproducts were seen through the detection of TPA (Fig. 9B).
Shown in Fig. 10 is the MPI of horizontal rectangular (1 × 40 μm2) pattern created in chicken bone cartilage. Following the creation of the pattern, we added rhodamine B (RB) solution onto the sample. Note that the RB solution flowed and was uniformly distributed inside the fs laser modified region (from −0.6 to + 0.6 μm, Fig. 10B, series 1). Moreover, in this region, SHG signal was absent (series 2) and two-photon fluorescence of RB solution had the same intensity as in the site out of cartilage tissue (series 3). Therefore, Fig. 10 clearly demonstrates that fs laser illumination was capable of both photomodification of collagen fibers and removal of photoproducts and other tissue components from the illuminated region. However, on the edges of the illuminated regions (Fig. 10B), both SHG and two photon fluorescence signals were weak and a black thin (~0.7 μm width) strips were observed. Most likely, these strips were occupied by photomodified tissues that were not capable of producing SHG and also did not allow the permeation of the RB solution.
Finally, we demonstrated the creation of 3D patterns by fs laser illumination in wet chicken bone cartilage tissue. Two cross patterns were engraved at the depths of 7.8 and 20 μm, and an additional cross pattern oriented at 45° from the other two cross patterns was engraved at the depth of 15 μm inside the sample. Scan time at each depth was 1.6 min. Axial profiles of SHG intensity in two regions of interest (ROIs, white rectangles in Fig. 11A) show that the average SHG intensity in a non-illuminated area was about 3 a.u., and full widths of axial profiles defined at 0.1 level of mean intensity were found to be 2.3, 2.7 and 3.2 μm at the depths of 7.8, 15 and 20 μm respectively. Therefore, our result shows that the region affected by the CFP process became larger with increasing depths: larger by ~1 μm in the x-y plane (see Fig. 10) and by 1.2 μm (at 7.8 μm depth) and 1.6 μm (at 20 μm depth) in the axial direction (Fig. 11). This observation is likely due to the depth-dependent broadening of point-spread-function caused by refractive index mismatch induced spherical aberration [13,22].
In this study, we investigated the fs-laser collagen photomodification (CFP) process at laser intensity below the optical breakdown threshold. The process accompanied with decrease of SHG signal and TPA increase at the damaged sites. We found that the modification efficiency was proportional to the laser pulse numbers, which had fallen onto the scanning area (Fig. 3) . We also found that during interruption of laser illumination, the CFP process seemed to cease (Fig. 4), suggesting that the CFP process was instantaneous and non-thermal. Furthermore, our observation of emanation of the CFP process from sites where autofluorescent photoproducts were present suggested that the CFP process was dependent on the presence of such products. Based on these results, we developed a kinetic model to describe the CFP process and found that our model could be used to describe the kinetics of collagen SHG. In addition, power study indicates that k, the CFP rate constant, depends on the sixth power on laser intensity.
The absorption of multiple photons can result in the photoionization of water and organic molecules, and subsequent chemical reactions became possible in the submicron focal region [22–28]. In particular, efficiency and pathways of laser photolysis of aromatic/aliphatic amino acids and related dipeptides were studied and it was shown that the absorption of one or more photons led to the destruction of amino acids and peptide bonds resulting in photo-degradation of collagen chains. The cleavage of side groups, formation of collagen crosslinks, conformational change with concomitant chain degradation, and suppression of fibril formation were processes that contributed to photo-degradation. Specifically, it was measured that quantum yield (QY) of decomposition of aromatic amino acids were 0.009-0.022 for one photon excitation at 254 nm, about 0.1 for 193 nm excitation, and 0.11-0.19 for two-step (biphotonic) picosecond laser excitation at 266 nm. Furthermore, the QY of peptide bond scission between glycine and proline, the two most widespread residues in collagen, was 0.059 at 193 nm excitation. These results indicate that for effective photomodification of collagen structure, direct absorption of 5 or more photons of our 780 nm laser source was enough for effective ionization of amino acids or scission of the peptide bonds and that such a process may contribute to the CFP process we observed. However, our observation of the likely involvement of the autofluorescent photoproducts in CFP suggests that in the case of MHz fs-laser illumination, two-photon absorption of the photoproduct followed by resonant absorption of several photons may contribute to increase in localized energy deposition up to photoionization energy or effective energy/charge transfer to the next, undamaged collagen molecules and contribute to the CFP process. Additional experiments are needed to clarify the exact mechanisms of collagen femtosecond photomodification.
Nonlinear illumination by 80 MHz fs laser can be used for processing 3D spatial features with precision less than 2 μm in collagen matrix. We demonstrated that moderate illumination power can be used to modulate collagen architecture such as cutting, bending, and the creation of spatially specific patterns in three-dimensions. However, in-depth CFP is sensitive to refractive index mismatch induced spherical aberration which can lead to point-spread-function broadening and degraded resolution of the created features. These results indicate that in case of 3D CFP, a careful selection of illumination parameters need to be determined in order to optimize the creation of three-dimensionally resolved features for biomedical applications. With additional development, the approach demonstrated in this work may be combined with fs laser ablation to become a scaffold manufacturing technique which may lead to the creation of tissue engineering scaffolds for cell proliferation and may be extended to medical procedures such as cartilage reshaping, skin rejuvenation, and intraocular surgery.
We would like to acknowledge the financial support received from National Science Council in Taiwan (grant number: NSC 98-2112-M-002-008-M43) for the support of this work.
References and links
1. L. G. Cima, J. P. Vacanti, C. Vacanti, D. Ingber, D. Mooney, and R. Langer, “Tissue engineering by cell transplantation using degradable polymer substrates,” J. Biomech. Eng. 113(2), 143–151 (1991). [CrossRef] [PubMed]
3. J. J. Klawitter and S. F. Hulbert, “Application of porous ceramics for the attachment of load bearing internal orthopedic applications,” Biomed. Mater. Symp. 5(6), 161–229 (1971). [CrossRef]
7. V. Venugopalan, A. Guerra 3rd, K. Nahen, and A. Vogel, “Role of laser-induced plasma formation in pulsed cellular microsurgery and micromanipulation,” Phys. Rev. Lett. 88(7), 078103 (2002). [CrossRef] [PubMed]
8. T. Juhasz, F. H. Loesel, R. M. Kurtz, C. Horvath, J. F. Bille, and G. Mourou, “Corneal refractive surgery with femtosecond lasers,” IEEE J. Quantum Electron. 5(4), 902–910 (1999). [CrossRef]
9. K. Koenig, O. Krauss, and I. Riemann, “Intratissue surgery with 80 MHz nanojoule femtosecond laser pulses in the near infrared,” Opt. Express 10(3), 171–176 (2002). [PubMed]
11. J. Bunte, S. Barcikowski, T. Puester, T. Burmester, M. Brose, and T. Ludwig, “Secondary hazards: Particle and X-ray emission,” In: Femtosecond Technology for Technical and Medical Applications. Topics Appl Phys96, F. Dausinger, F. Lichtner, H. Lubatschowski, editors (Springer, Berlin, 2004), p. 309–321.
13. M. Oujja, S. Pérez, E. Fadeeva, J. Koch, B. N. Chichkov, and M. Castillejo, “Three dimensional microstructuring of biopolymers by femtosecond laser irradiation,” Appl. Phys. Lett. 95(26), 263703 (2009). [CrossRef]
14. L. P. Cunningham, M. P. Veilleux, and P. J. Campagnola, “Freeform multiphoton excited microfabrication for biological applications using a rapid prototyping CAD-based approach,” Opt. Express 14(19), 8613–8621 (2006). [CrossRef] [PubMed]
15. S. Maruo and J. T. Fourkas, “Recent progress in multiphoton microfabrication,” Laser Photonics Rev. 2(1-2), 100–111 (2008). [CrossRef]
16. A. Vogel, J. Noack, G. Huttman, and G. Paltauf, “Mechanisms of femtosecond laser nanosurgery of cells and tissues,” Appl. Phys. B 81(8), 1015–1047 (2005). [CrossRef]
18. V. Hovhannisyan, W. Lo, C. Hu, S. J. Chen, and C. Y. Dong, “Non-ablative processing of biofibers by femtosecond IR laser,” Proc. SPIE-OSA 7373, 73731X1–7 (2009).
19. M. Wisniewski, A. Sionkowskaa, H. Kaczmarek, S. Lazare, V. Tokarev, and C. Belin, “Spectroscopic study of a KrF excimer laser treated surface of the thin collagen films,” J. Photochem. Photobiol., A 188(2-3), 192–199 (2007). [CrossRef]
20. E. Brown, T. McKee, E. diTomaso, A. Pluen, B. Seed, Y. Boucher, and R. K. Jain, “Dynamic imaging of collagen and its modulation in tumors in vivo using second-harmonic generation,” Nat. Med. 9(6), 796–801 (2003). [CrossRef] [PubMed]
21. V. A. Hovhannisyan, P. -J. Su, and C. -Y. Dong, “Quantifying thermodynamics of collagen thermal denaturation by second harmonic generation imaging,” Appl. Phys. Lett. 94(23), 233902 (2009). [CrossRef]
22. C. Y. Hsiao, Y. Sun, W. L. Chen, C. K. Tung, W. Lo, J. W. Su, S. J. Lin, S. H. Jee, G. J. Jan, and C. Y. Dong, “Effects of different immersion media in multiphoton imaging of the epithelium and dermis of human skin,” Microsc. Res. Tech. 69(12), 992–997 (2006). [CrossRef] [PubMed]
23. N. I. Kiskin, R. Chillingworth, J. A. McCray, D. Piston, and D. Ogden, “The efficiency of two-photon photolysis of a “caged” fluorophore, o-1-(2-nitrophenyl)ethylpyranine, in relation to photodamage of synaptic terminals,” Eur. Biophys. J. 30(8), 588–604 (2002). [CrossRef] [PubMed]
24. R. A. Meldrum, S. W. Botchway, C. W. Wharton, and G. J. Hirst, “Nanoscale spatial induction of ultraviolet photoproducts in cellular DNA by three-photon near-infrared absorption,” EMBO Rep. 4(12), 1144–1149 (2003). [CrossRef] [PubMed]
25. A. A. Oraevsky, D. B. Da Silva, A. M. Rubenchik, M. D. Feit, M. E. Glinsky, M. D. Perry, B. M. Mammini, W. Small, and B. C. Stuart, “Plasma mediated ablation of biological tissues with nanosecond-to femtosecond laser pulses: relative role of linear and nonlinear absorption,” IEEE J. Sel. Top. Quantum Electron. 2(4), 801–809 (1996). [CrossRef]
26. D. N. Nikogosyan and H. Gorner, “Laser-induced photodecomposition of amino acids and peptides: extrapolation to corneal collagen,” IEEE J. Sel. Top. Quantum Electron. 5(4), 1107–1115 (1999). [CrossRef]
27. S. Gaspard, M. Forster, C. Huber, C. Zafiu, G. Trettenhahn, W. Kautek, and M. Castillejo, “Femtosecond laser processing of biopolymers at high repetition rate,” Phys. Chem. Chem. Phys. 10(40), 6174–6181 (2008). [CrossRef] [PubMed]