2.1.

Femtosecond Laser Microscope

The fs laser Integral Pro 400 oscillator with extremely short pulses of 10 fs (Femtolasers Produktions GmbH, Vienna, Austria) with 0.2 MW peak power, 85 MHz, and $M^2<1.3$ was employed. The laser beam was combined with a ZEISS Axio Observer inverted fluorescence microscope equipped with a galvoscanning unit (JenLab GmbH, Jena, Germany) and piezodriven focussing optics ($z$ stage controller) (nanoMipos400 with digital controller, Piezosysteme Jena). For dispersion pre-compensation of the whole optical system, a set of chirped-mirrors was used. The in situ pulse duration was measured by means of a second order interferometric scanning autocorrelator using a non-linear photodiode at the focus of the objective (Femtometer, Femtolasers Production GmbH, Wien, Austria). Typical short pulse duration of 12 fs (in situ pulse width) and a bandwidth of about 100 nm, respectively, were measured at the sample plane. This broad spectrum had two maxima (770 and 830 nm) in this particular laser, as measured with a spectrometer. A center wavelength of 800 nm was used for calculations.

The ultracompact scanning microscope with highly-dispersive, large-NA objective (Zeiss EC Plan-Neofluar $40\times /1.3$ oil) was employed. Optionally, double flint glass wedges, and glass blocks have been inserted in the beam path to vary the in situ pulse length from 12 fs up to 3 ps. A schematic representation of the setup is demonstrated in Fig. 1.

Fig. 1

Schematic representation of the experimental setup. A broadband 85 MHz NIR fs laser was combined with an inverted fluorescence microscope equipped with galvoscanners and piezo-driven focussing optics. An in situ pulse width of 12 fs was measured at the sample plane. A dispersion pre-compensation module consisting of chirped mirrors was used to obtain 12 fs laser pulses at the target. Optionally, flint glass wedges and glass blocks have been inserted in the beam path to vary the in situ pulse length from 12 fs up to 3 ps. A CCD camera attached to the side port enabled online imaging of the target and the laser beam. A PMT attached to the front port was employed to image two-photon fluorescence.

The microscope was employed in the two-photon fluorescence excitation mode at mean powers in the microwatt range for nondestructive imaging and in the milliwatt range for nanoprocessing in two exposure modes: 1. scanning of a region of interest (ROI) and 2. by single point illumination where the galvoscanners were fixed to a point of interest. A CCD camera attached to the side port enabled online imaging of the target, the laser beam, and the formation of plasma-filled cavitation bubbles. A side-on photomultiplier tube PMT-H7732 (Hamamatsu) with a rise-time of 2.2 ns and a spectral detection range of 185 to 680 nm, according to the producer’s data sheet, was employed to detect two-photon fluorescence.

Laser powers were measured with the power meter FieldMaxII-TO™ (Coherent, USA) at the specimen plane. The exposure time was controlled by a shutter controller SC10 (Thorlabs. Inc., New Jersey, USA). Additionally, a motorized stage (Märzhäuzer, Wetzlar) was used to move the samples in $x$, $y$ directions.

2.2.

Metaphase Chromosomes and Cells

Metaphase chromosomes were prepared from human peripheral blood cells according to the standard procedure.³² One group of chromosomes was spread onto cover slips without staining, and the second group was stained with Giemsa and used for photobleaching studies. Other metaphase chromosomes were prepared for high resolution AFM microscopy purposes. Therefore, chromosomes were mounted onto glass cover slides (170 μm thickness) after cleaning with isopropanol/acetone ($1:1$) and coating with 0.01% Poly-L-Lysine (Sigma). Afterward, they were fixed with 2% glytaraldehyde in PBS (phosphate buffer saline) for 10 min, rinsed with distilled water and critical point dried. Figure 2 shows a fluorescent image of human metaphase chromosomes. About 200 μW (12 fs) mean power was required to image chromosome probes fixed with glytaraldehyde. Chromosome probes without staining and stained with Giemsa have been visualized with CCD camera during laser processing.

Fig. 2

Metaphase chromosomes of human leukocytes before laser nanoprocessing. The two-photon fluorescence image was taken with a 12 fs (central excitation wavelength at 800 nm, detection filter: BG39) laser pulses at a mean power of 200 μW. The intense fluorescence originates from the fixative 2% glytaraldehyde.

2.3.

Atomic Force Microscope and Image Analysis

The samples were imaged with a Dimension™ 3100 (Bruker Nano Surface Division, Santa Barbara, CA, USA) AFM in tapping mode (TM). The cantilevers (Micro Cantilever, OMCL-AC160TS-W2, Olympus) with silicon tips at a diameter of about 7 nm, a spring constant of $42\mathrm{N}/\mathrm{m}$, and a resonance frequency of 300 kHz were used. Images of chromosomes and cells were obtained with different resolutions. Figure 3 shows topographic AFM images of human metaphase chromosomes. As it is shown on the image, the length of human metaphase chromosomes varied from 1.6 to 7.0 μm and the thickness of two chromatid sisters varied from 1.6 to 2.0 μm.

Fig. 3

High resolution 2D and 3D representative topographic AFM images of human metaphase chromosomes. AFM measurements show that the size of metaphase chromosomes varied from 1.6 to 7.0 μm in length. The thickness of two chromatid sisters is between 1.6 and 2.0 μm.

WSxM 4.0 software was used to analyze AFM (.stp) files and measure profiles from line cuts and holes in the cells and chromosomes. AFM images were imported as ASCII matrix (.txt files) format in OriginLab 8G software. The mean cut width was calculated along the whole cut ($384\text{pixels}\equiv 24\mathrm{\mu m}$).

3.1.

Photobleaching Effects of the Laser

Giemsa stained human chromosomes and cell nuclei on cover slips were used for comparative photobleaching and nano- and microsurgery studies using 12 fs, 300 fs, and 3 ps laser pulses and different mean powers between 0.4 and 5 mW. The laser beam was focused directly into the samples through the cover slip which has a thickness of 170 μm. Cuts were realized by the mode “line scanning.” Focused laser beam and bleaching effects of Giemsa were observed with the CCD camera. Typically, a total of 10 s. exposure time was chosen to realize multiple scans per line (12 to $24\mathrm{ms}/\text{scan}$). Certain distances (500 nm to 2 μm) between lines were realized by means of the motorized stage. Laser processed samples were measured with the AFM microscope to confirm photobleaching and ablation effects on the biological material. Photobleaching effects were observed at powers as low as 0.4 mW (12 fs), 0.8 mW (300 fs), and 1.5 mW (3 ps) (images are not shown). The ablation effects started at higher power levels as confirmed by AFM (Sec. 3.2, Figs. 4 and 5).

Fig. 4

AFM images of three cell nuclei processed by laser pulses at 12 fs (left) vs 300 fs (middle) and 3 ps (right). The left figure (12 fs) demonstrates that line cuts can be created with a mean power as low as 0.4 mW. In the case of 300 fs, line cuts can be created with 1.5 mW. At 3 ps first ablation effects were seen at 5 mW. Note: AFM images demonstrated laser ablation effects. Photobleaching of Giemsa stained probes was monitored with a CCD camera during laser processing by line scanning with 0.4 to 5 mW mean powers. Photobleaching effects occurred at very low powers of 0.4 mW (12 fs), 0.8 mW (300 fs), and 1.5 mW (3 ps). Total scan time: 10 s. All line scans with 12 fs laser show ablation. Nearly no ablation effects were found in the 3 ps exposed cell nucleus. Scale bar: 4.8 μm.

Fig. 5

(a) AFM scans of Line cuts in a cell nucleus realized with 12 fs laser pulses at mean powers (mW) of 4.0, 3.5, 3.0, 2.5, 2.0, 1.5, 1.0, 0.8, 0.4, and 5.0. (b) The graph (mean profile line) indicates that a low laser power of 0.4 mW was sufficient to induce ablation effects. (c) Magnified AFM image with three incisions at 0.4, 0.8, and 1 mW. (d) Analysis of individual line cuts of the cell nucleus shows that sub-100 nm lines can be created with 12 fs laser pulses.

3.2.

Nano- and Microprocessing of the Cell Nucleus Dependence on Pulse Width

To characterize laser ablation effects of laser dependence on pulse width, samples were measured with the AFM microscope in the tapping mode. Interestingly, a very low power of 0.4 mW in the case of 12 fs can already induce visible modifications in the cell nucleus morphology. Due to the non-homogenous structure of the nucleus, the width of the line cuts varied along the line incision. Analysis of each line shows that the width of the line cuts became larger and more homogenous with increased laser powers (Fig. 5). Additionally, slower and longer scan time also resulted in better cuts.

Comparative experiments have been carried out with laser pulses of 12 fs versus 300 fs versus 3 ps. That means the laser pulse width was increased by a factor 25 and 250, respectively. Cells have been cut with different laser mean powers between 0.4 and 5.0 mW at the same conditions. Samples were measured with the AFM, and mean curves have been analyzed. The mean value has been performed over 384 profiles per line. Results show that line cuts can be created with a mean power as low as 0.4 mW ($E=4.2\mathrm{mJ}/{\mathrm{cm}}^2$, $I_{\text{peak}}=0.35\mathrm{TW}/{\mathrm{cm}}^2$) in the case of 12 fs (Figs. 4 and 5). When using 300 fs, ablation effects started at 1.5 mW ($E=15.9\mathrm{mJ}/{\mathrm{cm}}^2$, $I_{\text{peak}}=0.053\mathrm{TW}/{\mathrm{cm}}^2$). More power $P_{\text{th}}$ is required to induce ablation when using picosecond (ps) laser pulses. A value of 5.0 mW ($E=53.0\mathrm{mJ}/{\mathrm{cm}}^2$, $I_{\text{peak}}=0.017\mathrm{TW}/{\mathrm{cm}}^2$) was determined when using 3 ps. No visible ablation effects were found below 5 mW mean power as demonstrated with the AFM (Fig. 4).

A linear fit with a slope of $0.45\pm 0.03$ was determined, when depicting the power values ($P_{\text{th}}$) of 0.4, 1.5, and 5 mW for the onset of ablation versus the pulse length of 12 fs, 300 fs, and 3 ps on a bi-logarithmic scale (Fig. 6). That means that $P_{\text{th}}$ is nearly linearly dependent on $\sqrt{\tau }$, which corresponds to a ${P_{\text{th}}}^2/\tau$ relation.

Fig. 6

Threshold powers to induce ablation effects on the cell versus laser pulse width are plotted on a bi-logarithmic scale. The slope of $0.45\pm 0.03$ nearly indicates a $P^2/\tau$ dependence.

3.3.

Drilling and Cutting Chromosomes with 12 fs Laser Pulses

Low mean mW powers were sufficient to cut human chromosomes by line scanning and to drill holes by single point illumination (Fig. 7). About 0.7 to 1.0 mW ($E=7.4$ to $10.6\mathrm{mJ}/{\mathrm{cm}}^2$, $I_{\text{peak}}=0.62$ to $0.88\mathrm{TW}/{\mathrm{cm}}^2$) laser mean power was sufficient to drill holes into chromosomes with a diameter below 100 nm. Figure 7(b) and 7(c) shows the hole with a diameter of 90 nm and a FWHM of 52 nm created by 12 fs laser. Increased mean powers resulted in bigger holes. Five mW and higher mean powers resulted in disruption of larger parts of the chromosomes.

Fig. 7

Topographic AFM images showing cutting and drilling effects of the laser after processing on chromosomes. (a) The arrows show laser produced holes with different diameter size and cut incision. (b) and (c) The measured profile from a hole with a diameter of 90 nm (FWHM of 52 nm).

Experimental results show that the ablation threshold of the laser power depends on fixative or staining dyes, innate components, the density, and the structure of the biomaterials. Non-stained probes require a higher energy than those of stained probes. This corresponds to the findings of König et al.⁶ and Kuetemeyer et al.³³ In current study, cutting of metaphase chromosomes required higher intensities/energies than the cutting of the nucleus itself. It may be related to the highly condensed structure of chromosome material of DNA in the chromosomes compared to the nucleus.

However, fs lasers are much more precise than pico- and nanosecond laser pulses because the destructive effects occur after the laser exposure and the major process is multiphoton ionization.^34,35 Photodisruptive effects based on the formation of shock waves and cavitation bubble dynamics can be kept to a minimum. These effects are related to the pulse energy. When using fs laser pulses of low picojoule pulse energies, photodisruptive effects are negligible. There are no plasma shielding effects involved in contrast to nanosecond laser pulses. Detailed explanation on the laser-matter interactions of fs laser pulses versus ps and ns laser pulses was presented by Vogel et al.³⁵ They calculated thresholds for laser ablation, which are orders lower for ultrashort pulses versus ns pulses.

The comparative experiments described in this paper with laser pulses of 12 fs versus 300 fs versus 3 ps (laser pulse width increased by a factor 25 and 250, respectively) showed that higher power $P_{\text{th}}$ is required for the ablation effect when using ps laser pulses (0.4, 1.5, and 5 mW, respectively).