Open Access
Add to BibSonomy
Abstract
Infrared photothermal microscopy (IPM) has recently gained considerable attention as a versatile analytical platform capable of providing spatially resolved molecular insights across diverse research fields. This technique has led to numerous breakthroughs in the study of compositional variations in functional materials and cellular dynamics in living cells. However, its application to investigate multiple components of temporally dynamic systems, such as living cells and operational devices, has been hampered by the limited information content of the IP signal, which only covers a narrow spectral window (< 1 cm-1). Here, we present a straightforward approach for measuring two distinct IPM images utilizing the orthogonality between the in-phase and quadrature outputs of a lock-in amplifier, called dual-phase IR photothermal (DP-IP) detection. We demonstrate the feasibility of DP-IP detection for IPM in distinguishing two different micro-sized polymer beads.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
In the field of infrared (IR) imaging techniques, IR photothermal microscopy (IPM) stands out as a powerful tool capable of providing spatially-resolved microscopic images of chemicals resonant with the IR excitation field. By utilizing an IR-pulsed light source to induce a photothermal effect and a visible probe beam to measure the change in the photothermal lens created, IPM can achieve both molecular specificity and superior spatial resolution at sub-micrometer scales [1–5]. This capability unlocks a multitude of applications previously unattainable with conventional IR imaging techniques with a relatively low spatial resolution, such as the analysis of active pharmaceutical ingredients in tablets [6], the identification of compositional variations in photovoltaic materials [7], and even the exploration of metabolic functions within cells. IPM is particularly well-suited for elucidating molecular dynamics in biological systems. It has been used to study changes in biomolecule distribution [8], drug delivery [3], and the observation and tracking of protein vesicles and plastic micro-particles within individual living cells [8,9].
Despite its many achievements, the progress of IPM in revealing the intricacies of living systems has been hampered by several inherent limitations. One of them is the difficulty of precise spectral analysis due to the broad IR absorption spectrum of water, a major constituent of organisms. Another limitation is the slow frame rate for image acquisition, which is limited to a few seconds by the sample scanner, falling far short of achieving real-time video-rate imaging. The most critical constraint, however, is the limited acquisition of molecular information in a single scan, primarily due to the narrow spectral linewidth of the commonly employed light source, the quantum cascade laser (QCL) [9]. This limitation becomes especially pronounced when studying dynamic metabolic processes, as it requires the simultaneous capture of a broader range of IR spectral information within a short timeframe to investigate the mutual correlation of biomolecules. Significant efforts have been dedicated to developing an advanced detection scheme capable of addressing this limitation. One approach involves a substantial reduction in the time needed to capture a single image by employing widefield imaging, facilitating fast sequential acquisition of information about different molecules [10–19]. However, even in this case, it is still necessary to adjust the IR excitation frequency to target the desired IR spectral band between each measurement, leading to temporal discrepancies between individual IP images. Additionally, the optical configuration becomes more intricate compared to the point illumination scheme, resulting in significantly reduced light source intensities reaching the specimen. In this context, we have been working on a novel approach to simultaneously acquire a broader range of IR spectral information while preserving the simplicity of the previous point illumination optical setup. This would make it easily accessible for researchers across diverse fields.
Here, we introduce a novel IR photothermal (IP) detection scheme termed dual-phase IP (DP-IP) detection, which enables the simultaneous measurement of two distinct IR absorption bands in a sample. To achieve this, we leveraged the orthogonality of sinusoidal functions, which is the fundamental principle of lock-in amplification. The Ji group used a similar phase-sensitive detection method to demonstrate a two-color stimulated Raman scattering microscopy [20]. Here, the proposed approach, called dual-phase IPM, differs from the previous two-color IP detection method based on the modulation-frequency multiplexing (MFM) technique we developed [9]. In this MFM approach, temporal overlap of the two IP dynamics is, however, unavoidable due to the difference in the repetition rates of the two distinct IR pulse trains. Given that the IR absorption coefficient is a function of temperature, the temporal overlap of two IP dynamics could corrupt the measured IP signals due to temperature fluctuations during the measurement. In contrast, achieving temporal separation between them is a prerequisite for DP-IP detection, as it prevents mutual interference of the IP signals. These two techniques can also be used together to simultaneously measure multiple IR absorption bands within a sample. To validate our concept, we calculated the in-phase and quadrature outputs of lock-in amplification in response to the IR pulse-induced temperature change using a heat transfer model [9,21,22]. In line with the calculation result, experimental demonstrations were conducted on both liquid and solid samples, confirming that the DP-IP detection could simultaneously measure two distinct IR absorption bands, even when applied to imaging modality. Furthermore, we performed DP-IP imaging to distinguish between two types of polymer beads composed of different monomers, demonstrating a two-color IR photothermal imaging (IPI) modality.
2. Principle and experimental setup
2.1 Lock-in amplification of IP signal
To understand the principle of DP-IP detection, we provide brief descriptions of the detector response in IP measurement and its subsequent refinement in a lock-in amplifier (LIA). Consider a situation in which an IR excitation pulse train with a specific repetition rate induces a photothermal lensing effect around the target particle. Simultaneously, another coaxial probe beam passes through the lens and reaches the detector. The detector output reflects the intensity change of the probe beam as it passes through the sample. Fourier analysis shows that any periodic signal can be expressed as a sum of sinusoidal functions, irrespective of its specific shape [23]. Therefore, we could describe the voltage output of the detector measuring the time-dependent IR photothermal lens in the following form of Fourier series as
where ${\textrm{V}_{out}}(t )$ is the amplitude of time-dependent detector output, ${\textrm{V}_{IP}}$ is the IP signal amplitude, ${a_n}$ is the corresponding coefficient of the nth Fourier component, $\omega $ is the angular frequency, $\varphi $ is the phase, and $N(t )$ is the random noise.The detector output is fed into a LIA to isolate the IP signal from the background noise. ${\textrm{V}_{out}}(t )$ is initially multiplied by a reference sinusoidal function generated by an internal oscillator, which is synchronized with an external trigger signal operating at the same frequency as the repetition rate of the IR pulse [24]. This process, called mixing, produces the following output, ${\textrm{V}_{mix}}(t )$:
For computational convenience, we approximate ${\textrm{V}_{out}}(t )$ as a single sinusoidal function with ${\omega _{mod}}$ and ${\varphi _{IP}}$. Consequently, Eq. (2) is simplified as:
The product of sine functions in the first term of Eq. (3) can be rewritten as follows:
${\textrm{V}_{mix}}(t )$ is then passed through a low pass filter to remove the alternating current (AC) component. Since we have already known the modulation frequency (${\omega _{mod}} = {\omega _{ref}}$) of IP signal, the slow AC term representing the difference frequency (${\omega _n} - {\omega _{ref}}$) component is a direct current (DC) component. The filtered output, denoted as ${\textrm{V}_{f,1}}$, is thus expressed as
The dual-phase LIA performs the above process with another 90° phase-shifted reference replica to extract the amplitude and phase of the input signal accurately. Then, the second filtered output is represented as
By adjusting the reference signal (i.e., ${\textrm{V}_{ref}} = 2$ and ${\varphi _{ref}} = 0$), the amplitude (${\textrm{V}_{IP}})$ and phase (${\varphi _{IP}}$) of the IP signal are obtained as:
2.2 Principle of dual-phase IR photothermal detection
The two filtered outputs of the LIA, denoted as ${\textrm{V}_{f,1}}$ and ${\textrm{V}_{f,2}}$ in Sec 2.1, are conventionally referred to as the in-phase (X) and quadrature (Y) components of a signal. The X and Y outputs are orthogonal. This relationship is exploited in DP-IP detection to measure two distinct IR-resonant signals simultaneously. To demonstrate the feasibility of this concept, we simulated the transient IR photothermal effect around an IR absorber using a heat transfer model, as in our previous study [9]. The top panel of Fig. 1(A) shows two IR pulse trains with the same repetition rate (${f_{mod}}$) and period ($T$), but with a 90° phase difference. These pulse trains excite two vibrational modes in the same or different particles inside the beam focus, resulting in transient temperature changes (i.e., IP dynamics), as shown in the bottom panel of Fig. 1(A). The repetition rate, duration, and shape of the two IR pulses are set to 10 kHz, 500 ns, and rectangular, respectively. The details of simulating IP dynamics are described in Supplement 1.
Fig. 1. Schematic descriptions of the working principle and instrumental configuration of dual-phase IR photothermal microscope. (A) Simulated IP dynamics of an absorber under two IR excitation pulses with a 90° phase difference (red, blue). The temporal profiles of the two IR excitation pulse trains with the same repetition rate of fmod and the resulting temperature changes are described in the top and bottom panels, respectively. The absorber has a thermal time constant of 5 µs. The two IR excitation pulse trains have a repetition rate of 10 kHz and a pulse duration of 500 ns. Each IP dynamic was color-coded to correspond to the respective IR excitation pulses for clear differentiation. (B) In-phase and quadrature outputs calculated using the dynamics of Fig. 1(A). The colors of the output intensities correspond to those of the IR pulse. (C) Optical configuration of dual-phase IP microscope. FG, function generator; M, mirror; BS, beamsplitter; OL, objective lens; CL, condenser lens; PD, photodiode; LIA, lock-in amplifier.
In the previous demonstration utilizing dual-frequency IR excitation, slow heat dissipation in an aqueous environment caused the two IP dynamics to overlap in time [9]. However, it is straightforward to distinguish between the two IP dynamics in the time domain (indicated by the red and blue translucent areas at the bottom of Fig. 1(A)) under the experimental condition for DP-IP detection, where their phase difference is fixed at 90°. In this configuration, the rapid and selective measurement of the two IP dynamics can be achieved by exploiting the orthogonality of the two LIA outputs (X and Y) through optimal tuning of the phase for each IR pulse train.
Calculating expected outputs can be an effective way to confirm the feasibility of the concept before experimentation. We calculated the X and Y outputs of the lock-in amplified IP signal. Since the IP dynamics induced by an IR excitation is recorded as changes in detected probe beam intensity, we assumed the detector output, ${\textrm{V}_{out}}(t )$, fed to the LIA as equivalent to the simulated IP dynamics, accounting for a unit change. The signal (i.e., the detector output) is multiplied with reference sine and cosine functions to yield two mixing results for the selective extraction of the X and Y outputs, respectively. Suppose the modulation frequency of the IP signal and the frequency of reference functions are identical. In that case, each mixing result contains a DC component proportional to the IP signal amplitude, as shown in Eq. (4). These DC components could be obtained by using a low-pass filter, which is mathematically expressed using a moving average filter as follows [25]:
The first step to ensure reliable calculations is to determine the optimal value for the time constant. We calculated the X and Y outputs using Eq. (9) by varying the time constant. A single pulse train was used, and the parameters for the pulses are set to be the same as those of Fig. 1(A). In Fig. S1B, the result clearly shows that a larger time constant relative to the period of the signal is required to obtain more precise DC components as the X and Y outputs. Interestingly, applying a time constant corresponding to an integer multiple of the signal period consistently yielded the same X and Y output, regardless of its magnitude. These values match those obtained through convergence with an infinite time constant. Considering the time-consuming nature of calculations with a large time constant, we have used an integer multiple of the signal period as the time constant. We also verified the constancy of the X and Y outputs over time (Fig. S1C). To examine the dependence of the calculated outputs on the phase difference between the IR pulse train and reference functions, we calculated the X and Y outputs by varying the starting point of the IR pulse train (Fig. S1D). The amplitude of the IP signal was observed to remain constant, independent of the phase difference. At specific phase differences, denoted as ${\mathrm{\varphi }_X}$ for maximum X and ${\mathrm{\varphi }_Y}$ for maximum Y, the IP signal could be selectively measured as either the X or Y output.
We investigated whether the two interdependent IP dynamics could be simultaneously distinguished as the X and Y outputs to check the feasibility of DP-IP detection. For calculations, the identical IP dynamics simulated at the bottom of Fig. 1(A) were utilized, and the time constant was set to 1 ms. Using a single IR pulse train, the phases of both the leading and trailing IP dynamics (red and blue) were adjusted to maximize Y and X, respectively (see the left bars of Fig. 1(B)). Even in a situation employing two simultaneous IR pulse trains, the calculated X and Y outputs were confirmed to be identical to the corresponding outputs obtained with each IR excitation, validating the reasonability of our concept.
2.3 Experimental setup
The setup of dual-phase IR photothermal microscopy (DP-IPM) is depicted in Fig. 1(B), which has been slightly modified from our two-color IPM setup to allow for adjustment and locking of the phase difference between two IR pulse trains. The general optical configuration remains the same, but an additional function generator (Agilent, 33522A) was introduced to lock the phase difference precisely at 90° (one-quarter of the period, T⁄4). The dashed lines in Fig. 2(B) depict the flow of trigger signals from the function generator to the QCLs and the LIA, while the solid lines represent the flow of the measured IP signals.
Fig. 2. Analysis of LIA outputs by changing the phases of IR pulse trains. In-phase and quadrature outputs (X and Y) were measured by varying the phases of the LIA (A) and the function generator (B). Black and red lines represent the X and Y outputs, respectively, and the amplitude of IP signals is plotted at the bottom. The data were averaged for 4 ms at each phase. Phase stepsize, 100 ns. (C) IR pulse trains optimized for DP-IP detection. The position of the detector is marked in Fig. 1(C). The MCT outputs measuring the IR pulse trains were recorded with an oscilloscope. (D) Temporal intensity profiles of X and Y outputs and amplitude under different IR excitation conditions. The IR excitation beams were illuminated during specific time intervals indicated in yellow, with the marks Q1 and Q2 denoting IR excitation of QCL1 and QCL2, respectively.
The two outputs of the function generator, which are rectangular-shaped electrical pulses with frequency ${f_{mod}}$, are directed to the trigger input channels of each QCL. Their frequencies are set in coupling mode, ensuring the two QCL pulse trains are phase-locked. One of the two outputs also serves as a trigger signal for the LIA to generate an internal reference for phase-sensitive detection.
3. Experimental results and discussion
3.1 Phase orthogonality of IP signals generated by two QCLs
To achieve the desired 90° phase difference between the two IP signals for DP-IP detection, we examined the in-phase (X) and quadrature (Y) outputs of the IP signal by changing the phase of each trigger fed to the QCLs. We used a mixture of acetone and hexane, as these liquids are miscible and have distinct IR absorption bands within the IR spectral windows of our setup (see Fig. S2).
Figure 2(A) shows the two orthogonal outputs of the LIA in response to IR excitation from QCL1, measured by varying the phase difference between the IP signal and the internal reference of the LIA. Adjusting the phase of the IR pulse train while fixing that of the LIA reference appears to be a more general approach. However, it was unfeasible within the constraint of the current IPM setup, wherein both QCL1 and the LIA should share a common trigger (see Fig. 1(C)). We thus measured the phase-dependent IP signal by utilizing the internal phase adjustment of the LIA, instead of changing the phase of the trigger for QCL1. If the phase is scanned in the opposite direction, both methods were confirmed to yield the same results (Fig. S3). The IR excitation frequency (repetition rate: 25 kHz) of QCL1 was adjusted to 1,723 cm-1 to excite the acetone carbonyl (C = O) stretch mode. The X and Y outputs exhibited sinusoidal waveforms with a 90° phase difference. The amplitude of the IP signal ($= \sqrt {{X^2} + {Y^2}} $) is independent of the phase setting (bottom of Fig. 2(A)). The X and Y were maximized at the phases of 18 and 28 µs, respectively. This suggests that they could represent the optimal phases for the reference function of the LIA to align the IP signal with the in-phase and out-of-phase relative to them, respectively. The two values that maximize the X and Y outputs are denoted as ${\mathrm{\varphi }_{X1}}$ and ${\mathrm{\varphi }_{Y1}}$, respectively.
We also investigated the phase-dependent X and Y outputs of the IP signal with IR excitation from QCL2, where the IR excitation frequency (repetition rate: 25 kHz) was set to 2,860 cm-1 to excite the hexane C-H stretch mode (Fig. 2(B)). Note that the phase of the LIA was fixed at ${\mathrm{\varphi }_{X1}}$, allowing the IP signal from QCL1 to be measured at the X output. The X and Y outputs exhibited similar results as before, with only the phase being different. In this case, the X and Y components reached their maximum values at phases of 27.5 and 37.5 µs, respectively, designated as ${\mathrm{\varphi }_{X2}}$ and ${\mathrm{\varphi }_{Y2}}$. Figure 2(C) displays the IR pulse trains measured with a HgCdTe (mercury cadmium telluride, MCT) detector located in the middle of the beam path after the spatial overlap of all three beams, with the phases of the LIA and the trigger signal for QCL2 set to ${\mathrm{\varphi }_{X1}}$ and ${\mathrm{\varphi }_{Y2}}$, respectively. The phase difference between the detected IR pulse trains was observed to be 90° (10 µs in the time domain), the desired condition for DP-IP detection. We shall refer to this condition as dual-phase IR excitation.
Furthermore, we demonstrated whether the two IP signals at different times could be independently and simultaneously measured as X and Y outputs. Figure 2(D) presents the X and Y outputs measured under varying IR excitation conditions over time. As shown in the yellow region of Q1 and Q2 corresponding to single IR excitation, the X and Y outputs were confirmed to selectively respond to QCL1 and QCL2, respectively. Even in the stage of dual-phase IR excitation (indicated as Q1 + Q2), IP signals for QCL1 and QCL2 were independently measured as X and Y outputs. These values were confirmed to be unchanged compared to the results from adjacent single excitations.
3.2 DP-IP measurement on micro-sized polymer particle
The DP-IP detection scheme was subsequently applied to measure two distinct molecular vibrations, separated in the spectral domain, in a 1 µm polystyrene (PS) bead. Two IR absorption bands (aromatic C = C and C-H stretching vibrations of PS) were within the IR spectral range of our DP-IP setup (Fig. S2). A single PS bead was subjected to DP-IP measurements, setting the IR frequencies of QCL1 and QCL2 to 1,600 cm-1 and 2,850 cm-1 to excite the aromatic C = C and CH2 symmetric vibrations, respectively. To compensate for the effect of potential experimental variations caused by sample change, we slightly adjusted the phase values (${\mathrm{\varphi }_{X1}}$ and ${\mathrm{\varphi }_{Y2}}$) to maximize the two independent IP signals into X and Y outputs, as demonstrated in Fig. 2(A) and 2(B).
We measured the X and Y outputs within a 5 µm by 5 µm area encompassing the PS bead, constructing IP images of the PS bead under three different IR excitation conditions (Fig. 3(A)). As depicted in the left and middle column images, the IP signals generated by two distinct single IR excitations (with a 90° phase difference) were successfully isolated into X and Y outputs during sample scanning. Under dual-phase IR excitation (right column), two distinct IP images corresponding to the IR responses of 1,600 and 2,850 cm-1, respectively, were simultaneously acquired.
Fig. 3. Dual-phase IR photothermal measurement on a solid sample. (A) IP images of 1 µm polymer bead constructed using X and Y outputs with three different IR excitations. Time constant, 1 ms. Stepsize, 50 nm. Pixel dwell time, 4 ms. Probe power, 15 mW. IR excitation power, 2 mW (at 1,600 cm-1) and 0.35 mW (at 2,850 cm-1). Scale bars, 1 µm. (B) Line-cut profiles of the PS beads marked in (A) (LC1, LC2, and LC3). (C) Comparison of X and Y output intensities between single and dual-phase IR excitations at the center of the PS bead. Each data was averaged for 10 s, and the error bars represent the standard deviation of the IP signal intensities. (D) IP spectra of PS bead obtained with single and dual-phase IR excitations. In DP-IP measurement, the IR frequency of QCL2 was only scanned to obtain the spectrum, while that of QCL1 was fixed at 1,600 cm-1. Each data point in the IP spectra was obtained by 50-ms averaging. Stepsize, 1 cm-1.
Figure 3(B) displays the line-cut profiles of the PS bead indicated by yellow dotted lines in Fig. 3(A). We found that the line-cut profiles of dual-phase IR excitation (LC3 of X and Y outputs) were consistent with those of the corresponding single IR excitations (LC1 of X output and LC2 of Y output, respectively). These experimental results demonstrate the applicability of the DP-IP detection scheme to imaging modalities. We also confirmed the long-term stability of DP-IP detection by measuring the X and Y outputs at the center of the PS bead for 10 seconds (Fig. 3(C)). We further measured the IP spectra associated with CH2 vibrations of the PS bead with single IR excitation and dual-phase IR excitation conditions (Fig. 3(D)). The two IP spectra showing consistent results irrespective of the presence of the other IR excitation (1,600 cm-1) confirm the stability and usefulness of DP-detected IPM.
3.3 Identification of polymer beads with DP-IPI
Finally, we applied this DP-IP detection scheme to acquire two distinct IR-resonant images of a polymer bead mixture simultaneously. Two polymer microspheres were used: poly(methyl methacrylate) (PMMA) and PS. Figure 4(A) shows the bright-field image of the sample, but there are no noticeable differences between the beads. Previously developed two-color IPI using the MFM technique was performed on the regions indicated in Fig. 4(A). Two distinct IR excitation frequencies were applied with different modulation frequencies: 1,725 cm-1 (55 kHz) for the carbonyl (C = O) group in PMMA and 2,850 cm-1 (45 kHz) for the methylene (-CH2-) group in PS (Fig. 4(B) and S2). The dual-frequency IP image obtained with the MFM-IP imaging technique could exhibit clearly distinguished PMMA and PS beads (Fig. 4(B)), where the red and green false colors indicate IPI contrasts of 1,725 cm-1 and 2,850 cm-1, respectively.
Fig. 4. Identification of polymer beads with dual-phase IR photothermal imaging. (A) Bright-field image and (B) dual-frequency IP images of a mixture sample containing PMMA and PS microspheres (with a diameter of 1 µm). IR excitation frequencies of 1,725 cm-1 (55 kHz) and 2,850 cm-1 (45 kHz) were employed to differentiate between PMMA and PS microspheres. Time constant, 1 ms. Stepsize, 50 nm. Pixel dwell time, 4 ms. Probe power, 15 mW. IR excitation power, 0.15 mW (at 1,725 cm-1) and 0.35 mW (at 2,850 cm-1). Scale bars, 3 µm. (C) Respective in-phase (X) and quadrature (Y) images under different IR excitation conditions and (D) corresponding dual-phase IP images. IR excitation powers and frequencies used were the same as those in Fig. 4(B). The first and second rows in Fig. 4(C) represent the IP images constructed using the X and Y outputs, respectively. Each column corresponds to different IR excitation conditions. Time constant, 1 ms. Stepsize, 50 nm. Pixel dwell time, 4 ms. Scale bars, 3 µm.
Next, dual-phase IPI measurements were performed on the same region under three different IR excitation conditions (Fig. 4(C)). The modulation frequencies for both QCLs were fixed to 50 kHz, and the phase values were the same as those used in Fig. 3. Under two different single IR excitations (first two columns), each IP image generated by the respective IR excitation was exclusively obtained in the X and Y outputs, respectively. The PMMA and PS beads were clearly distinguishable in the two individual IP images, where the IPI contrasts were consistent with those in Fig. 4(B). The DP-IPM images were simultaneously acquired (last column), and merged for comparison (Fig. 4(D)). The color scheme is the same as in Fig. 4(B). This demonstrates consistency with the previous dual-frequency IPM, confirming the preservation of chemical selectivity. Based on these findings, it is evident that DP-IPM holds promise for multi-contrast label-free imaging.
3.4 Discussion
In this study, we demonstrate that the DP-IP detection scheme implemented in the IPM enables the simultaneous monitoring of two distinct IR absorption bands. This approach harnesses the orthogonality of in-phase and quadrature outputs in an LIA to distinguish between the two sequentially generated IP dynamics. By carefully controlling the phase condition of the utilized IR excitation pulse trains, we could achieve precise temporal separation of the two distinct IP dynamics. This separation effectively mitigates the mutual interference between the two IP signals that were simultaneously measured due to temperature changes, which was noted as a limitation of the previous two-color IP measurements based on MFM. In this regard, DP-IP detection has the potential to advance the field of molecular analysis based on IP measurements by providing more precise molecular information.
Unfortunately, the current optical configuration of DP-IPM is not suitable for studying samples with spatial heterogeneity in thermal diffusivity. As mentioned earlier and demonstrated in our experiments, it is essential to precisely control and firmly lock the relative phases of the IP signals with respect to the reference signal generated in the LIA during DP-IP measurement. However, the phase or time-dependent rise-and-decay pattern of the IP signal, demodulated by the LIA, may change when measuring heterogeneous specimens due to spatial variations in heat transfer or diffusion properties. Note that the phase information of the IP signal has been considered as an alternative imaging contrast for spatially mapping the thermal diffusivity instead of the signal amplitude for molecular distribution [26–29].
One potential solution to overcome this limitation is to replace the continuous wave (CW) probe beam with a pulsed one, having a duration significantly shorter than the temporal scale of both the IR excitation pulse and the resulting IP dynamics. Several studies have already demonstrated the feasibility of time-resolved measurements of IP dynamics using temporally short visible probes [11,12,14,18]. In such an optical configuration, we could selectively probe temperature changes at specific time points. Note that our primary focus is on the temperature change immediately after the end of the IR pulse duration, which represents the peak temperature during IP dynamics. Thus, the optimal phase condition determined by using visible probe pulses instead of a CW beam will be independent of spatial variations in heat diffusion properties within the sample, e.g., cytoplasmic matrix in cells.
4. Conclusion
In summary, we developed a novel IR photothermal microscopy technique that allows for the simultaneous measurement of two distinct IR photothermal signals. This achievement was made possible by exploiting the working principles of a lock-in amplifier and the orthogonal properties of its two outputs. Our results demonstrate that the developed DP-IP detection, when applied to molecular imaging studies, enables the simultaneous acquisition of two different pieces of molecular information without signal interference due to the temporal overlap of two independent IR photothermal signals. We believe that the DP-IP detection scheme will expand the applicability of various IR photothermal spectroscopy and microscopy techniques and enhance their potential for various applications.
Funding
Institute for Basic Science (IBS-R023-D1).
Disclosures
The authors declare no competing financial interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. E. S. Lee and J. Y. Lee, “Nonlinear optical infrared microscopy with chemical specificity,” Appl. Phys. Lett. 94(26), 261101 (2009). [CrossRef]
2. R. Furstenberg, C. A. Kendziora, M. R. Papantonakis, et al., “Chemical Imaging using Infrared Photothermal Microspectroscopy,” Proc. SPIE 8374, 837411 (2012). [CrossRef]
3. D. Zhang, C. Li, C. Zhang, et al., “Depth-resolved mid-infrared photothermal imaging of living cells and organisms with submicrometer spatial resolution,” Sci. Adv. 2(9), e1600521 (2016). [CrossRef]
4. Z. Li, K. Aleshire, M. Kuno, et al., “Super-Resolution Far-Field Infrared Imaging by Photothermal Heterodyne Imaging,” J. Phys. Chem. B 121(37), 8838–8846 (2017). [CrossRef]
5. S. Adhikari, P. Spaeth, A. Kar, et al., “Photothermal Microscopy: Imaging the Optical Absorption of Single Nanoparticles and Single Molecules,” ACS Nano 14(12), 16414–16445 (2020). [CrossRef]
6. C. Li, D. Zhang, M. N. Slipchenko, et al., “Mid-Infrared Photothermal Imaging of Active Pharmaceutical Ingredients at Submicrometer Spatial Resolution,” Anal. Chem. 89(9), 4863–4867 (2017). [CrossRef]
7. R. Chatterjee, I. M. Pavlovetc, K. Aleshire, et al., “Subdiffraction Infrared Imaging of Mixed Cation Perovskites: Probing Local Cation Heterogeneities,” ACS Energy Lett. 3(2), 469–475 (2018). [CrossRef]
8. J. M. Lim, C. Park, J. S. Park, et al., “Cytoplasmic Protein Imaging with Mid-Infrared Photothermal Microscopy: Cellular Dynamics of Live Neurons and Oligodendrocytes,” J. Phys. Chem. Lett. 10(11), 2857–2861 (2019). [CrossRef]
9. C. Park, J. M. Lim, S. C. Hong, et al., “Two-color infrared photothermal microscopy,” Analyst 148(10), 2395–2402 (2023). [CrossRef]
10. K. Toda, M. Tamamitsu, Y. Nagashima, et al., “Molecular contrast on phase-contrast microscope,” Sci. Rep. 9(1), 9957 (2019). [CrossRef]
11. Y. Bai, D. Zhang, L. Lan, et al., “Ultrafast chemical imaging by widefield photothermal sensing of infrared absorption,” Sci. Adv. 5(7), eaav7127 (2019). [CrossRef]
12. D. Zhang, L. Lan, Y. Bai, et al., “Bond-selective transient phase imaging via sensing of the infrared photothermal effect,” Light: Sci. Appl. 8(1), 116 (2019). [CrossRef]
13. M. Schnell, S. Mittal, K. Falahkheirkhah, et al., “All-digital histopathology by infrared-optical hybrid microscopy,” Proc. Natl. Acad. Sci. U.S.A. 117(7), 3388–3396 (2020). [CrossRef]
14. M. Tamamitsu, K. Toda, H. Shimada, et al., “Label-free biochemical quantitative phase imaging with mid-infrared photothermal effect,” Optica 7(4), 359–366 (2020). [CrossRef]
15. Y. Zhang, C. Yurdakul, A. J. Devaux, et al., “Vibrational Spectroscopic Detection of a Single Virus by Mid-Infrared Photothermal Microscopy,” Anal. Chem. 93(8), 4100–4107 (2021). [CrossRef]
16. E. M. Paiva and F. M. Schmidt, “Ultrafast Widefield Mid-Infrared Photothermal Heterodyne Imaging,” Anal. Chem. 94(41), 14242–14250 (2022). [CrossRef]
17. J. Yin, M. Zhang, Y. Tan, et al., “Video-rate mid-infrared photothermal imaging by single-pulse photothermal detection per pixel,” Sci. Adv. 9(24), eadg8814 (2023). [CrossRef]
18. G. Ishigane, K. Toda, M. Tamamitsu, et al., “Label-free mid-infrared photothermal live-cell imaging beyond video rate,” Light: Sci. Appl. 12(1), 174 (2023). [CrossRef]
19. K. Kniazev, E. Zaitsev, S. Zhang, et al., “Hyperspectral and Nanosecond Temporal Resolution Widefield Infrared Photothermal Heterodyne Imaging,” ACS Photonics 10(8), 2854–2860 (2023). [CrossRef]
20. R. He, Y. Xu, L. Zhang, et al., “Dual-phase stimulated Raman scattering microscopy for real-time two-color imaging,” Optica 4(1), 44–47 (2017). [CrossRef]
21. I. M. Pavlovetc, E. A. Podshivaylov, R. Chatterjee, et al., “Infrared photothermal heterodyne imaging: Contrast mechanism and detection limits,” J. Appl. Phys. 127(16), 165101 (2020). [CrossRef]
22. J. Yin, L. Lan, Y. Zhang, et al., “Nanosecond-resolution photothermal dynamic imaging via MHZ digitization and match filtering,” Nat. Commun. 12(1), 7097 (2021). [CrossRef]
23. A. Pinkus and S. Zafrani, Fourier Series and Integral Transforms (Cambridge University Press, 1997)
24. A. Mandelis, “Signal-to-noise ratio in lock-in amplifier synchronous detection: A generalized communications systems approach with applications to frequency, time, and hybrid (rate window) photothermal measurements,” Rev. Sci. Instrum. 65(11), 3309–3323 (1994). [CrossRef]
25. C. A. West, S. A. Lee, J. Shooter, et al., “Nonlinear effects in single-particle photothermal imaging,” J. Chem. Phys. 158(2), 024202 (2023). [CrossRef]
26. O. O. Dada, P. E. Feist, and N. J. Dovichi, “Thermal diffusivity imaging with the thermal lens microscope,” Appl. Opt. 50(34), 6336–6342 (2011). [CrossRef]
27. P. D. Samolis and M. Y. Sander, “Phase-sensitive lock-in detection for high-contrast mid-infrared photothermal imaging with sub-diffraction limited resolution,” Opt. Express 27(3), 2643–2655 (2019). [CrossRef]
28. P. D. Samolis, D. Langley, B. M. O’Reilly, et al., “Label-free imaging of fibroblast membrane interfaces and protein signatures with vibrational infrared photothermal and phase signals,” Biomed. Opt. Express 12(1), 303–319 (2021). [CrossRef]
29. P. D. Samolis, X. Zhu, and M. Y. Sander, “Time-Resolved Mid-Infrared Photothermal Microscopy for Imaging Water-Embedded Axon Bundles,” Anal. Chem. 95(45), 16514–16521 (2023). [CrossRef]
