Non-linear optical (NLO) systems have become very sophisticated in recent years, ranging from super-resolution fluorescence microscopes to multi-model biomedical instruments. These advancements allow for resolutions in the tens of nanometers, the visualization of both fluorescently tagged and unlabeled molecules in one image, and the capability to conduct in-vivo cancer research.
Ultra-short pulse femtosecond lasers with an extremely high peak pulse intensity are needed in many NLO systems to produce NLO signals. In these instruments, the pulsed beam reaches the sample after being reflected off or transmitted through many different optical components.
If these mirrors and optical filters and not specifically designed for use with femtosecond lasers, the peak pulse intensity will decrease due to group delay dispersion (GDD) (Figure 1) every time the pulse train is transmitted or reflected. The compounded effect will ultimately degrade instrument performance because a reduction in peak pulse intensity can reduce the number of NLO signals.
Figure 1. Example showing how decreased peak pulse intensity due to GDD can result when a femtosecond laser pulse is transmitted through an optical filter.
Non-Linear Optical (NLO) Systems
Within a fluorophore, linera fluorescence occurs when photons of light at a specific wavelength excite the electrons. The electrons emit photons of a longer wavelength than those that were initially absorbed when they return to the ground state. Therefore, the excitation light is always higher in energy than the emission light in linear fluorescence systems.
Conversely, in NLO techniques such as multiphoton (MP) fluorescence microscopy, the simultaneous absorption of multiple longwave photons can lead to emissions with a shorter wavelength than the excitation light (Figure 2).
Figure 2. Jablonski diagrams showing linear vs. non-linear fluorescence. In linear single-photon excitation, the absorption of short wavelength photons results in a longer wavelength fluorescence emission. In non-linear two-photon excitation (2PE), the absorption of two long wavelength photons results in a shorter wavelength fluorescence emission.
In addiiton to MP fluorescence microscopy, there are other NLO methods that produce different results (Figure 3). For instance, the second and third harmonic generation fluorescence microscopy (SHG and THG) techniques produce an NLO response in molecules lacking a center of symmetry .
Photons that are ½ or ⅓ of the original wavelength are emitted when several longwave photons are simultaneously absorbed by these molecules. Laser pulses produce Raman signals that are generated from the vibrational motion of molecules within the sample in coherent anti-stokes Raman scattering (CARS) and stimulated Raman scattering (SRS) techniques [2,3].
Figure 3. Jablonski diagrams showing multiple different NLO responses. Dashed lines indicate virtual states.
Photon absorption does not always lead to excitation of the sample. For example, longwave photons can be used to bring the electrons of an already-excited sample back to the ground state in stimulated emission depletion fluorescence microscopy (STED), without producing a spontaneous fluorescence response.
In this technique, an excitation laser is combined with the STED beam, where the laser produces either a single or MP fluorescence response. A highly focused region of spontaneous fluorescence surrounded by a region of stimulated emission is produced when a vortex phase plate (VPP) is use to produce a doughnut-shaped pattern of the STED beam .
A super-resolution composite image is formed when the beams in this configuration are mechanically scanned across the sample. The image allows the visualization of structures that can be as small as a single molecule .
Researchers now have the ability to detect several NLO responses within a single sample due to the recent advancements in multi-modal NLO instruments. Researchers are able to gather a large quantity of information about the fluorescently unlabeled and labeled molecules within a tissue sample using the imaging techniques .
Thin-Film Mirrors and Optical Filters for NLO Systems
Controlling Group Delay Dispersion (GDD)
Thin-film mirrors and filters are made when alternating layers of materials are deposited with varying refraction indices onto a substrate. Internal interference takes place when part of the light reflects at each layer as light makes its way through the filter. The configuration and thicknesses of the layers determine if certain wavelengths of light are transmitted through the filter and others are either reflected off of it or absorbed by it.
A HR dielectric mirror, also known as a Bragg reflector, is one of the simplest thin-film designs. This thin-film coating contains layer pair stacks, where each pair consists of a high-index-material layer and a low-index-material layer, with each of them having ¼ wavelength of optical thickness.
These dielectric mirrors have more than 99.9% reflectivity across a wide range of wavelengths, but they can result in a wavelength dependent phase shift of reflected light (Figure 4). This is especially true if the mirror is designed to reflect over a broad wavelength range when several quarter-wave stacks at different center wavelengths are optimized.
This phase delay ultimately results in reduced peak pulse intensity and GDD when a femtosecond laser is reflected off of the mirror.
Figure 4. Example showing how a wavelength dependent phase shift of light reflected off of a dielectric mirror can occur when that mirror has not been optimized to control GDD.
There are many thin-film optical components available that attempt to counteract the effect of GDD. One option is a simple low-dispersion mirror that can minimize GDD over a short region of high reflection when one quarter-wave stack at a single wavelength is optimized (Figure 5).
This option is adequate to minimize dispersion for pulsed lasers that function at a single wavelength. However, a simple low dispersion mirror would not be sufficient to control GDD, because Ti:Sapphire lasers can have a relatively wide emission range.
Another option is to use a combination of different chirped mirrors to direct the beam to the sample, where the mirrors have a layer thicknesses that vary across the layer stack. By altering the layer thicknesses, a variety of GDD effects can be produced that vary with wavelength in a controlled way.
The net GDD can be controlled over a wide range of wavelengths than that is possible by using a simple low-dispersion mirror when a pulsed laser is reflected off of one mirror and another is designed to have the opposite GDD.
Figure 5. Reflection and GDD for a simple low-dispersion dielectric mirror designed by optimizing one quarter wave stack at a single wavelength.
Due to steady improvements in coating capability and thin-film design, low dispersion thin-film mirrors can now be produced to achieve more than 99.5% reflection and low GDD across a wide wavelength range, as shown in Figure 6. In these dispersion-controlled thin-film coatings, a single optical component preserves the peak pulse intensity of femtosecond lasers.
Figure 6. Reflection and GDD for a dispersion controlled thin-film mirror. Group delay dispersion is less than ± 45 fs2 across a broad range of wavelengths where reflection is close to 100%.
While GDD is able to be controlled across a broad range of wavelengths corresponding to regions of reflection, this is not the case for regions of transmission. This is because a Hilbert transform links transmission amplitude and phase responses. Therefore, changes in the phase response immediately show up in the amplitude response, and vice-versa.
This limitation is applicable to all light transmitted through thin-film filters, but if the filter has an asymmetric design this does not apply to light reflected by the filter. Therefore, control of dispersion is possible without compromising reflectivity, but control of dispersion over a wide range of wavelengths in a transmissive filter usually results in a decrease in transmission.
This results in a trade off as NLO systems can be configured where a dichroic beamsplitter either reflects the pulsed laser to the sample, or transmits it to the sample. If transmitted to the sample, the peak pulse intensity may decrease due to GDD, although the dichroic’s high reflectivity allows weak emission signals to reach the detector.
When reflected to the sample, peak pulse intensity can be maintained by controlling GDD, although weaker signals may not reach the detector. This is because GDD control in the reflection band of the dichroic beamsplitter usually results in reduced amplitudes of the transmission band.
Laser-Induced Damage Threshold (LIDT) Testing
The thin-film optical components that are integrated into NLO systems must be designed to withstand high-intensity laser radiation for many years.
It is recommended that a laser damage rating is specified with the thin-film manufacturer for any mirrors, laser excitation filters, and dichroic beamsplitters that are integrated into NLO instruments, although many hard-coated thin-films have high resistance to laser damage. The laser damage rating can be determined by conducting an LIDT testing with an appropriate picosecond or femtosecond laser.
Surface Flatness and Coating Stress
When selecting optical components for NLO systems, surface flatness is another property that needs to be considered. When the coating is deposited onto a substrate, the stress of a thin-film coating causes the substrate to bend, leading to a bowl or dome shaped curvature (Figure 7).
Increased RWE is caused by this coating-stress-induced curvature, which can lead to image distortion. Therefore, flatness control is necessary for the dichroic mirrors and thin-film mirrors in NLO and other laser imaging systems.
Figure 7. Interferometric surface flatness measurement showing the coating-stress induced curvature of a typical thin-film dichroic filter. Flatness was measured at 2.87 wave P-V over the clear aperture.
Multiple techniques can be used to minimize the curvature, with the simplest being the use of a thicker substrate that is less susceptible to coating stress. If a thinner substrate is needed, the next option has traditionally been to add a backside compensation coating to the filter. However, both of these options come with a trade-off.
To minimize autofluorescence from the substrate to the detector, dichroics used for fluorescence applications are coated on fused silica, which becomes quite expensive when thickness is increased. When a backside compensation technique is used, there is an increase in the coating time and complexity of the design, as the backside coating must be thick enough to balance the coating stress of the filter .
Another option is a low-stress manufacturing method that produces ultra-flat mirrors and dichroics, without the need for backside compensation (Figure 8). The dichroic filters shown in Figures 7 and 8 have the same spectral response, coating thickness, substrate thickness, and substrate material. However, the dichroic produced by the low-stress process is much flatter than that by a standard method.
Figure 8. Interferometric measurement showing the low coating stress of an ultra-flat dichroic filter produced using a low-stress process. Flatness was measured at 0.21 wave P-V over the clear aperture.
Some of the greatest recent advancements in biological disciplines have been due to NLO systems. They help researchers to accurately quantify super-resolution fluorescence images and imaging of in-vivo samples in a label-free and non-invasive manner. To achieve optimal performance, the optical components integrated into NLO systems need to be specifically designed for such instruments.
This is particularly important for NLO systems that use a femtosecond laser, as these instruments need reflective components to minimize dispersion and preserve peak pulse intensity. The engineers at Alluxa can help to customize high-performance thin films specially designed for NLO systems, regardless of system requirements.
 Campagnola, P., Millard, A. C., Terasaki, M., Hoppe, P. E. Malone, C. J., and W. A. Mohler. (2002). Three-Dimensional High- Resolution Second-Harmonic Generation Imaging of Endogenous Structural Proteins in Biological Tissues. Biophysical Journal, 82(1): 493-508.
 Cheng, J., Volkmer, A., Book, L. D., and S. Xie. (2001). An Epi-Detected Coherent Anti-Stokes Raman Scattering (E-CARS) Microscope with High Spectral Resolution and High Sensitivity. Journal of Physical Chemistry B, 105(7): 1277-1280.
 Ji, M., Orringer, D. A., Freudiger, C. W., Ramkissoon, S., Liu, X., Lau, D., Golby, A. J., Norton, I., Hayashi, M., Agar, N. Y., Young, G. S., Spino, C., Santagata, S., Camelo-Piragua, S., Ligon, K. L., Sagher, O., and X. S. Xie. (2013). Rapid, label-free detection of brain tumors with stimulated Raman scattering microscopy. Science Translational Medicine, 5(201): 201ra119.
 Kasper, R., Harke, B., Forthmann, C., Tinnefeld, P., Hell, S. W., and M. Sauer. (2010). Single-molecule STED microscopy with photostable organic fluorophores. Small, 6(13): 1379-1384.
 Madsen, C. K. and J. H. Zhao. (1999). Optical Filter Design and Analysis. Hoboken, N. J. John Wiley & Sons, Inc. New York, NY.
 Tang, S., Krasieva, T. B., Zhongping, C., Tempea, G, and B. J. Tromberg. (2006). Effect of pulse duration on two-photon excited fluorescence and second harmonic generation in nonlinear optical microscopy. Journal of Biomedical Optics 11(2): 020501.
 Thorley, J. A., Pike, J., and J. Z. Rappoport. (2014). Super-resolution microscopy: A comparison of commercially available options. Fluorescence Microscopy: Super Resolution and Other Novel Techniques. Edited by Cornea, A. and P. M. Conn. Academic Press / Elsevier. London, UK. 185-197.
 Yamada, M., Lin, L. L., and T. W. Prow. (2014). Multiphoton microscopy applications in biology. Fluorescence Microscopy: Super Resolution and Other Novel Techniques. Edited by Cornea, A. and P. M. Conn. Academic Press / Elsevier. London, UK. 199-212.
 Yue, S., Slipchenko, M. N., and J. Cheng. (2011). Multimodal Nonlinear Optical Microscopy. Laser Photon Review, 5(4): 0.1002/lpor.201000027.
 Alluxa Engineering Staff. (2012). Thin Substrate, Dichroic and Polychroic Thin Film Filters Featuring Flatness Less Than 0.1 Waves RMS. Alluxa White Paper Series. http://www.alluxa.com/learning-center.
This information has been sourced, reviewed and adapted from materials provided by Alluxa.
For more information on this source, please visit Alluxa.