Open Access
Add to BibSonomyAbstract
In the present paper we demonstrate the possibility to image dyed solids, i.e. Rhodamine B dyed polyethylene spheres, by means of two-photon absorption-induced photoacoustic scanning microscopy. A two-photon luminescence image is recorded simultaneously with the photoacoustic image and we show that location and size of the photoacoustic and luminescence image match. In the experiments photoacoustic signals and luminescence signals are generated by pulses from a femtosecond laser. Photoacoustic signals are acquired with a hydrophone; luminescence signals with a spectrometer or an avalanche photo diode. In addition we derive the expected dependencies between excitation intensity and photoacoustic signal for single-photon absorption, two-photon absorption and for the combination of both. In order to verify our setup and evaluation method the theoretical predictions are compared with experimental results for liquid and solid specimens, i.e. a carbon fiber, Rhodamine B solution, silicon, and Rhodamine B dyed microspheres. The results suggest that the photoacoustic signals from the Rhodamine B dyed microspheres do indeed stem from two-photon absorption.
© 2013 Optical Society of America
1. Introduction
Two-photon photoacoustic imaging (TP-PAI) can be considered as the photoacoustic analogue of two-photon laser scanning fluorescence microscopy. In both cases a highly focused pulsed laser beam is scanned over a sample. In two-photon laser scanning fluorescence microscopy the fluorescence of the sample is recorded [1], while in TP-PAI ultrasonic signals caused by the rapid temperature rise due to the light absorption are measured. The advantages of two-photon laser scanning fluorescence microscopy in comparison with confocal microscopy are its deeper tissue penetration, less photodamage, and a better axial resolution [2]. Two-photon absorption scales quadratic with the laser intensity. If a material does only absorb via two-photon processes then the volume of absorption is localized within the focal region, leading to an axial resolution in the order of the Rayleigh length. The good axial resolution of two-photon laser scanning fluorescence microscopy allows for depth discrimination, therefore enabling three-dimensional fluorescence imaging [2]. It is expected that there exists a similar correlation between TP-PAI and single-photon photoacoustic imaging (PAI). Actually, by using non-linear photoacoustic microscopy an increase in resolution with respect to single-photon PAI was recently demonstrated [3].
Two-photon photoacoustic imaging was first claimed by Yamaoka et al. [4–6]. In these publications two-photon photoacoustic generation was demonstrated for Rhodamine B solutions which were filled into capillaries. Neither the Rhodamine B solution nor the used materials of the capillaries show single-photon absorption in the near infra-red. However, a quadratic response of the photoacoustic signal as function of incident laser intensity was reported at a wavelength of 1064nm, thus giving evidence of two-photon generated ultrasound. By employing high laser light intensities photoacoustic imaging of solid targets, i.e. gold nanoparticles dispersed in a Rhodamine B solution or copper wires, was demonstrated in [5, 7], respectively. In these publications, however, it was not proven that the photoacoustic signals stem from two- or multi-photon absorption processes.
In the present paper we demonstrate two-photon photoacoustic imaging on solid targets, i.e. on polyethylene (PE) spheres dyed with Rhodamine B. Dyes like Rhodamine B have a certain probability to make photo-mediated chemical reactions with the surrounding molecules. If such a reaction takes place the new molecule possesses a modified absorption and emission characteristic and is likely not to contribute to the luminescence or PA signal. This effect is known under the name of photobleaching (see e.g [8, 9].). In two-photon induced photoacoustics one expects a quadratic dependence of the photoacoustic signal strength on the incident light intensity. If photobleaching occurs the quadratic signal response is hampered. Eventually, when all dyes are bleached no signal response will occur. In liquid samples fresh dye solution can diffuse into the excitation volume if a reservoir which is large enough is present and photobleaching is observable at high excitation intensities only. In real applications the amount of dye is usually limited, e.g. when investigating dyed cells. To demonstrate that two-photon photoacoustic microscopy is capable of being a complementary technique to two-photon fluorescence microscopy one has to perform imaging on samples with localized dyes. To this end we use Rhodamine B dyed PE spheres where the dye is localized within the volume of the sphere and no fresh dye can diffuse into the volume. Here photobleaching becomes a major issue and only comparatively low excitation intensities can be used in order to observe a quadratic response in luminescence and photoacoustic signals. Obviously the strength of the photoacoustic signal increases with excitation intensity and too low intensities will not provide enough signal strength. On the other hand too high pulse energies may lead to photobleaching. Hence, one has to find a compromise for the excitation intensity.
In the present paper we were interested in two-photon absorption-induced photoacoustic signals only. Two-photon absorption-induced photoacoustics refers to photoacoustic signals that are directly induced via two-photon absorption (see Fig. 1) and do not stem from energy transfer from the dye to surrounding media as described e.g. in [10]. In order to verify two-photon absorption-induced photoacoustic signal generation we had to measure a quadratic response of the PA signal as a function of laser light intensity, as e.g. published in [5]. Further, we had to exclude single-photon absorption in Rhodamine B as source of the ultrasonic signals. Single-photon absorption may be caused by light generated due to second harmonic generation (SHG) [11], white light generation [12,13], or optical breakdown [14]. To exclude these effects we in situ measured the luminescence spectrum between 350nm and the region of the Rhodamine B absorbance, i.e. up to 640nm. In case of optical breakdown all light frequencies up to the plasma frequency become absorbed [15] and only the knowledge of the charge carrier density allows to judge if two-photon absorption is still dominant. SHG and white light generation are non-linear effects and thus we expect that, like for two-photon absorption, they improve depth discrimination in PAI. SHG and white light generation are effects which occur in the matrix and may indirectly excite chromophores by single-photon excitation. However, this paper is dedicated to the examination of two-photon absorption only, which directly occurs in the chromophore and is therefore less sensitive to its surrounding matrix. If we did observe one of the aforementioned effects the PA signal was considered to be single-photon induced. Additionally, we had to consider a possible temperature rise of the sample. Because of the temperature dependency of the Grüneisen coefficient Γ(T) the strength of the photoacoustic signals may vary with the temperature rise caused by the incident light intensity. This behavior can be easily confused with two-or multi-photon effects.
Fig. 1 Sketch of the physical mechanism of two-photon absorption-induced photoacoustics and luminescence. Simultaneously two photons bring an electron from the quantum mechanical ground state (thick black solid line at the bottom) to an excited quantum state. In the sketch this process is labeled: 2p excitation. When the electron returns to the ground state part of the energy is released via luminescence and the other part via thermal energy. The thermal energy is proportional to the PA signal. Alternatively, the electron could relax purely thermally to the ground state without the emission of a luminescence photon (not shown).
The paper is constructed as follows. In section 2 the expected proportionalities between photoacoustic signal and excitation laser intensity are derived. Section 3 is dedicated to the setup, the methodology and the description of the samples. In order to prove our experimental method we measured different solid and liquid samples and present the results together with theoretical fits in section 4. Finally, we demonstrate multimodal two-dimensional imaging on a Rhodamine B dyed microsphere by recording the two-photon absorption-induced photoacoustic and fluorescence signals simultaneously.
2. Theory
In this section we calculate the expected proportionalities between photoacoustic signal and excitation laser intensity. For short laser pulses, the pressure of the ultrasonic wave is proportional to the absorbed laser light energy density [16]. The absorbed energy A within a thin layer with thickness z, in which the laser diameter can be considered to be constant, is proportional to I0 – I(z), where I0 is the incident intensity and I(z) is the intensity at z. The intensity as a function of z is given by dI/dz = -(α I + β I2) where α is the single-photon absorption coefficient and β is the two-photon absorption coefficient (see e.g [17].). In case of single-photon absorption – this means that β is zero – the absorbed energy, A1, is proportional to the incident laser light intensity I0,
In case of pure two-photon absorption – here α = 0 – the absorbed energy A2 is given by:If the product β⋅z⋅I0 is much smaller than one, the absorbed energy is proportional to the square of the incident intensity. If the product is much larger than one, the absorbed energy is linear proportional to the incident intensity. For the case of a combination of single and two-photon absorption we find the proportionality of the absorbed energy A1,2 to beThese three Eqs. are only valid if the laser beam profile does not change along the direction of propagation within the particle or layer.For very high laser light intensities (the threshold is material dependent but typically above 1016W/m2 [18]) a plasma is generated. Due to plasma generation the single-photon absorption coefficient, α, increases rapidly. This consequently leads to an increase in reflectivity, as can be seen from the Fresnel equations [18, 19]. Therefore we expect that in this regime the proportionalities described in Eqs. (1)-(3) do not hold. Due to the increase in reflectivity it is expected that for very high intensities the photoacoustic signal saturates or even decreases with an increase in intensity. Such a behavior was, for example, observed in [7].
3. Method
3.1 Experimental setup
For generation of the photoacoustic and luminescence signals we used a Ti:sapphire femtosecond laser (FEMTOSOURCETM scientificTM XL 650 from Femtolasers) with a minimum pulse duration τ of 50 fs, a maximum energy per pulse E of 650 nJ, a pulse repetition frequency f of 3.968 MHz, and a center wavelength of 800 nm. The energy was regulated by a variable beam attenuator for linearly polarized light (model 990-0071-800H from Eksma Optics) and a neutral density filter and was measured by a powermeter (Fieldmate and PM10-19C from Coherent) before each experiment. For acquisition of two-dimensional images the beam was scanned over the sample by two silver-coated galvanometer mirrors (G6215H, Cambridge Technology) and focused onto the sample by a 10x objective optimized for the near infra-red (M Plan Apo, Mitutoyo). The objective had a numerical aperture (NA) of 0.26. A simplified sketch of the setup is depicted in Fig. 2. The spatial half-width of the laser spot below the objective in air, w0 = 10µm, was determined by a self-built beam-profilometer [20]. In the experiments the samples were always immersed in a water bath. In order to avoid ripples on the surface of the water we put a glass slide directly on top of the water and the excitation beam was focused through the glass slide into the water. Due to the beam broadening that was caused by glass and water the actual beam waist w was slightly larger than w0. The actual beam waist w was estimated by using the Gaussian matrix formalism [21] to be around 12µm. The pulse duration, τ, was recorded in situ by an autocorrelator (Femtometer, from Femtolasers). Measurements with the beam-profilometer and the autocorrelator showed that the pulses had spatial and temporal Gaussian shape. Luminescent light was collected by the same objective and directed to a spectrometer (StellarNet EPP2000) for in situ recording the spectra or to an avalanche photo diode (Thorlabs APD110A2) for fluorescence imaging. Light with wavelengths above 640nm was removed by broadband filters (FGB37S, Thorlabs) to protect the avalanche photodiode and the spectrometer from the high intensity laser light and to remove light originating from the femtosecond laser.
Fig. 2 Simplified schematic of the setup. Femtosecond laser pulses (red line) are focused onto the sample via an objective. Luminescence light (green) was collected with the same objective and directed to a spectrometer or to an avalanche photodiode via a cold mirror. Broadband filters protected the avalanche photodiode and the spectrometer from the high intensity laser light. The hydrophone signal was either detected via a digital scope or via a Lock-In amplifier.
3.2 Determination of the laser pulse peak intensity I0
If the average laser power P is measured with a power meter one can deduce the pulse energy E via
where f is the laser’s pulse repetition rate. If the laser pulses have spatial and temporal Gaussian shape the intensity as a function of distance from the center r and time t is given byHere τ is the pulse duration measured at full width at half maximum and w is the spatial half-width at intensity e−2. A temporal and spatial integration [22] over the total 4-dimensional space yields the pulse energyTogether with Eq. (4) which relates the pulse energy with the measured average power we calculated the peak laser intensity I0. The calculation was performed by neglecting reflections at the sample’s interfaces. Therefore the actual intensities might be slightly lower than calculated.3.3 Samples
To verify our experimental setup and evaluation method we chose samples which we expected to express the behaviors of Eqs. (1)-(3). The solid samples investigated in this paper were a carbon fiber, PE spheres dyed with Rhodamine B (RHBPMS-0.98, 90-106µm, from Cospheric), and a [111] silicon wafer. The carbon fiber showed strong absorption in the visible and the near infra-red spectral region. Therefore it was expected that the photoacoustic signal intensity shows a linear proportionality to the incident laser intensity (Eq. (1)). For the dyed microspheres a quadratic dependency, as described by Eq. (2), was expected. We expected a combination of single-photon absorption and two-photon absorption for the silicon wafer, and thus the photoacoustic signal should exhibit the dependency of Eq. (3).
The Rhodamine B dyed microspheres had nominal diameters of 98 ± 8 µm. Undyed polyethylene has its absorption peak below 200nm [23]. It was therefore expected that mainly two-photon absorption of the Rhodamine B contributes to the photoacoustic signal. Consequently, we expected a dependence on the incident laser intensity regarding to Eq. (2). The used PE spheres contained 2 weight-% of Rhodamine B [24]. Under the assumption that the two-photon cross section of Rhodamine B in polyethylene is the same as when dissolved in methanol (σ(2) = 150 GM [25]) we calculated β to be 15 × 10−12m/W for a wavelength of 800nm.
As sample which should exhibit single- and two-photon absorption we chose a silicon wafer. Silicon does weakly absorb 800nm by single-photon processes with α being in the order of 0.1 × 106 m−1 [26]; the two-photon absorption coefficient β is around 2 × 10−11 m/W [27]. The PA signal should show the proportionality as described by Eq. (3).
As additional verification of our evaluation method (see section 3.4) and for comparison with literature, we also measured a glass pipette filled with a degassed Rhodamine B/water solution (calculated β ≈7 × 10−13 m/W). Degassing was performed by boiling the solution for some minutes prior to the measurement. We expected to reproduce the quadratic dependence of the photoacoustic signal as a function of the incident laser intensity as reported in [5].
3.4 Measurement principle
All samples were put into a water bath. The water bath ensured the acoustic coupling between the sample and the hydrophone (HNC1000, from Onda) which was used to record the ultrasonic signals. The hydrophone had its peak sensitivity at around 7MHz and a −3dB bandwidth of approximately 12MHz [28]. Photoacoustic signals were either acquired directly by recording the temporal signal with a digital scope (LeCroy WaveRunner HRO 66Zi) or by measuring with a Lock-In amplifier (SR844, Stanford Research). A signal acquired with the digital scope and the corresponding calculated Fourier transform are depicted in Figs. 3(a) and 3(b), respectively. The Fourier transform of a pulse train shows equidistant lines which are located at integer multiples of the laser pulse repetition rate of 3.968MHz; the envelope, i.e. the height of the peaks, is determined by the Fourier transformed single shot response of the sample. If data acquisition is done with the Lock-In amplifier (LIA), the LIA is locked to an integer multiple of the excitation frequency, i.e. to one of the lines in Fig. 3(b), and the amplitude of the corresponding Fourier component can be accessed directly.
Fig. 3 (a) Trace of PA signal stemming from a carbon fiber recorded with a digital scope. (b) Corresponding FFT.
The output voltage V of the hydrophone, which corresponds to the photoacoustic signal, follows the instantaneous value of the ultrasonic pressure [29]. For a constant excited volume the ultrasonic pressure p0 is related to the Grüneisen coefficient Γ and to the absorbed energy A via [16]
The proportionality of the absorbed energy A for single and two-photon processes is discussed in section 2. The photoacoustic signal is the pressure as a function of time p(t) and we writewhere g(t) is a normalized function that defines the shape of the ultrasonic pulse and is the amplitude of the signal. We assumed that g(t) is constant and does not change with the incident laser intensity, i.e. the volume from which the ultrasound originates does not change. Consequently, also the characteristic shape of the Fourier transform of g(t) does not change with the laser intensity. In this case p0 is proportional to the height of the Fourier components. A similar type of data evaluation was previously employed by van Raaij [7].As the recorded PA signal is sensitive to the hydrophone alignment, it was ensured that the position of the sample and the hydrophone were kept constant during a measurement series. The only quantities that changed were the incident laser light intensity and the position of the laser focus in case images were acquired. Hence during image acquisition the signal strength changed as a function of the focus position. For the small scan fields that were employed during the measurements the variation in signal strength was found to be insignificant. The experiments were started with low laser light intensities and the intensity was increased until sample modifications, e.g. photo-bleaching of the Rhodamine B dye, became significant.
High laser irradiation might cause melting and evaporation of material. Additionally, chemical reactions with surrounding species might be initiated. Surface melting or evaporation remains visible even after the experiment and chemical reactions often cause gaseous molecules that form bubbles in water [30]. In case modifications or gas bubbles were found at the sample during or after the experiment the results were either rejected or the modifications are mentioned within the paper.
4. Experiments and discussion
The dependency of the PA signal strength as function of the incident light intensities was measured simultaneously with the luminescence. The luminescence spectra and the dependency of the PA signals are presented separately in sections 4.1 and 4.2, respectively. In section 4.3 we demonstrate TP-PAI and two-photon fluorescence imaging by recording the PA signals and fluorescence light simultaneously.
4.1 Luminescence spectra
The luminescence spectra were recorded in order to exclude SHG, white light generation, optical breakdown or thermal radiation as the origin of the ultrasonic signals. In case of SHG a line at 400nm would appear in the spectrum. For white light generation, optical breakdown or thermal radiation a broad luminescence underground is expected. In case a broadband spectrum was observed, the photoacoustic signal was not longer considered as being two-photon induced. Typical luminescence signals of a Rhodamine B solution, Rhodamine B dyed PE spheres and silicon are shown in Fig. 4. Two-photon luminescence of Rhodamine B dissolved in deionized water is shown in Fig. 4(a). Note, that the two-photon luminescence spectrum has the same shape as the single-photon luminescence spectrum [25]. The luminescence of Rhodamine B in the PE microspheres [Fig. 4(b)] occurs at around 590nm which matches with the specified value [31]. Neither for the Rhodamine B solution nor for the Rhodamine B dyed PE spheres we did observe SHG, white light generation, thermal radiation or plasma formation with the employed laser light intensities. Thus, we excluded single-photon absorption as the origin of the PA signal. We concluded that if the photoacoustic signal together with the luminescence signal scaled purely quadratic with the incident laser light intensity the absorption was solely due to a two-photon process.
Fig. 4 Luminescence spectra of (a) Rhodamine B dissolved in water, (b) Rhodamine B dyed PE spheres, (c) silicon exposed to approximately 5.3 × 1015W/m2.
The luminescence spectrum of silicon at a peak laser light intensity of approximately 5.3 × 1015W/m2 is shown in Fig. 4(c). Here we observed a weak but broad signal which is basically following the transmittance curve of the microscope objective and the filter that protected the CCD device from the laser light. When removing the silicon wafer no light emission was recorded. We observed slight modifications on the silicon surface after the experiments and concluded that the light was either generated by optical breakdown or thermal radiation at the silicon surface and not by white light generation inside the water. It was not clear to which extend two-photon induced processes contributed to the PA signal.
4.2 Dependence of the PA signal strength
Measured PA signal intensities as a function of incident laser light intensities are shown in Figs. 5(a)-5(d), for a carbon fiber, Rhodamine B solution in a glass pipette, silicon, and Rhodamine B dyed microspheres, respectively. In Fig. 5(c) the data points represent the FFT amplitudes of averaged time traces that were recorded with the digital scope over 0.3s. Data points in Figs. 5(a), 5(b) and 5(d) were acquired with the LIA and an integration time of 1s. The error bars represent the measurement noise.
Fig. 5 PA signal as a function of the incident laser peak intensity of (a) a carbon fiber and (c) silicon. PA und luminescence (LUM) signal as a function of the incident laser peak intensity of (b) Rhodamine B dissolved in water and (d) Rhodamine B dyed microspheres. Red crosses represent the PA signal and the circles the luminescence intensity. Signal amplitudes for PA and LUM were normalized to fit the same curve fitting (solid lines).
As expected we found a pure linear dependence of the PA signal and no luminescence signal for the carbon fiber at low excitation laser intensities [Fig. 5(a)]. In order not to damage the sample we applied low laser light intensities. To bring the sample into the focal spot we optimized the distance between sample and objective by maximizing the PA signal amplitude. As the PA signal strength is also dependent on the hydrophone position we cannot be sure that we accomplished a perfect adjustment. However, we can ensure that the maximum laser peak intensity at the sample was below 2 × 1014W/m2. For samples which exhibit luminescence we optimized the sample position by recording the luminescence, which is more reliable.
For the Rhodamine B solution [Fig. 5(b)] we found a quadratic dependence of the PA signal and the luminescence (LUM) signal as function of the excitation intensity. In the figure the signals were normalized to match the same curve fitting.
The PA data for the carbon fiber and the Rhodamine B solution fulfill the expectations from Eqs. (1) and (2). Additionally, the two-photon induced LUM data of the Rhodamine B solution reproduces the well known quadratic dependence (see, e.g [32].). From this we concluded that our data acquisition and evaluation method worked correctly.
For silicon we observed a sub-linear behavior [Fig. 5(c)]. A similar functionality between the PA signal and the incident laser light was reported for copper in [7]. After the experiments we observed slight modifications on the silicon surface which might have been caused by plasma formation or due to surface melting. In section 4.1 we presented the corresponding luminescence spectrum which we think originates from optical breakdown or thermal radiation. As mentioned in section 2, optical breakdown leads to an increase of the single-photon absorption coefficient and consequently to an increase in reflectivity. This behavior might explain the observed sub-linear behavior. In case of thermal radiation and surface melting another explanation could be given. As the melting point of silicon is far above the evaporation temperature of water the signals could origin from cavitations. A sub-linear dependence of the PA amplitudes caused by cavitations with respect to fluence was e.g. reported in [33].
For the Rhodamine B dyed PE spheres we observed a quadratic dependence in PA signal and in luminescence signal [Fig. 5(d)]. We tried to fit the data with Eq. (2) in order to verify the calculated two-photon absorption coefficient of 15 × 10−12m/W. The corresponding fit lies within the measurement uncertainties. Anyhow, as the measurement uncertainty was rather large we could not determine the exact two-photon coefficient with a reasonable accuracy. We can, however, be sure that the two-photon absorption coefficient of the spheres was below 150 × 10−12m/W, because otherwise the PA signal would approach the linear regime according to Eq. (2). The fit shown in Fig. 5(d) is purely quadratic.
In order to exclude that the PE became single-photon absorptive for 800nm during illumination, e.g. due to carbonization, we exposed the spheres with a laser light intensity of approximately 4 × 1015W/m2 and recorded the PA signal and the luminescence spectrum over time. We observed photobleaching which led to a decrease in luminescence. The decrease in luminescence went hand in hand with a decrease in PA signal. In case of carbonization or any other mechanism that makes the PE spheres absorptive for 800nm, the PA signal is expected to increase due to single-photon absorption. However, even after several minutes of exposure no increase in PA signal was observed. With this experiment we excluded PA signal generation resulting from single 800nm-photon absorption in PE.
Due to the high laser light intensities it was likely that the samples became locally heated up to several Kelvin even in the case of pure two-photon absorption. A change in temperature will influence the strength of the PA signal via the temperature dependent Grüneisen coefficient [34, 35]. E.g., an approximately linear dependency of the Grüneisen parameter of water was reported in [34]. For pure single-photon absorption in water the strength of the PA signal then shows a linear and a quadratic dependency as a function of the incident intensity. For the Rhodamine B/water solution we did not find any linear dependency of the photoacoustic signal. Therefore, the quadratic dependency cannot be explained by single-photon absorption and heating of the specimen. For pure two-photon absorption, and considering a temperature dependent Grüneisen coefficient, one finds a quadratic plus a quartic dependence as a function of the incident laser intensity. Within the measurement uncertainty it is possible to fit the data in Fig. 5(b) with a quadratic and an additional quartic term. The data of Fig. 5(b) is in agreement with the theory for a temperature increase up to 30K for the largest employed laser intensities. Consequently, we cannot exclude an increase in temperature. Anyhow, fitting of the data is only possible when assuming two-photon absorption. For dyed PE spheres we expect the Grüneisen coefficient to be constant or even slightly to decrease as a function of temperature [36, 37]. This would lead to a negative quartic term. For the data presented in Fig. 5(d) we did not find this behavior. We conclude that the effect is smaller than the uncertainty of the measurement. For the carbon fiber we found a pure linear dependence. No additional quadratic dependency was found and we conclude that we can neglect the effect of heating in this case. In summary we conclude that the temperature dependency of the Grüneisen coefficient was not a dominant effect in our measurements. However, for this kind of experiments one should keep the influence of the temperature dependent Grüneisen parameter in mind as one can easily confuse the resulting effect with multi-photon effects. This is especially crucial when mixed single-photon and two (or multi) photon absorption occurs in a sample.
4.3 Multimodal TP-PAI and two-photon fluorescence imaging
Finally, we demonstrated TP-PAI by imaging Rhodamine B dyed PE spheres. First experiments showed that Rhodamine B from the microspheres slightly dissolved in water. This bleeding was not observable by bare eye but was detectable by two-photon induced luminescence. In order to exclude that the photoacoustic signal stemmed from dissolved Rhodamine B we embedded the microspheres into a transparent resin (Viscovoss GTS). The acoustic impedance of the resin is close to that of polyethylene and close to that of water. The ultrasonic reflection losses at each interface were below 10%.
Two-dimensional photoacoustic and fluorescence images of a Rhodamine B dyed microsphere embedded in the resin are shown in Fig. 6. The images were acquired simultaneously; the acquisition time for the images was 8 min at a scan frequency of 30 Hz in x-direction. One line in the image was obtained by averaging over 20 x-scans. The laser peak intensity I0 in the center of the image was approximately 5 × 1014 W/m2. The image was obtained with the LIA locked on the first harmonics of the laser pulse repetition rate (3.968MHz).
Fig. 6 Two-dimensional PA image of a 100µm PE sphere containing Rhodamine B. (a) pure PA signal, (b) pure luminescence signal, (c) PA signal and luminescence signal. The sphere was embedded in a transparent resin. Photoacoustic signal generation was due to two-photon absorption.
In Fig. 6(a) the PA image is shown. In our setup the Rayleigh length for the 800nm laser light was longer than 100 µm. Therefore the sphere was excited throughout the entire intersection of the light path and the sphere. Hence, in the middle of the sphere more energy was deposited with respect to the border. We think that this caused a stronger PA signal in the center than at the border, as can be seen in Fig. 6(a). The image was low-pass filtered with a Gaussian function of 5µm FWHM (full width at half maximum), which is below the laser beam diameter, in order to reduce noise. The two-photon absorption-induced photoacoustic signals were only slightly above the noise level. One disadvantage of the present setup is that the laser excitation and the photoacoustic measurement are not confocal. Also, the fluorescence signals and the photoacoustic signals are not detected from the same direction in the present setup. We think that a confocal detection scheme where the acoustic and the optic foci are aligned will improve the sensitivity of the setup significantly. Therefore, we plan to improve the setup by employing a confocal detection scheme as e.g. presented in [38] in future.
The fluorescence of the sphere was recorded simultaneously to the PA signals and is shown in Fig. 6(b). The image displays the unprocessed luminescence signal of the 100 µm sphere. The spatial resolution of the fluorescence images was determined to be in the order of 10 µm, which corresponds to the beam waist of the laser focus. The fluorescence of the 100 µm sphere is strongest on the shell and weak in the center. A possible explanation is that the foci for excitation and luminescence light did not perfectly match. In Fig. 6(c) an overlay of both pictures is presented. The size and location of the PA image match the luminescence image.
5. Conclusion
In the presented paper we derived the expected dependencies between excitation intensity and photoacoustic signal strength for single-photon absorption, two-photon absorption and for the combination of both. Experimentally, photoacoustic signals and luminescence signals were generated by pulses from a femtosecond laser and acquired simultaneously. In order to check if our data acquisition method worked correctly we measured the photoacoustic intensity as a function of laser light intensity for several types of samples, i.e. a carbon fiber, Rhodamine B/water solution in a glass capillary, a silicon wafer, and Rhodamine B dyed polyethylene microspheres. As expected, the photoacoustic intensity of the carbon fiber showed a linear proportionality. For the Rhodamine B/water solution and the dyed microspheres a quadratic dependence was found. Here we excluded single-photon absorption as the origin of the photoacoustic signals by spectroscopic means. The photoacoustic signal of silicon showed a sub-linear dependence as a function of the incident laser light intensity.
Finally, the possibility to image dyed solids, i.e. Rhodamine B dyed polyethylene spheres, by means of two-photon absorption-induced photoacoustic imaging was demonstrated. Location and size of the sphere in the photoacoustic image and in the simultaneously recorded two-photon luminescence image matched. In such samples the dye is localized and cannot diffuse. Therefore photobleaching becomes a major issue. This limits the maximum employable excitation intensities and the measurement time. We nevertheless were able to obtain two-photon absorption-induced photoacoustic signals using a commercial unfocused hydrophone. This indicates that two-photon photoacoustic imaging could be used in real applications to gain complementary information to two-photon fluorescence microscopy.
Acknowledgments
This work has been supported by the Christian Doppler Research Association (Christian Doppler Laboratory for Photoacoustic Imaging and Laser Ultrasonics), the Federal Ministry of Economy, Family and Youth, the Austrian Science Fund (FWF), project number S10503-N20, the European Regional Development Fund (EFRE) in the framework of the EU-program Regio 13, and the federal state Upper Austria.
References and links
1. W. Denk, J. H. Strickler, and W. W. Webb, “Two-photon laser scanning fluorescence microscopy,” Science 248(4951), 73–76 (1990). [CrossRef] [PubMed]
2. P. T. C. So, “Two-photon fluorescence light microscopy,” Encyclopedia of Life Sciences, Nature Publishing Group (2002).
3. A. Danielli, K. I. Maslov, A. Garcia-Uribe, A. M. Winkler, C. Li, L. Wang, and L. V. Wang “Non-linear photoacoustic microscopy with optical sectioning, ” presented at SPIE Photonics West, San Francisco, USA, 2–7 Feb. 2013.
4. Y. Yamaoka, M. Nambu, and T. Takamatsu, “Fine depth resolution of two-photon absorption-induced photoacoustic microscopy using low-frequency bandpass filtering,” Opt. Express 19(14), 13365–13377 (2011). [CrossRef] [PubMed]
5. Y. Yamaoka and T. Takamatsu, “Enhancement of multiphoton excitation-induced photoacoustic signals by using gold nanoparticles surrounded by fluorescent dyes,” Proc. SPIE 7177, 71772A (2009). [CrossRef]
6. Y. Yamaoka, M. Nambu, and T. Takamatsu, “Frequency-selective multiphoton-excitation-induced photoacoustic microscopy (MEPAM) to visualize the cross sections of dense objects,” Proc. SPIE 7564, 75642O (2010). [CrossRef]
7. M. E. van Raaij, M. Lee, E. Cherin, B. Stefanovic, and F. S. Foster, “Femtosecond photoacoustics: integrated two-photon fluorescence and photoacoustic microscopy,” Proc. SPIE 7564, 75642E (2010). [CrossRef]
8. L. Song, E. J. Hennink, I. T. Young, and H. J. Tanke, “Photobleaching Kinetics of Fluorescein in Quantitative Fluorescence Microscopy,” Biophys. J. 68(6), 2588–2600 (1995). [CrossRef] [PubMed]
9. P. Wilhelm and D. Stephan, “Photodegradation of rhodamine B in aqueous solution via SiO2 @ TiO2 nano-spheres,” J. Photochem. Photobiol. A 185(1), 19–25 (2007). [CrossRef]
10. Y. H. Lai, C. F. Chang, Y. H. Cheng, and C. K. Sun, “Two-Photon Photoacoustic Ultrasound Measurement by A Loss Modulation Technique,” Proc. of SPIE Vol. 8581, 85812R (2013). [CrossRef]
11. P. A. Franken, A. E. Hill, C. W. Peters, and G. Weinreich, “Generation of Optical Harmonics,” Phys. Rev. Lett. 7(4), 118–119 (1961). [CrossRef]
12. R. R. Alfano and S. L. Shapiro, “Observation of self-phase modulation and small-scale filaments in crystals and glasses,” Phys. Rev. Lett. 24(11), 592–594 (1970). [CrossRef]
13. A. Brodeur and S. L. Chin, “Band-Gap Dependence of the Ultrafast White-light Continuum,” Phys. Rev. Lett. 80(20), 4406–4409 (1998). [CrossRef]
14. C. B. Schaffer, A. Brodeur, and E. Mazur, “Laser-induced breakdown and damage in bulk transparent materials induced by tightly focused femtosecond laser pulses,” Meas. Sci. Technol. 12(11), 1784–1794 (2001). [CrossRef]
15. C. Kittel, “Introduction to Solid State Physics” (Wiley, 2004).
16. M. Xu and L. V. Wang, “Photoacoustic imaging in biomedicine,” Rev. Sci. Instrum. 77(4), 041101 (2006). [CrossRef]
17. J. H. Bechtel and W. L. Smith, “Two-photon absorption in semiconductors with picoseconds laser pulses,” Phys. Rev. B 13(8), 3515–3522 (1976). [CrossRef]
18. D. F. Price, R. M. More, R. S. Walling, G. Guethlein, R. L. Shepherd, R. E. Stewart, and W. E. White, “Absorption of Ultrashort Laser Pulses by Solid Targets Heated Rapidly to Temperatures 1-1000 eV,” Phys. Rev. Lett. 75(2), 252–255 (1995). [CrossRef] [PubMed]
19. B. E. A. Saleh and M. C. Teich, Fundamentals of Photonics (Wiley, 2007).
20. G. Langer, A. Hochreiner, P. Burgholzer, and T. Berer, “A webcam in Bayer-mode as a light beam profiler for the near infra-red,” Opt. Lasers Eng. 51(5), 571–575 (2013). [CrossRef] [PubMed]
21. H. Kogelnik, “On the Propagation of Gaussian Beams of Light Through Lenslike Media Including those with a Loss or Gain Variation,” Appl. Opt. 4(12), 1562 (1965). [CrossRef]
22. I. N. Bronstein, K. A. Semendjajew, G. Musiol, and H. Mühlig, Taschenbuch der Mathematik (Harri Deutsch, 1995).
23. R. H. Partridge, “Near Ultraviolet Absorption Spectrum of Polyethylene,” J. Chem. Phys. 45(5), 1679 (1966). [CrossRef]
24. Private communication from Cospheric (e-mail from 27.02.2013).
25. N. S. Makarov, M. Drobizhev, and A. Rebane, “Two-photon absorption standards in the 550-1600 nm excitation wavelength range,” Opt. Express 16(6), 4029–4047 (2008). [CrossRef] [PubMed]
26. Properties of Silicon (INSPEC, 1988).
27. A. D. Bristow, N. Rotenberg, and H. M. van Driel, “Two-photon absorption and Kerr coefficients of silicon for 800-2200nm,” Appl. Phys. Lett. 90(19), 191104 (2007). [CrossRef]
28. Onda Certificate of Hydrophone Calibration for HNC-1000 S/N:1193 (2011).
29. P. N. T. Wells, “Ultrasonic Imaging of the human body,” Rep. Prog. Phys. 62(5), 671–722 (1999). [CrossRef]
30. C. Frez and G. J. Diebold, “Laser generation of gas bubbles: Photoacoustic and photothermal effects recorded in transient grating experiments,” J. Chem. Phys. 129(18), 184506 (2008). [CrossRef] [PubMed]
31. Product information on Rhodamin B Polyethlyene Microspheres from Cospheric, http://www.cospheric-microspheres.com/Fluorescent_Microspheres_Rhodamine_B_p/rhodamine%20b%20microspheres.htm?1=1&CartID=0 (19.03.2013).
32. C. Xu and W. W. Webb, “Measurement of two-photon excitation cross sections of molecular fluorophoros with data from 690 to 1050nm,” J. Opt. Soc. Am. B 13(3), 481 (1996). [CrossRef]
33. H. Ju, R. A. Roy, and T. W. Murray, “Gold nanoparticle targeted photoacoustic cavitation for potential deep tissue imaging and therapy,” Biomed. Opt. Express 4(1), 66–76 (2013). [CrossRef] [PubMed]
34. M. Pramanik and L. V. Wang, “Thermoacoustic and photoacoustic sensing of temperature,” J. Biomed. Opt. 14(5), 054024 (2009). [CrossRef] [PubMed]
35. A. Danielli, K. Maslov, J. Xia, and L. V. Wang, “Wide range quantitative photoacoustic spectroscopy to measure nonlinear optical absorption of hemoglobin,” Proc. SPIE 8223, 82233H (2012). [CrossRef]
36. R. E. Barker, “An approximate relation between elastic moduli and thermal expansivities,” J. Appl. Phys. 34(1), 107 (1963). [CrossRef]
37. M. Shen, W. N. Hansen, and P. C. Romo, “Thermal expansion of the polyethylene unit cell,” J. Chem. Phys. 51(1), 425 (1969). [CrossRef]
38. S. Hu, K. Maslov, and L. V. Wang, “Second-generation optical-resolution photoacoustic microscopy with improved sensitivity and speed,” Opt. Lett. 36(7), 1134–1136 (2011). [CrossRef] [PubMed]
