Open Access
Add to BibSonomy
Abstract
We discuss how to realize rigorous single pulse imaging using a fiber mode-locked laser for the purpose of ultrafast interferometric observation of fast varying dynamic objects. Sub-picosecond pulses are readily picked up in synchronization with the camera operation, allocating one pulse per frame, but rigorous ultrashort single pulse imaging is disturbed by the accumulation of amplified spontaneous emission (ASE) over the exposure time of the camera. Here, we propose four distinct methods to eliminate the ASE-accumulated disruption in the ultrashort optical gating by pulse interferometry and then evaluate their merits and limitations individually by experiments. The proposed four methods are referred to respectively as the time averaged phase modulation, unbalanced pulse overlapping, tandem pulse picking, and second harmonic generation.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Optical imaging preferably in three dimensions plays important roles for the progress of many fields of science and technology including biology, materials science and precision manufacturing [1–3]. The emphasis in optical imaging has long been put on the enhancement of lateral and axial resolutions for volumetric reconstruction of tiny objects with sharp contrast [4]. Much attention has also been paid to fast optical imaging for instantaneous observation of dynamic physical phenomena and chemical reactions [5–8]; for example, fuel combustion [9], laser-irradiated ablation [10], nuclear fusion [11] and thermal reactions [12]. For fast optical imaging, not only improved stroboscopic techniques [13,14] but ultrashort pulsed lasers are more preferably employed for sub-picosecond ultrafast dynamic imaging [15]. In addition, ultrashort pulses arranged in the form of temporal pulse train are used for pump-probe monitoring of fast phenomena with time delay control [16,17]. Furthermore, incorporating pulse chirping techniques permits ultrafast imaging without variable delay lines [18–20].
One technical issue in implementing ultrafast single pulse imaging using ultrashort pulses is how to avoid the image contrast degradation caused by the accumulation of amplified spontaneous emission (ASE) during the camera exposure time. A certain level of incoherent ASE unavoidably coexists with coherent stimulated emission during the lasing process within the gain medium of a laser cavity or amplifier. The ASE’s optical power in the time domain is not strong in itself; merely being of the order of 10−9 to the pulse peak power for a solid-state Ti:sapphire laser [21], or a 10−7 level for a fiber mode-locked laser [22]. However, in the process of fast imaging, the ASE’s power accumulates itself over the camera exposure time that is in practice long to reach up to a few milliseconds, whereas the temporal duration of a single pulse is usually shorter than 0.1 picosecond. In consequence, the accumulated ASE power within a single frame image becomes comparable to the total power of a single pulse, making it extremely difficult to achieve rigorous single pulse imaging with sharp image contrast by suppressing background noise.
In this study, we deal with rigorous single pulse imaging using a fiber mode-locked laser for the purpose of ultrafast interferometric observation of fast varying dynamic objects. As an initial step, sub-picosecond pulses are picked up in synchronization with the camera frame rate to implement ultrashort imaging using a single pulse for frame. Then four methods are proposed to eliminate the ASE’s accumulation over the camera exposure time, along with verification of their individual advantages and limitations by experiments. The four proposed methods are respectively referred to as the time averaged phase modulation, unbalanced pulse overlapping, tandem pulse picking, and second harmonic generation.
2. Ultrafast single pulse imaging
Figure 1(a) illustrates a Twyman-Green type interferometer configured in this study to evaluate the ASE’s effect on the acquired interferometric image. The light source is an Er-doped fiber mode-locked laser, emitting 100 fs optical pulses at a 100 MHz repetition rate with a 1560 nm center wavelength. Output pulses are picked up using an acousto-optic pulse picker (AOPP) and amplified through an Er-doped fiber amplifier (EDFA). The interferometric image is captured using a charge-coupled device (CCD) camera of 752 × 582 pixels with a 10.6 μm pixel size). The CCD camera operates at a 200 Hz frame rate with each frame being equipped with a 500 ns exposure time. The AOPP is regulated to operate in synchronization with the CCD camera so as to pick pulses at a 200 kHz rate with a 10 ns on-time slot for each pulse. This synchronization condition allocates 1,000 pulses per camera frame, while only a single pulse is precisely allocated for every camera exposure time slot. The pulse-to-pulse interferometer is aligned with one flat mirror (M1) being slightly tilted with respect to the other flat mirror (M2), so the resulting interference pattern produces line-shaped fringes.
Fig. 1 (a) Twyman-Green type interferometer setup to observe the amplified spontaneous emission (ASE) effect on single-pulse imaging. M: mirror, CL: collimating lens, BS: beam splitter, IL: imaging lens. (b) Interferometric fringe formed by a single pulse with ASE. The pulse duration is 100 fs while the camera exposure time τexp for a single frame is 500 nm. (c) interferometric fringe formed by ASE only.
Figure 1(b) shows the interferometric fringes formed with a single pulse being captured per camera frame. Note that the camera exposure time (500 ns) per frame is much longer, i.e., more than six orders of magnitude, than the single pulse duration (100 fs). Thus, a significant amount of ASE existing in the light source is accumulated in the camera frame. In consequence, the resulting interferometric pattern is constructed by not only the captured single pulse but also the penetrated ASE. For comparison, only ASE-dominated interference can readily be acquired by breaking the AOPP-camera synchronization, i.e. allocating no pulse at all within the camera exposure time, as depicted in Fig. 1(c). The result shows that the captured ASE is basically incoherent but it produces a significant contribution to the equal-path interference pattern, of which the fringe contrast reaches 0.31. The fringe contrast of the pulse plus ASE interferogram of Fig. 1(b) is 0.89, so the sole contribution of a single pulse can be isolated by the subtraction of 0.89 – 0.31 = 0.58. Thus, the signal-to-noise (SNR) due to ASE turns out to be 0.58 / 0.31 = 1.87, indicating that the ASE’s contribution to the equal-path interference pattern is comparable to that of a single pulse, not being ignorable if rigorous ultrashort single pulse imaging has to be achieved with clear sharp image contrast.
Single pulse interferometric imaging may be performed simply by treating the ASE-induced interference as background intensity noise and subsequently suppressing it by adopting phase-shifting techniques. This enables one to determine the interferometric phase formed only by a single pulse by taking three or four phase-shifted frames of interferometric image. However, the sequential approach is not applicable to ultrafast imaging where only a single frame of interferometric image is given in observation of a very fast time-varying object. Inevitably, for rigorous imaging using a single pulse, the ASE-induced interference has to be removed in the very process of image capturing, for which hereafter we consider four methods;
- (1) Time averaged phase modulation
- (2) Unbalanced pulse overlapping
- (3) Tandem pulse picking
- (4) Second harmonic generation
2.1 Time averaged phase modulation
Figure 2(a) depicts the method of time averaged phase modulation in which fast phase modulation is provided in an interferometer arm using an electro-optical modulator (EOM). During the long camera exposure time, the ASE-induced interference is averaged out, while the pulse-induced interference remain intact as it is formed just during the very short pulse duration. In other words, the total interferogram intensity I is regarded as a linear incoherent superposition of the intensity IA of penetrated ASE and the intensity of IP of a single pulse as
Fig. 2 Optical configurations of four distinct ways of suppressing the effect of amplified spontaneous emission (ASE) on ultrafast single pulse imaging. (a) Time averaged phase modulation. (b) Unbalanced pulse overlapping. (c) Tandem pulse picking. (d) Second harmonic generation (SHG). AOPP: acoustic-optic pulse picker, AMP: amplifier, EOM: electro-optical modulator, and other interferometer-related symbols are the same as depicted in Fig. 1.
By the EOM’s phase modulation, IA and IP are averaged as
where α and β are the signal strengths of IA and IP, respectively, ϕ is the interferometric phase, and ap and f are the modulation depth and frequency. Both the phase-modulated interferogram intensities are averaged over the camera exposure time τexp but the effective averaging time for IP is restricted to the pulse duration τp. In addition, τexp is much longer than the modulation period, so IA turns to 0 while IP remains asThe acquired interferogram is subsequently affected only by the pulse, permitting ultrafast imaging while the phase term apcos(2πfτp) may be regarded as a small negligible constant.2.2 Unbalanced pulse overlapping
The second method relies on an unbalanced pulse overlapping configuration as shown in Fig. 2(b). The mode-locked laser light source provides coherent pulses in the form of a pulse train, with each pulse being able to interfere with its neighboring pulses over a long coherence length [23]. On the other hand, the ASE’s coherence length is usually shorter than the laser cavity length, making no interference in the unbalanced overlapping configuration. Thus, when the optical path-length difference of the interferometer is unbalanced as much as the laser cavity length, only the pulse-induced interference remains intact with suppression of the ASE-induced interference. For practical implementation, two adjacent neighboring pulses (a and b in Fig. 2(b)) are picked up by the AOPP at a time and then the total interference intensity IT is expressed as
where Ia,r, Ia,m, Ib,r, Ib,m denote the non-interference background terms attributable to the individual intensities of the reference (r) and measurement (m) pulses. Besides, IASE is also the non-interference intensity caused by ASE, while Iinter represents only the interference intensity formed by two neighboring pulses. As a result, compared to the previous method of time averaged phase modulation, this unbalanced pulse overlapping method suffers a relatively high level of non-interference background that deteriorates the achievable best image contrast of the single pulse interference.2.3 Tandem pulse picking
Single pulse imaging requires high pulse energies to be detected by the image sensor. To efficiently amplify the pulse energy, a pulse picker is used to select pulses at a slow repetition rate from the mode-locked oscillator and them an optical amplifier is used to increase the pulse energy as in Fig. 1(a). In the amplification process, however, the ASE level inevitably rises and spreads over the entire temporal domain. As shown in Fig. 2(c), the method of tandem pulse picking incorporates another pulse picker AOPP2 in synchronization to the preliminary AOPP1 after the optical amplifier. This scheme reduces the ASE’s level augmented by the optical amplifier, leaving only a small amount ASE to leak during the pulse picking time that can be reduced to as short as a few nanoseconds.
2.4 Second harmonic generation
The high peak power of ultrashort pulses leads to second harmonic generation (SHG), of which the optical power P0 is proportional to the square of the original pulse intensity I as
where the factor represents the SHG coefficient of the nonlinear material in use. On the other hand, the ASE’s power is not strong enough to induce SHG and consequently removed in a selective way. As illustrated in Fig. 2(d), with a single pulse being picked up by the AOPP, only the second harmonic pulse is delivered to the interferometer as the light source. This method of second harmonic generation requires the original pulse to be prepared with a sufficient power amplification compared to other previous three methods.3. Experimental results
Figure 3 shows an experimental result of single pulse imaging, which captures the dynamic motion of thin sheet pellicle in vibration response to an external acoustic wave generated from a sound speaker. The pellicle is placed on the position of the mirror (M2) in Fig. 1(a) and its 3D surface profile is measured during plate vibration. For comparison, a super-luminescent diode (SLD) is employed as a continuous wave (CW) source while single pulse imaging is implemented using a Q-switched pulse laser with the method of tandem pulse picking technique. The exposure time of the CCD camera is set at 2.86 ms. Resulting interferograms are analyzed by adopting the well know algorithm of 2D Fourier transformation [24]. Figure 3(a) shows the interferograms obtained with the sound speaker turned off, in which spatial carrier fringes are clearly seen. However, the SLD-interferogram disappears when the speaker sends a sound wave at 600 Hz to the pellicle. It is because the plate vibration motion is averaged out during the exposure time of the CCD camera.
Fig. 3 Static and dynamic interferometric measurements of a thin-film pellicle. (a) Static interferograms measured using a super-luminescent diode (SLD) and a single Q-switched pulse extracted by tandem pulse picking. (b) Dynamic interferograms measured using the same SLD and pulse with the pellicle in vibratory motion.
On the other hand, the pulse-interferogram is clearly seen as in Fig. 3(b), enabling 3-D motion analysis on the pellicle plate by the sound wave. Figure 4 shows a sequence of plate vibration of the pellicle with a time delay of 0.33 ms in every step. The vibration frequency of the pellicle plate is identified as 600 Hz from the camera frame rate of 52.75 Hz in consideration of aliasing effect. The vibration sequence is reconstructed by capturing 100 images consecutively along with the time difference determined by the acoustic vibration period as well as the camera frame rate.
In order to demonstrate rigorous single pulse imaging, the proposed four methods were implemented with their performances being estimated in terms of the interference fringe contrast. Figure 5 shows a set of interferometric fringes obtained with either only ASE or single pulse. In the case of time averaged phase modulation given at 6 MHz, the ASE-disturbance is eliminated as shown in Fig. 5(b) with a fringe contrast of 0.28. The unbalanced pulse overlapping presented in Fig. 5(c) results in a low fringe contrast of 0.14, mainly due to the non-interferometric terms of Eq. (5) as well as the unbalanced dispersion caused by the optical path delay. In the tandem pulse picking configuration, the second AOPP was operated at 200 kHz with 10 ns pulse picking duration, synchronized with the preliminary AOPP. As a result,, the ASE-interference is significantly reduced as in Fig. 5(f) and the image by only a single pulse is successfully obtained with no detection of even a very weak level ASE at all.
Fig. 5 Interferometric fringes with only ASE or single pulse. (a&b) Time averaged phase modulation. (c&d) Unbalanced pulse overlapping. (e&f) Tandem pulse picking. (g&h) Second harmonic generation.
For the SHG imaging, the pulses from an EDFA pre-amplifier are further amplified through an EDFA main amplifier with the pulse peak power reaching 95.2 kW. Other techniques use 17.5 kW pulse peak power. As expected, the image of second harmonic light reveals no ASE-interference as shown in Fig. 5(h). The optical components used in the SHG imaging are replaced with those operating in the visible range. The spurious noise present in Fig. 5(h) is attributed to the contamination of optical components, but the fringe contrast is worked out as high as 0.90. Table 1 summarizes test results of the proposed four techniques.
Table 1. Summary of fringe contrast and background intensity for four single pulse imaging methods
In view of the fringe contrast, the tandem pulse picking and the SHG imaging are found superior to the others because they leave no background ASE intensity in the image. Although the time averaged phase modulation and unbalanced pulse overlapping yield relatively lowered fringe contrasts as shown in Table 1, they are still useful for instantaneous phase imaging with the absence of the ASE interferometric fringes. In addition, the fringe contrast can be improved by incorporating the dispersion balance between the two interferometer arms. When the non-interferometric intensity imaging is considered, the time averaged phase modulation and unbalanced pulse overlapping methods are not suitable due to high ASE background intensity, even though the background intensity may be compensated by subtraction using a pre-acquired calibration pattern.
4. Comparison of single pulse imaging methods
As discussed so far, the ASE-induced interference is found the main factor limiting the instantaneous imaging capability using single pulses. Among the four proposed methods, the time averaged phase modulation is relatively easy to implement by incorporating an electro-optic modulator to suppress the ASE-induced interference as explained in Eq. (4). However, phase-modulating only one arm in the interferometer causes unbalanced pulse dispersion between the reference and measurement arms, which consequently lowers the fringe contrast. Further, the modulation frequency should be high enough to sufficiently average out the ASE-induced interference over the camera exposure time.
The unbalance pulse overlapping requires no additional active hardware components, but the resulting fringe contrast is relatively lower compared to other methods. In order to improve the contrast, a very long unbalanced may be used as shown in Fig. 6. In this case, the amount of non-interferometric light can be reduced significantly with enhancement of the fringe contrast. For the purpose, dispersion unbalance between two arms should be compensated by the use of a dispersion compensation fiber. The unbalance interferometric configuration can be also applied to the instantaneous phase imaging only.
Fig. 6 Unbalanced pulse overlapping with a very long delay line to improve the fringe contrast by more effective suppression of the ASE temporal noise.
The tandem pulse picking method can reduce the ASE level embedded in the source light significantly even though a small amount of the ASE noise remains in the pulse picking duration that is as short as 10 ns, rendering the ASE power negligibly small as in Fig. 5(f). This method is very beneficial also to direct non-interferometric intensity imaging when high-quality AOPPs are used. Finally, the SHG imaging method is found the most effective way to eliminate the ASE-induced interference as confirmed by experiments. No background noise is observed at all by ASE. The weak power of the ASE permits second harmonic generation to be generated only by the pulse, enabling instantaneous intensity and phase imaging. However, pulse broadening by SHG should be taken into consideration so as not to limit ultrafast imaging. In addition, the use of high pulse energy increases the system cost and complexity.
5. Summary
We have discussed how to achieve rigorous ultrashort optical imaging using a single pulse emitted from a fiber mode-locked laser for the purpose of ultrafast interferometric observation of fast varying dynamic objects. Using an acousto-optic pulse picker, sub-picosecond pulses are readily rearranged in synchronization with the camera frame rate to implement ultrashort imaging using a single pulse for each frame shot. Then the accumulation of amplified spontaneous emission (ASE) over the long exposure time of the camera operation is reduced by applying four distinct methods devised to facilitate the ASE-induced problem associated with ultrafast single pulse imaging. With subsequent verification of their individual advantages and limitations by experiments, the methods are referred to respectively as time averaged phase modulation, unbalanced pulse overlapping, tandem pulse picking, and second harmonic generation. We expect the comparison results for single pulse imaging approaches can support to the ultrafast imaging, manufacturing and other related fields.
Funding
National Research Foundation of the Republic of Korea (NRF-2012R1A3A1050386) and Programme of Introducing Talents of Discipline to Universities of China (B12019).
References
1. B. Huang, W. Wang, M. Bates, and X. Zhuang, “Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy,” Science 319(5864), 810–813 (2008). [CrossRef] [PubMed]
2. C. Mann, L. Yu, C. M. Lo, and M. Kim, “High-resolution quantitative phase-contrast microscopy by digital holography,” Opt. Express 13(22), 8693–8698 (2005). [CrossRef] [PubMed]
3. M. Eisner, N. Lindlein, and J. Schwider, “Confocal microscopy with a refractive microlens-pinhole array,” Opt. Lett. 23(10), 748–749 (1998). [CrossRef] [PubMed]
4. M. J. Mlodzianoski, M. F. Juette, G. L. Beane, and J. Bewersdorf, “Experimental characterization of 3D localization techniques for particle-tracking and super-resolution microscopy,” Opt. Express 17(10), 8264–8277 (2009). [CrossRef] [PubMed]
5. C. Dunsby, Y. Gu, and P. French, “Single-shot phase-stepped wide-field coherencegated imaging,” Opt. Express 11(2), 105–115 (2003). [CrossRef] [PubMed]
6. S. Yun, G. Tearney, J. de Boer, N. Iftimia, and B. Bouma, “High-speed optical frequency-domain imaging,” Opt. Express 11(22), 2953–2963 (2003). [CrossRef] [PubMed]
7. D. W. Yang, D. Xing, S. H. Yang, and L. Z. Xiang, “Fast full-view photoacoustic imaging by combined scanning with a linear transducer array,” Opt. Express 15(23), 15566–15575 (2007). [CrossRef] [PubMed]
8. G. Rodriguez-Zurita, C. Meneses-Fabian, N. I. Toto-Arellano, J. F. Vázquez-Castillo, and C. Robledo-Sánchez, “One-shot phase-shifting phase-grating interferometry with modulation of polarization: case of four interferograms,” Opt. Express 16(11), 7806–7817 (2008). [CrossRef] [PubMed]
9. J. Hult, M. Richter, J. Nygren, M. Aldén, A. Hultqvist, M. Christensen, and B. Johansson, “Application of a high-repetition-rate laser diagnostic system for single-cycle-resolved imaging in internal combustion engines,” Appl. Opt. 41(24), 5002–5014 (2002). [CrossRef] [PubMed]
10. A. Borowiec, D. M. Bruce, D. T. Cassidy, and H. K. Haugen, “Imaging the strain fields resulting from laser micromachining of semiconductors,” Appl. Phys. Lett. 83(2), 225–227 (2003). [CrossRef]
11. T. Ditmire, J. Zweiback, V. P. Yanovsky, T. E. Cowan, G. Hays, and K. B. Wharton, “Nuclear fusion from explosions of femtosecond laser-heated deuterium clusters,” Nature 398(6727), 489–492 (1999). [CrossRef]
12. D. G. de Oteyza, P. Gorman, Y. C. Chen, S. Wickenburg, A. Riss, D. J. Mowbray, G. Etkin, Z. Pedramrazi, H.-Z. Tsai, A. Rubio, M. F. Crommie, and F. R. Fischer, “Direct imaging of covalent bond structure in single-molecule chemical reactions,” Science 340(6139), 1434–1437 (2013). [CrossRef] [PubMed]
13. M. B. Bouchard, B. R. Chen, S. A. Burgess, and E. M. Hillman, “Ultra-fast multispectral optical imaging of cortical oxygenation, blood flow, and intracellular calcium dynamics,” Opt. Express 17(18), 15670–15678 (2009). [CrossRef] [PubMed]
14. I. Grulkowski, K. Szulzycki, and M. Wojtkowski, “Microscopic OCT imaging with focus extension by ultrahigh-speed acousto-optic tunable lens and stroboscopic illumination,” Opt. Express 22(26), 31746–31760 (2014). [CrossRef] [PubMed]
15. K.-N. Joo and S.-W. Kim, “Absolute distance measurement by dispersive interferometry using a femtosecond pulse laser,” Opt. Express 14(13), 5954–5960 (2006). [CrossRef] [PubMed]
16. K. D. Rao, M. A. Choma, S. Yazdanfar, A. M. Rollins, and J. A. Izatt, “Molecular contrast in optical coherence tomography by use of a pump-probe technique,” Opt. Lett. 28(5), 340–342 (2003). [CrossRef] [PubMed]
17. C. K. Sun, H. K. Choi, C. A. Wang, and J. G. Fujimoto, “Studies of carrier heating in InGaAs/AlGaAs strained‐layer quantum well diode lasers using a multiple wavelength pump probe technique,” Appl. Phys. Lett. 62(7), 747–749 (1993). [CrossRef]
18. K. Minoshima, H. Matsumoto, Z. Zhang, and T. Yagi, “Simultaneous 3-D Imaging Using Chirped Ultrashort Optical Pulses,” Jpn. J. Appl. Phys. 33(29B), L1348–L1351 (1994). [CrossRef]
19. K. Goda, K. K. Tsia, and B. Jalali, “Serial time-encoded amplified imaging for real-time observation of fast dynamic phenomena,” Nature 458(7242), 1145–1149 (2009). [CrossRef] [PubMed]
20. K. Nakagawa, A. Iwasaki, Y. Oishi, R. Horisaki, A. Tsukamoto, A. Nakamura, K. Hirosawa, H. Liao, T. Ushida, K. Goda, F. Kannari, and I. Sakuma, “Sequentially timed all-optical mapping photography (STAMP),” Nat. Photonics 8(9), 695–700 (2014). [CrossRef]
21. S. Keppler, A. Sävert, J. Körner, M. Hornung, H. Liebetrau, J. Hein, and M. C. Kaluza, “The generation of amplified spontaneous emission in high-power CPA laser systems,” Laser Photonics Rev. 10(2), 264–277 (2016). [CrossRef] [PubMed]
22. R. Herda and O. G. Okhotnikov, “Effect of amplified spontaneous emission and absorber mirror recovery time on the dynamics of mode-locked fiber lasers,” Appl. Phys. Lett. 86(1), 011113 (2005). [CrossRef]
23. W. D. Joo, S. Kim, J. Park, K. Lee, J. Lee, S. Kim, Y.-J. Kim, and S.-W. Kim, “Femtosecond laser pulses for fast 3-D surface profilometry of microelectronic step-structures,” Opt. Express 21(13), 15323–15334 (2013). [CrossRef] [PubMed]
24. M. Takeda and K. Mutoh, “Fourier transform profilometry for the automatic measurement of 3-D object shapes,” Appl. Opt. 22(24), 3977–3982 (1983). [CrossRef] [PubMed]
