We demonstrate in situ and real time characterization of two-photon polymerization (TPP) by means of broadband coherent anti-Stokes Raman scattering (CARS) microscopy. The same experimental setup based on one femtosecond oscillator is used to perform both TPP and broadband CARS microscopy. We performed in situ imaging with chemical specificity of three-dimensional microstructures fabricated by TPP, and successfully followed the writing process in real time. Broadband CARS microscopy allowed discerning between polymerized and unpolymerized material. Imaging with good vibrational contrast is achieved without causing any damage to the microstructures or undesired polymerization within the sample.
©2010 Optical Society of America
Over the past decade, the fabrication of three-dimensional microstructures using two-photon polymerization (TPP) has experienced a rapid growth . Several research fields are indeed benefiting from the ability of TPP to create complex patterns with resolution less than 100 nm. For example, three-dimensional microstructures created by TPP are successfully employed in nanophotonics, microelectronics, microfluidics, microelectromechanical systems, and bioengineering [2–6]. In these applications, TPP is either the only method capable of creating the sought after three-dimensional objects or the most time and cost effective way of producing them.
The main characteristic that renders TPP unique in making three-dimensional microstructures is the confinement of matter transformation within the focal volume of a focused laser beam . This phenomenon is based upon the nonlinear optical interaction between the sample and the excitation field, and it permits to localize two-photon absorption and subsequent polymerization in volume elements (voxels) as small as fractions of a femtoliter. Complex three-dimensional microstructures can be formed by carefully overlapping polymerized voxels in predetermined trajectories. In a recent report it was demonstrated that electron generation and hence chemical bond breaking in TPP is mostly caused by avalanche absorption . Nonlinear absorption then plays the role of generating “seed” electrons for subsequent electronic excitations. This mechanism describing the initial stages of TPP, which has been tested on a sol-gel based photocurable material using low repetition rate fs lasers and high irradiance excitation, would explain why so many resins can be structured by TPP using NIR photons even when their two-photon cross-sections are usually negligible in this region of the spectrum.
Although TPP has become an enabling technology for several applications, there are still some fundamental aspects that have not yet been fully characterized. One of these aspects is the influence of the scanning pattern on the mechanical properties of the microstructure [9,10]. Even if the same object can be realized utilizing different writing strategies, some approaches are better than others in producing accurate and dimensionally robust microstructures. The inevitable shrinkage that occurs during the transition from liquid to solid upon polymerization promotes buildup of internal stresses . The judicious choice of scanning patterns can minimize or offset this negative effect, resulting in the production of microstructures of high fidelity. Another aspect of TPP that requires a better understanding is the influence of the developing process on polymer swelling and solvent inclusion which have a large impact on the structural integrity of the microstructures. Finally, with ever increasing writing resolution [12,13], deeper insight into the differences between the material properties of the bulk polymer and those of the microstructures due to size effects is required [14,15].
We have recently demonstrated that coherent anti-Stokes Raman scattering (CARS) microscopy is an ideal optical technique for probing TPP from the aforementioned point of views [16–18]. CARS microscopy provides in fact imaging of polymeric microstructures with molecular selectivity and a sufficiently high spatial resolution. We have shown that when acquiring signals generated by carbon-hydrogen stretching modes (~3000 cm−1) of polymers, imaging by CARS microscopy provides qualitative information on the density heterogeneities of the microstructure. We have shown also that quantitative details on the degree of polymers cross-linking can be inferred by CARS microscopy when acquiring signals generated by vibrations of chemical bonds that participate in the polymerization process, such as the carbon-carbon double bonds in acrylate monomers (~1640 cm−1).
In CARS microscopy, two laser beams are focused in the sample, a pump beam at a frequency ωp and a Stokes beam at a frequency ωs. Vibrational sensitivity is obtained when the difference frequency ωp - ωs matches the frequency of a Raman active vibrational mode of the sample. The anti-Stokes signal at ωas = 2ωp - ωs is coherently driven by the two laser beams, resulting in coherent radiation [19–22]. In our previous reports, we employed a picosecond laser and a synchronously pumped OPO as sources of pump and Stokes beams. To acquire images in different spectral regions of the Raman spectrum of the polymeric microstructures, the output wavelengths of the OPO was adjusted accordingly.
Although CARS microscopy by means of narrowband excitation fields permits imaging with high spectral resolution, it has the drawback of requiring different sources of light than those needed for TPP. It would be desirable to perform both TPP and CARS microscopy utilizing the same experimental setup. In this way, microstructures fabricated by TPP can be monitored in situ and ultimately in real time during microfabrication.
To explore this possibility, we investigate the use of broadband CARS microscopy [23–30] for the characterization of TPP. In broadband CARS microscopy the spectral shape of the Stokes and pump beams are broad and narrow, respectively. Application of the pump and Stokes beams simultaneously excites multiple Raman transitions within the bandwidth of the Stokes pulse. Thus, broadband CARS microscopy allows gathering of a large section of the vibrational spectrum of the sample at once. Furthermore, it has been demonstrated that imaging by broadband microscopy can be efficiently performed on biological and non biological samples using a single Ti:sapphire oscillator [29–32]. Among a variety of approaches to perform broadband CARS microscopy, we choose the method based on a single Ti:sapphire oscillator and supercontinuum generated in a photonic crystal fiber [25–29,33–36].
Integrating TPP and broadband CARS microscopy into one experimental setup that shares the same source of light provides the advantage of rapidly investigating TPP fabricated microstructures in situ, without the need to move the sample to another instrument. In addition, this approach makes possible monitoring TPP in real time.
2. Experimental methods
The experimental setup employed to perform TPP and CARS microscopy is schematically illustrated in Fig. 1 . The laser light source is a commercial Ti:sapphire laser (MaiTai DeepSeeTM, Spectra-Physics) that produces pulses with a repetition rate of 80 MHz, pulse duration of 100 fs, and a center wavelength that can be tuned from 690 nm to 1040 nm. All experiments are performed using an excitation wavelength at 800 nm where the laser average output power exceeds 3 Watts. By means of a beam splitter, the laser beam is divided into two optical paths. The optical path with the transmitted beam is used in the wavelength extension unit (WEU, Newport Corp.) where it is divided once again to produce the narrow and broadband laser beams that act respectively as pump and Stokes beams for generating CARS signal in the sample. The pump and Stokes beams exit the WEU in a collinear fashion. The polarization of both beams is vertical, and their relative time delays can be varied with a delay line within the WEU.
The laser beam reflected off the beam splitter bypasses the WEU and goes through a mechanical shutter, and a combination half-wave plate/polarizer serving as a variable attenuator. This beam maintains temporal and spectral characteristics of the oscillator and performs TPP. Since its polarization is horizontal, the laser beam for TPP is recombined with the pump and Stokes beams using a polarizing cube beam splitter that transmits vertical and reflects horizontally polarized beams. The three beams, at this point traveling collinearly, are launched into a home-built laser scanning microscope . In order to perform CARS microscopy while TPP is taking place, the polarizing cube beam splitter is positioned between the galvanometric mirrors and the scan lens. Thus, while the pump and Stokes beams can be rapidly scanned in a raster mode at the focal plane in the sample, the TPP beam is set stationary in the middle of the objective lens (40x 0.75NA) field of view.
The sample rests on a xyz computer-controlled stage assembly. Microfabrication by TPP is realized using software written in LabView that controls the movements of the stages and the action of the mechanical shutter. The CARS signal generated at the sample is collected with a condenser lens (0.55 NA) which is positioned right after the stage assembly. Thus, all CARS images presented in this work were acquired by recording the forward directed signal only. After exiting the condenser, the signal goes through a set of filters to remove the excitation beams, and by means of a lens it is focused on the active area of a PMT detector. Images are acquired by a computer running MPScope software . The CARS signal can also be monitored by a spectrometer with the aid of a mirror on a flip mount.
A schematic diagram of the WEU is shown in Fig. 2(a) . The laser beam is first divided into two by means of a beam splitter. The reflected beam is sent into an optical Faraday isolator to eliminate feedback into the laser and then launched into a photonic crystal fiber (PCF) (SCG-800-CARS, Newport Corp.). A broad supercontinuum ranging from below 700 nm to over 1100 nm is generated. The near-infrared portion of the supercontinuum (> 800 nm) is selected with a long pass filter and it is used as the Stokes beam. The transmitted part of the input laser beam is spectrally narrowed to 3 nm with a bandpass filter and is used as the pump beam. Thus, the spectral resolution of the CARS microscope defined by the bandwidth of the pump beam is 50 cm−1. The pump and Stokes beam are recombined and spatially overlapped using a long pass filter. To ensure that the two beams overlap in the time domain as well, a delay line is inserted into the pump beam path prior to being recombined with the Stokes beam. Both beams in the WEU box have variable attenuators to allow independent control of laser intensities and polarizations. Typical average powers for the pump and Stokes beams measured after they exit the WEU are 400 mW and 30 mW, respectively. The reported average power of the Stokes beam relates to a portion of the supercontinuum that goes from 1010 nm to 1070 nm. The average powers for the Stokes beam addressed in this work hereafter refer to the portion of the supercontinuum within this window only.
The pump and Stokes beams spectra are shown in Fig. 2(b). A narrow peak centered at 800 nm is observed. This is the residual input beam from the oscillator that has gone through the 3 nm bandpass filter. At wavelengths shorter than 800 nm no signals are measured because of the long pass filters employed in the WEU. At wavelengths longer than 800 nm, a broad spectrum originating from the PCF is observed with a peak corresponding to a Stokes shift at around 3000 cm−1.
The details of microfabrication by TPP are described elsewhere . Briefly, it is accomplished by using an acrylic-based resin. It consists of a photoinitiator and two monomers. The photoinitiator is Irgacure 369 (Ciba Specialty Chemicals) and it is used at a concentration of 3% by weight. The remainder of the resin is equally divided between two trifunctional acrylic monomers, SR499 and SR368 (Sartomer Inc.). The three components are mixed at room temperature until forming a homogenous viscous liquid which can be applied onto microscope cover slips by either spin or drop casting.
3. Results and discussions
3.1 Signal characterization
The Raman spectrum of the polymerized resin in the frequency region between 2000 cm−1 and 3000 cm−1 is shown in Fig. 3(a) (red line). A broad and structured peak centered at around 2900 cm−1 is observed. This peak corresponds to the stretching modes of aliphatic and aromatic carbon-hydrogen bonds that are abundantly present in the resin. The CARS spectrum of the polymerized resin is also shown in Fig. 3(a) (blue line). This spectrum was recorded using a bandpass filter (650 ± 50 nm) in front of the detector and was not corrected by normalizing it to the signal generated in the glass substrate. The profile of the CARS spectrum shows two typical characteristics of CARS signals. One is that its spectrum is slightly broader than the Raman spectrum. The other is that the maximum peak of the CARS spectrum is red-shifted relative to maximum peak of the Raman spectrum. These features derive from the composite nature of the CARS signal. In addition to a contribution from a resonant term, CARS signals always contain a nonresonant background that is independent of the Raman shift . Interference between the resonant and nonresonant components of CARS signals give rise to the characteristic broad and red-shifted spectrum profiles. Although the experimental setup resolution (~50 cm−1) in the CARS spectrum does not permit to discern the fine structure of the Raman signal, the CARS and Raman spectra overlap quite well.
A quantitative way to confirm that the collected signal stems from a CARS process is to measure the signal dependence on the intensity of the excitation fields. Theory predicts that the CARS signal varies linearly with the Stokes beam intensity, while it exhibits quadratic dependence on the pump beam intensity (ICARS ∝ I2pump IStokes). To verify this relationship, we imaged by CARS microscopy a microstructure fabricated by TPP and measured its signal as a function of pump and Stokes beam intensities. The sample consisted of a rectangular shaped slab 7 μm wide, 35 μm long and 20 μm thick attached to a glass substrate. The image plane was set 15 μm within the sample to avoid any contribution to the signal from the substrate. Furthermore, in this study, the unpolymerized portion of the resin was washed away before imaging. The results of our investigation are shown in Fig. 3(b). A CARS image of the microstructure used in this experiment is shown in false color in the inset in Fig. 3(b). Within the error of our measurement, we observe no variation from the expected trends.
3.2 Imaging with chemical contrast
One of the main advantages of CARS microscopy over other laser scanning imaging techniques is the ability to create images based on chemical contrast without the aid of extrinsic dyes. The signal is generated directly from the interaction of the excitation fields with specific Raman active vibration modes within the sample. Unfortunately, images collected by CARS microscopy include contributions from both resonant and nonresonant signal. Since the latter is not chemical specific, it is desirable to remove the nonresonant contribution. In this way, images with pure chemical information can be gathered.
To accomplish imaging with chemical contrast, we performed broadband CARS microscopy, consecutively using two bandpass filters in front of the detector one at a time. Both filters are 10 nm wide; one centered at 650 nm (A), the other centered at 670 nm (B). Under the experimental conditions described in section 2, these filters pass signals that derive from Stokes shifts ranging from 3000 cm−1 to 2750 cm−1 for filter A, and from 2520 cm−1 to 2300 cm−1 for filter B. Figure 4(a) shows superimposed transmission windows of filters A and B on the CARS spectrum of the polymerized resin. While filter A overlaps with the polymer CARS signal centered at 2900 cm−1, filter B is in a region of the polymer spectrum where no specific Raman active modes contribute to the polymer CARS signal.
The sample used in this study consists of a microstructure fabricated by TPP having the shape of a cross with both the horizontal and vertical sections 70 μm long and 10 μm wide. An image of the microstructure recorded by scanning electron microscopy (SEM) is shown in Fig. 4(b). The microstructure was built 30 μm tall to permit imaging by CARS microscopy at a plane above the glass substrate interface. Imaging was performed on the sample after washing away the surrounding unpolymerized resin.
Figure 4(c) shows the CARS image of the microstructure recorded using filter A. The signal measured in this image is the sum of both the resonant and nonresonant contributions. Figure 4(d) shows the nonresonant part of the signal, independently recorded using filter B. The nonresonant signal in Fig. 4(d) originates only from the polymer that constitutes the microstructure. Since excitation occurs far enough away from the interface, no nonresonant signal is contributed by the glass substrate. The images shown in Figs 4(c) and 4(d) are normalized to the signal from the glass substrate obtained under same excitation conditions.
An image with pure chemical contrast can be obtained by subtracting the image in Fig. 4(d) from the image in Fig. 4(c). The result is shown in Fig. 4(e). The ratio of resonant to nonresonant signal is 2:1. The contrast observed in Fig. 4(e) depends exclusively on differences in the resonant contributions to the CARS signal. The discontinuities in CARS signal present in Fig. 4(e) are directly correlated to the localized density of the material. The denser the material, the higher the concentration of carbon-hydrogen bonds. The periodic structure seen in Fig. 4(e) is a result of the overlapping sample scanning pattern chosen to create the microstructure. In particular, the periodic structure is proportional to the spacing between the single laser passes employed for TPP microfabrication.
3.3 In situ imaging
In situ imaging of polymeric microstructures by broadband CARS microscopy puts some stringent requirements on the laser intensities used for pump and Stokes beams. On one hand, the intensities of the pump and Stokes beams must be low enough to ensure that no polymerization of the unsolidified portion of the resin occurs. On the other hand, they must be high enough to create strong CARS signal. We found that the resin employed for TPP in this study consents the use of laser intensities for pump and Stokes beams that satisfy both requirements. Representative images of microstructures created by TPP still immersed in a bath of unpolymerized resin using broadband CARS microscopy are shown in Fig. (5) .
All images in Fig. (5) were recorded with a wide bandpass filter (650 ± 50 nm) in front of the detector. The difference in signal strengths observed between the polymerized and unpolymerized resin is due to density differences between the two states of the resin. Since the polymerized resin is highly cross-linked, it is also denser than the unpolymerized resin, resulting in a larger concentration per unit volume of the excited vibration modes.
To obtain the images in Fig. (5) we used a scan rate of 2 frames per second, and average laser powers for the pump and Stokes beams of 15 mW and 20 mW, respectively. The latter values were measured after the scan lens. Every image is the average of six frames. Under these imaging conditions, no polymerization of the resin is observed. This can be explained by considering that the pixel rate is only 2 μs (512 by 512 pixels frame), too short of an exposure time for creating enough radicals to start polymerization. Furthermore, both pump and Stokes beams posses much lower peak intensities than that of the beam used for TPP. Thus, images with good contrast such as the one shown in Fig. 5(a) can be obtained. The microstructure in this picture is suspended in the unpolymerized resin with the aid of two 50 μm tall towers (not shown) that anchor the object to the substrate.
The image of another microstructure recorded in situ by broadband CARS microscopy is shown in Fig. 5(b). A series of cantilevers were fabricated by TPP on top of a rectangular shaped tower that is 50 μm tall. Each cantilever was made by using the same stage velocity and average laser power for the excitation beam. The difference among the cantilevers is the spacing between the vertical polymeric lines that composes them. In particular, from top to bottom they are 0.1 μm, 0.2 μm, 0.4 μm, 0.6 μm, 0.8 μm, and 1.0 μm. Although the intensity of the signal in the field of view is not flat, it is possible to discern the difference in signal distribution along the cantilevers. The larger the spacing used to create the cantilevers, the more inhomogeneous the CARS signal.
Figures 5(c) and 5(d) are CARS images of another microstructure, but recorded at different planes. While Fig. 5(c) is recorded at a distance of 30 μm from the substrate, Fig. 5(d) is recorded at a distance of 60 μm from the substrate. The microstructure, created by TPP, is a box with sides 30 μm long and walls 60 μm high. The walls are 2 μm thick. In the middle of the microstructure at a height of 30 μm, two perpendicular lines were made by single laser passes. Finally, the box is capped with a roof. Other than the two polymerized lines, the microstructure is filled with unpolymerized resin. The CARS image recorded in the middle of the microstructure (Fig. 5(c)) shows that the cross pattern is warped. The cause of this shape alteration is most likely due to the poor structural integrity of the polymerized line created by a single laser pass. Portions of the microstructure walls in Fig. 5(c) appear dark because of the shadowing effect of the thick remaining part of the wall on the forward generated signal. When comparing the images in Figs. 5(c) and 5(d), no cross-talking between the signals in the two different planes is observed. Clear images of the cross pattern and of the roof are recorded in Fig. 5(c) and 5(d), respectively. This sectioning capability of CARS microscopy is a desirable characteristic for monitoring TPP since this technique permits the fabrication of three-dimensional microstructures.
In Fig. 5(c), the CARS signal of the unpolymerized resin trapped within the microstructure is higher than the signal of the unpolymerized resin that is outside. Although this observation requires further investigation for a full explanation, it is possible that the difference is caused by partial polymerization. Since the resin inside the microstructure cannot escape from its confinements, some of the radicals generated while writing the cross pattern can diffuse into the remaining volume and induce polymerization.
In order to determine the achievable degree of contrast between the CARS signals of the polymerized and unpolymerized resin, a test sample was fabricated by TPP. A CARS image of this sample is shown in Fig. 5(e). The microstructure is made of two solid walls 60 μm long, 15 μm wide and 50 μm tall. The two walls are attached to the glass substrate and are separated by a 50 μm gap. At a height of 25 μm, five lines are written between the two walls by single laser passes. The lines are made using a stage velocity of 20 μm/s and laser average powers of 20 mW, 18 mW, 16 mW, 14 mW, and 12 mW for lines I, II, III, IV, and V, respectively. The average powers of the laser beam employed for TPP were measured after the scan lens. A stark contrast between the signals of line I and the surrounding unpolymerized resin is observed. This contrast diminishes accordingly in the lines made with decreased laser average powers, confirming that the signal we are measuring is dependent on the degree of cross-linking of the material. Furthermore, the lines polymerized with less laser average power show signs of distress due to their poor structural integrity. The same effect noticed in Fig. 5(c) is also apparent in Fig. 5(e). The forward generated signal of the walls recorded at 25 μm cannot reach the detector because it is scattered by the additional portions of the walls above.
An intensity profile of the CARS signal along the blue line in Fig. 5(e) is shown in Fig. 5(f). The graph is normalized to the signal of the unpolymerized resin. Line I has a CARS signal more than 2.5 times larger than the signal generated by the unpolymerized resin. This value steadily decreases with the other lines until reaching a magnitude barely higher than that of the background for line V. The large contrast in signal intensities between polymerized and unpolymerized resin shown in Figs. 5(e) and 5(f) can be explained by taking into consideration the change in density of the resin upon polymerization. For example, resins composed of multifunctional monomers such as the one used in this work present volume shrinkage around 15-20% upon polymerization [39,40].
Although in the case of acrylic based resins a measurable contrast between polymerized and unpolymerized material can be observed trough conventional optical methods because of the large change in index of refraction that occurs upon polymerization, CARS microscopy provides two benefits that makes it desirable as a monitoring toll. The first benefit is the intrinsic sectioning capabilities while imaging (Figs. 5(c) and 5(d)). The second benefit is the nature of the collected signal. Since it originates from the excitation of vibrational modes in the sample, CARS microscopy can be used to monitor specific molecular bonds within the components of the resin. The second point is particularly useful for resins that do not present large change in index of refraction when exposed to light such as SU-8 .
3.4 Imaging in real time
The experimental setup described in Fig. 1 permits movement of the sample around a fixed laser beam that induces TPP and, at the same time, it allows for the pump and Stokes beams to probe the sample in a raster scan fashion. If the area scanned for generating CARS images is larger than the size of the desired microstructure, the entire process of microfabrication by TPP can then be monitored with chemical specificity.
To demonstrate this capability, a two-dimensional grid pattern sample was fabricated by TPP while a large coplanar area of this pattern was concurrently imaged by broadband CARS microscopy. A laser average power of 18 mW was used for microfabrication. The imaging part of the process was simultaneously performed by using laser average powers of 12 mW and 13 mW for the pump and Stokes beams, respectively. The average powers of the three laser beams were measured after the scan lens. To create the sample grid pattern, the stage was moved at a velocity of 10 μm/s. For imaging, a frame rate of 2 frames per second was chosen. Forward directed CARS signal was collected using a wide bandpass filter (650 ± 50 nm) in front of the detector, permitting real-time imaging with contrast based on the degree of cross-linking in the sample. The grid-like microstructure was created by first writing six parallel lines 50 μm long with 10 μm spacings. Then, the same pattern was transposed 90° and repeated, superimposing it on the first one.
A sequence of representative frames acquired during real time imaging of TPP is shown in Fig. 6 . When assembled together, these images form a movie (Media 1) that clearly displays the process of writing in the resin. Although a higher frame rate would have delivered a more precise time capture of TPP, Fig. 6 proves that TPP can be monitored in real time by using broadband CARS microscopy. The contrast between the CARS signals of the polymerized and unpolymerized resin is large enough to distinguish without a doubt regions of the sample that have been patterned by TPP and regions of the sample that have not.
We have shown that broadband CARS microscopy is an effective tool for the microscopic characterization of structures fabricated by TPP. The resin used in this study permits analysis with low laser average power avoiding interferences due to boiling and/or modification of the sample while imaging. Structural information on microstructures can be retrieved by detecting the strong CARS signal generated by the carbon-hydrogen stretching modes which are abundantly present in our resin. Furthermore, we have demonstrated that one experimental setup based on a single femtosecond laser oscillator is sufficient for performing both TPP and broadband CARS microscopy. Microstructures created by TPP can be characterized when still immersed in the bath of unpolymerized resin or during the microfabrication process itself.
We foresee that the methodology described in this article for characterizing the process of TPP in situ and in real time can be applied to photosensitive materials other than acrylic based resins, such as hybrid organic-inorganic sol-gels, and epoxy resins [42,43]. After a full characterization of the signatures peaks in the sample Raman spectrum, broadband CARS microscopy can be used to monitor the dynamic of specific chemical bonds over periods of time that are relevant for microfabrication in TPP.
References and links
2. M. Deubel, G. von Freymann, M. Wegener, S. Pereira, K. Busch, and C. M. Soukoulis, “Direct laser writing of three-dimensional photonic-crystal templates for telecommunications,” Nat. Mater. 3(7), 444–447 (2004). [CrossRef] [PubMed]
3. R. A. Farrer, C. N. LaFratta, L. J. Li, J. Praino, M. J. Naughton, B. E. A. Saleh, M. C. Teich, and J. T. Fourkas, “Selective functionalization of 3-D polymer microstructures,” J. Am. Chem. Soc. 128(6), 1796–1797 (2006). [CrossRef] [PubMed]
4. S. Maruo and H. Inoue, “Optically driven viscous micropump using a rotating microdisk,” Appl. Phys. Lett. 91(8), 084101 (2007). [CrossRef]
5. S. Maruo, K. Ikuta, and H. Korogi, “Force-controllable, optically driven micromachines fabricated by single-step two-photon micro stereolithography,” J. Microelectromech. Syst. 12(5), 533–539 (2003). [CrossRef]
6. P. Tayalia, C. R. Mendonca, T. Baldacchini, D. J. Mooney, and E. Mazur, “3D Cell-Migration Studies using Two-Photon Engineered Polymer Scaffolds,” Adv. Mater. 20(23), 4494–4498 (2008). [CrossRef]
7. S. Juodkazis, V. Mizeikis, and H. Misawa, “Three-Dimensional Structuring of Resists and Resins by Direct Laser Writing and Holographic Recording,” Adv. Polym. Sci. 213, 157–206 (2008).
8. M. Malinauskas, A. Zukauskas, G. Bickauskaite, R. Gadonas, and S. Juodkazis, “Mechanisms of three-dimensional structuring of photo-polymers by tightly focussed femtosecond laser pulses,” Opt. Express 18(10), 10209–10221 (2010). [CrossRef] [PubMed]
9. S. O. Onuh and K. K. B. Hon, ““An esperimental investigation into the effect of hatch pattern in stereolithography,” CIRP Annals - Manuf Tech. 47(1), 157–160 (1998). [CrossRef]
10. S. O. Onuh and K. K. B. Hon, “Improving stereolithography part accuracy for industrial applications,” Int. J. Adv. Manuf. Technol. 17(1), 61–68 (2001). [CrossRef]
11. Q. Sun, S. Juodkazis, N. Murazawa, V. Mizeikis, and H. Misawa, “Freestanding and movable photonic microstructures fabricated by photopolymerization with femtosecond laser pulses,” J. Micromech. Microeng. 20(3), 035004 (2010). [CrossRef]
12. W. Haske, V. W. Chen, J. M. Hales, W. T. Dong, S. Barlow, S. R. Marder, and J. W. Perry, “65 nm feature sizes using visible wavelength 3-D multiphoton lithography,” Opt. Express 15(6), 3426–3436 (2007). [CrossRef] [PubMed]
13. L. Li, R. R. Gattass, E. Gershgoren, H. Hwang, and J. T. Fourkas, “Achieving λ/20 resolution by one-color initiation and deactivation of polymerization,” Science 324(5929), 910–913 (2009). [CrossRef] [PubMed]
14. S. Nakanishi, S. Shoji, S. Kawata, and H. B. Sun, “Giant elasticity of photopolymer nanowires,” Appl. Phys. Lett. 91(6), 063112 (2007). [CrossRef]
16. R. Zadoyan, T. Baldacchini, M. Karavitis, and J. Carter, “CARS microspectrometer with a suppressed nonresonant background,” Ultrafast Phenomena XVI, Springer Series in Chemical Physics 92, 997–999 (2009). [CrossRef]
17. T. Baldacchini, M. Zimmerley, E. O. Potma, and R. Zadoyan, “Chemical mapping of three-dimensional microstructures fabricated by two-photon polymerization using CARS microscopy,” Proc. SPIE 7201, 72010Q72011 (2009). [CrossRef]
18. T. Baldacchini, M. Zimmerley, C. H. Kuo, E. O. Potma, and R. Zadoyan, “Characterization of microstructures fabricated by two-photon polymerization using coherent anti-stokes Raman scattering microscopy,” J. Phys. Chem. B 113(38), 12663–12668 (2009). [CrossRef] [PubMed]
19. A. Zumbusch, G. R. Holtom, and X. S. Xie, “Three-dimensional vibrational imaging by coherent anti- Stokes Raman scattering,” Phys. Rev. Lett. 82(20), 4142–4145 (1999). [CrossRef]
20. J. X. Cheng and X. S. Xie, “Coherent anti-Stokes Raman scattering microscopy: Instrumentation, theory, and applications,” J. Phys. Chem. B 108(3), 827–840 (2004). [CrossRef]
21. C. L. Evans and X. S. Xie, “Coherent anti-stokes Raman scattering microscopy: chemical imaging for biology and medicine,” Annu Rev Anal Chem (Palo Alto Calif) 1(1), 883–909 (2008). [CrossRef]
25. H. Kano and H. Hamaguchi, “Ultrabroadband (> 2500 cm(−1)) multiplex coherent anti-Stokes Raman scattering microspectroscopy using a supercontinuum generated from a photonic crystal fiber,” Appl. Phys. Lett. 86(12), 121113 (2005). [CrossRef]
26. H. Kano and H. Hamaguchi, “Near-infrared coherent anti-Stokes Raman scattering microscopy using supercontinuum generated from a photonic crystal fiber,” Appl. Phys. B 80(2), 243–246 (2005). [CrossRef]
27. H. Kano and H. O. Hamaguchi, “In-vivo multi-nonlinear optical imaging of a living cell using a supercontinuum light source generated from a photonic crystal fiber,” Opt. Express 14(7), 2798–2804 (2006). [CrossRef] [PubMed]
28. A. C. T. Ko, A. Ridsdale, M. S. D. Smith, L. B. Mostaço-Guidolin, M. D. Hewko, A. F. Pegoraro, E. K. Kohlenberg, B. Schattka, M. Shiomi, A. Stolow, and M. G. Sowa, “Multimodal nonlinear optical imaging of atherosclerotic plaque development in myocardial infarction-prone rabbits,” J. Biomed. Opt. 15(2), 020501 (2010). [CrossRef] [PubMed]
29. B. von Vacano, L. Meyer, and M. Motzkus, “Rapid polymer blend imaging with quantitative broadband multiplex CARS microscopy,” J. Raman Spectrosc. 38(7), 916–926 (2007). [CrossRef]
30. M. Muller and J. M. Schins, “Imaging the thermodynamic state of lipid membranes with multiplex CARS microscopy,” J. Phys. Chem. B 106(14), 3715–3723 (2002). [CrossRef]
31. Z. D. Schultz, M. C. Gurau, and L. J. Richter, “Broadband coherent anti-Stokes Raman spectroscopy characterization of polymer thin films,” Appl. Spectrosc. 60(10), 1097–1102 (2006). [CrossRef] [PubMed]
32. S. H. Lim, A. G. Caster, O. Nicolet, and S. R. Leone, “Chemical imaging by single pulse interferometric coherent anti-stokes Raman scattering microscopy,” J. Phys. Chem. B 110(11), 5196–5204 (2006). [CrossRef] [PubMed]
33. S. Murugkar, C. Brideau, A. Ridsdale, M. Naji, P. K. Stys, and H. Anis, “Coherent anti-Stokes Raman scattering microscopy using photonic crystal fiber with two closely lying zero dispersion wavelengths,” Opt. Express 15(21), 14028–14037 (2007). [CrossRef] [PubMed]
34. A. F. Pegoraro, A. Ridsdale, D. J. Moffatt, Y. W. Jia, J. P. Pezacki, and A. Stolow, “Optimally chirped multimodal CARS microscopy based on a single Ti:sapphire oscillator,” Opt. Express 17(4), 2984–2996 (2009). [CrossRef] [PubMed]
35. Y. J. Lee, S. H. Parekh, Y. H. Kim, and M. T. Cicerone, “Optimized continuum from a photonic crystal fiber for broadband time-resolved coherent anti-Stokes Raman scattering,” Opt. Express 18(5), 4371–4379 (2010). [CrossRef] [PubMed]
36. K. M. Hilligsøe, T. V. Andersen, H. N. Paulsen, C. K. Nielsen, K. Mølmer, S. Keiding, R. Kristiansen, K. P. Hansen, and J. J. Larsen, “Supercontinuum generation in a photonic crystal fiber with two zero dispersion wavelengths,” Opt. Express 12(6), 1045–1054 (2004). [CrossRef] [PubMed]
37. M. Balu, T. Baldacchini, J. Carter, T. B. Krasieva, R. Zadoyan, and B. J. Tromberg, “Effect of excitation wavelength on penetration depth in nonlinear optical microscopy of turbid media,” J. Biomed. Opt. 14(1), 010508 (2009). [CrossRef] [PubMed]
40. K. S. Anseth, C. N. Bowman, and N. A. Peppas, “Polymerization Kinetics and Volume Relaxation Behavior of Photopolymerized Multifunctional Monomers Producing Highly Cross-Linked Networks,” J. Polym. Sci. A 32(1), 139–147 (1994). [CrossRef]
41. K. K. Seet, V. Mizeikis, K. Kannari, S. Juodkazis, H. Misawa, N. Tétreault, and S. John, “Templating and replication of spiral photonic crystals for silicon photonics,” IEEE J. Sel. Top. Quantum Electron. 14(4), 1064–1073 (2008). [CrossRef]
42. R. Houbertz, L. Frohlich, M. Popall, U. Streppel, P. Dannberg, A. Brauer, J. Serbin, and B. N. Chichkov, “Inorganic-organic hybrid polymers for information technology: from planar technology to 3D nanostructures,” Adv. Eng. Mater. 5(8), 551–555 (2003). [CrossRef]
43. W. H. Teh, U. Durig, G. Salis, R. Harbers, U. Drechsler, R. F. Mahrt, C. G. Smith, and H. J. Guntherodt, “SU-8 for real three-dimensional subdiffraction-limit two-photon microfabrication,” Appl. Phys. Lett. 84(20), 4095–4097 (2004). [CrossRef]