Open Access
Add to BibSonomy
Abstract
Confocal Raman microscopy (CRM) has found applications in many fields as a consequence of being able to measure molecular fingerprints and characterize samples without the need to employ labelling methods. However, limited spatial resolution has limited its application when identification of sub-micron features in materials is important. Here, we propose a differential correlation-confocal Raman microscopy (DCCRM) method to address this. This new method is based on the correlation product method of Raman scattering intensities acquired when the confocal Raman pinhole is placed at different (defocused) positions either side of the focal plane of the Raman collection lens. By using this correlation product, a significant enhancement in the spatial resolution of Raman mapping can be obtained. Compared with conventional CRM, these are 23.1% and 33.1% in the lateral and axial directions, respectively. We illustrate these improvements using in situ topographic imaging and Raman mapping of graphene, carbon nanotube, and silicon carbide samples. This work can potentially contribute to a better understanding of complex nanostructures in non-real time spectroscopic imaging fields.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Confocal Raman microscopy (CRM) is widely used in materials science [1–3], biomedicine [4–6], and physical chemistry [7–10] owing to its ability to provide molecular fingerprint and tomography information using non-destructive detection. However, the intensity of Raman scattered light is 10−6–10−8 times weaker than that which is Rayleigh scattered. Thus, conventional CRM often uses large pinholes (∼hundreds of micrometers) to obtain high signal-to-noise (SNR) spectra. This significantly reduces the spatial resolution of the technique and hinders applications in the micro/nano area, e.g., examination of microcomponents and nanostructures. Therefore, improving the spatial resolution of Raman mapping has become an increasingly active research area.
Typical methods of improving the spatial resolution of Raman mapping are divided into areas associated with near-field [7,11–14], far-field [15–19], and image restoration technologies [20–25]. In the near-field Raman area, researchers combined scanning probe microscopy e.g., AFM, STM, with Raman spectroscopy to improve the spatial resolution. Liu et al. used tip-enhanced Raman spectroscopy (TERS) to study single Co(II)–tetraphenylporphyrin molecules and so realized atomic resolution Raman mapping [13]. Chen et al. used scanning tunneling microscope based tip-enhanced Raman spectroscopy (STM-TERS) to study individual carbon nanotubes and achieved 1.7 nm spatial resolution [14]. However, these techniques have practical limitations, including high instrument complexity, small dynamic range, strict requirements for surface roughness, and strictly controlled measurement environment.
In the far-field area, beam shaping is used to get a high spatial resolution [15–19]. For example, Rieger et al. proposed depletion of the vibrational ground-state by using a doughnut-shaped beam to suppress spontaneous Raman scattering in a spatially defined area; this improved the spectral spatial resolution by a factor of two [15,16]. However, it is unsuitable for all systems as the suppression efficiency and resolution enhancement strongly depend on molecular properties, such as the excited-state lifetime and absorption cross-section. Watanabe et al. employed structured illumination Raman microscopy to reach the theoretical resolution limit [17,18]. However, since spontaneous Raman scattering is typically weak, the separation of high-frequency spatial components in Fourier space is difficult. Li et al. introduced radially polarized light compact-focusing technology and image restoration technology into CRM to get the spatial resolution to around 160 nm [19]. However, it reducing the excitation energy of the focal spot and so leading to a reduction in signal intensity.
As an alternative to instrumental methods, image restoration including a regularized super-resolution algorithm [21], a deep learning algorithm based on U-Net convolutional neural network [22], and the interior point least squares (IPLS) method [23], are used in Raman mapping and have also become popular recently as a means improve the spatial resolution and image quality. These methods have good compatibility with existing systems since often neither the system acquisition conditions, nor sample format need be change. However, they do nevertheless require sufficiently high SNR data to generate high quality images. Furthermore, the lack of focus tracking ability in conventional CRM leads to uncertainties in the processed images, resulting in pseudo-detail information and even restoration errors.
In the work reported here, we now present the differential correlation-confocal Raman microscopy (DCCRM) technique, which is not only suitable for far-field measurement, but also operates without loss of Raman excitation intensity or imposition of specific image quality requirements.
The DCCRM method implements both accurate focus tracking and high spatial resolution topographic imaging by using the reflected and Rayleigh scattered light as previously described in our differential confocal work [26,27], and obtains an improvement of over 20% in the spatial resolution of Raman mapping by the correlation product method applied to the collected Raman scattering intensities. Taken together, this leads to in-situ topographic imaging and Raman mapping with high spatial resolution.
2. Experimental methods
2.1 Differential correlation-confocal Raman microscopy method
A schematic of the DCCRM system is shown in Fig. 1. The DCCRM system comprises a differential confocal system (as the reflected light path from the notch filter) and a correlation-confocal Raman spectroscopy system (as the transmitted path through the notch filter). After the sample is excited, the reflected light and Raman scattered light are collected along the original light path through objective and separated by notch filter. In the differential confocal system, the reflected light is reflected by notch filter and beam splitter, converged through L1 and L2 and transmitted through PH1 and PH2, which are placed in front and behind of the focal plane with the same offset. Signals are then detected by PMT1 and PMT2. The reflected light intensity detected by the two detectors are processed by differential subtraction to achieve high-precise axial focusing capability and high spatial resolution topographic imaging [26,27]. In correlation-confocal Raman spectroscopy system, two Raman spectra are obtained by offsetting the Raman pinhole by small amounts (uRM) either side of the focal plane of the collecting lens LR1 that is one of the elements of the focusing optics in front of the Raman spectrometer. These two Raman spectra are then processed using the correlation product method to compress the point spread function of Raman system. This improvement on Raman mapping has not been found in our previous studies [26,27].
Fig. 1. Differential correlation-confocal Raman microscopy (DCCRM). (a) Schematic of DCCRM, where NF is notch filter, PZT is Piezo-Actuated objective scanner, BS is the beam splitter, PH is the pinhole, PMT is the photomultiplier tubes, A and B are the positions of the Raman pinholes where two Raman spectra are obtained; (b) three-dimensional (3D) Raman spectra intensity point spread function (IPSF) of IRA (v, u, +uRM); (c) 3D IPSF of IRB (v, u, −uRM); (d) 3D IPSF of IRC (v, u); and (e) 3D IPSF of IRR (v, u, uRM).
To improve the axial focusing capability of the DCCRM described above, two intensity response signals I1 (v, u, −uM) and I2 (v, u, +uM) received from PMT1 and PMT2 were processed by differential subtraction. Here the differential confocal axial intensity response curve ID (v, u, uM) is shown as Eq. (1).
After focusing on the sample position of interest, Raman spectra are collected when the Raman pinhole is moved to positions either side of the focal plane of the collection lens LR1. The intensities of the two Raman spectra (IRA (v, u, +uRM) and IRB (v, u, −uRM)) can be expressed as Eq. (2).
Thus, to improve the spatial resolution of CRM, we processed Raman scattering intensity responses IRA (v, u, +uRM) and IRB (v, u, −uRM) using a correlation product. This leads to the correlated intensity responses IRR(v, u, uRM) being expressed as Eq. (5).
2.2 Simulation of differential correlation-confocal Raman microscopy
To illustrate improvements in the lateral and axial resolutions that are hypothesized using the above method, simulations using different offset values of uRM were performed evaluate the properties of a DCCRM system. As discussed in the section 2.1, previous studies using differential CRM have not improved the spatial resolution of Raman mapping compared with CRM. For simplicity, in the following comparison of CRM and DCCRM based techniques, the spectroscopic spatial resolutions simulated for conventional CRM systems will be used (since these are similar to those that also employ differential CRM).
The normalized lateral response curve IRR (v, 0, uRM) of DCCRM and that IRC (v, 0) of CRM are shown in Fig. 2(a). The full width at half maximum (FWHM) of DCCRM for different offset valuve uRM are indicated in Fig. 2(b).
Fig. 2. (a) Comparison of normalized lateral response curves between CRM and DCCRM with different offset values of uRM; (b) FWHM of the normalized DCCRM lateral response curve for different offset value uRM.
Figure 2(a) shows that the main lobe of the spectral lateral response curve IRR (v, 0, uRM) of DCCRM is sharper than that of IRC (v, 0) of CRM. As illustrated in Fig. 2(b), the FWHM of IRR (v, 0, uRM) increases as the offset value uRM increases. When uM ≤ 5, the FWHM of IRR this change is gradual, whereas for uRM > 5, the change FWHM of IRR with uRM is more dramatic. Therefore, simulations suggest the normalized offset uRM should be ≤ 5.
Similarly, the normalized axial response curve IRR (0, u, uRM) of DCCRM and IRC (0, u) of CRM are shown in Fig. 3(a) with the FWHM of DCCRM for different offset value uRM being plotted in Fig. 3(b).
Fig. 3. (a) Comparison of normalized axial response curves between CRM and DCCRM with different offset value uRM. (b) FWHM of the normalized DCCRM axial response curve for different offset value uRM.
Figure 3(a) again shows that the main lobe of the spectral axial response curve IRR (0, u, uRM) of DCCRM is sharper than that of the spectral axial response curve IRC (0, u) of CRM. However, in contrast to the lateral response dependency on uRM, for the axial response, the FWHM of IRR (0, u, uRM) decreases as uRM increases (Figure 3(b)). Furthermore, significant sidelobes of appear in the IRR (0, u, uRM) response uRM > 7; these sidelobes would degrade the SNR of DCCRM measurments. Thus from the perspective of axial resolution, the operating range should be 0 < uRM ≤ 7.
Since the axial and lateral FWHM curves of Fig. 2(b) and Fig. 3(b) have opposite dependencies on uRM displacements, to balance optimal improvements in each of the lateral resolution, axial resolution, and Raman scattering intensity a normalized offset is uRM = 5 was chosen. Thus, the lateral and axial normalized intensity response curves of DCCRM and CRM for this value of uRM are shown in Fig. 4. Here, the improvements in the lateral and axial resolution of DCCRM (compared to CRM) suggested by these simulations are 28.6% and 34.6%, respectively.
Fig. 4. Comparison of normalized intensity response curves between CRM and DCCRM: (a) Normalized lateral intensity responses; (b) Normalized axial intensity responses.
To further illustrate the hypothesized improvement of DCCRM's resolution compared with CRM, we carried out simulations in which the IPSF was applied to a pattern (image) comprising features of differing Raman scattering intensities (Figure 5). Figure 5(a) is the original image and its corss-section profile of the dot line, which comprises four white and two orange stripes. The intensity of the white stripes is twice that of the orange stripes. The result of the DCCRM and CRM IPSF’s convolved with the original image results in the degraded images Figures 5(b) and 5(c).
Fig. 5. Images and their cross-section profiles (at the dotted line) comparing the measured Raman scattering intensities expected when using CRM and DCCRM methods: (a) Original sample intensity pattern its cross-section profile; (b) Degraded image its cross-section profile when using CRM; (c) Degraded image its cross-section profile with DCCRM.
For each of the images in Figure 5, a cross section profile is shown to clearly illustrate the improvements in spatial resolution that might be expected when using the DCCRM method. Figure 5(b) suggest that CRM would only be able to resolve the white line, and the orange stripes are not clearly discernable in either the image or the cross-section profile; in contrast the orange stripes can be seen in both the image and the cross-section profile of the DCCRM IPSF simulation, Fig. 5(c). Therefore, these simulations clearly support the hypothesis that DCCRM has higher spatial resolution than CRM.
2.3 Differential correlation-confocal Raman microscopy system
As stated, in our DCCRM, a diode-pumped solid-state laser (532 nm, Samba 150, Cobolt, Sweden) was used as the excitation source. A 100× objective (100×, NA = 0.95, MPlanApo N, Olympus, Japan) and a high-precision piezo-actuated objective scanner (P-754.1CD, Physik Instrumente, Germany) were used for axial scanning. An x-y piezo-electrical driving stage (P-542.2CD, Physik Instrumente, Germany) was used for lateral scanning. Two 30 µm pinholes (PH-30, Newport, USA) and two photomultiplier tubes (PMT, H10723-01, Hamamatsu, Japan) were used in the differential confocal system. A Notch Filter (LPD01-532RU-25, Semrock, USA) was employed to separate the reflected and Raman scattered light. The Raman scattered light passed through a 100-µm pinhole (PH-100, Newport, USA) before being detected by a spectrometer (Omni-λ 500, Zolix, China) equipped with an EMCCD (DU970P-BVF, Andor, UK).
3. Results and discussion
3.1 Axial focusing capability performance
To verify the axial focusing capability of DCCRM, a silver-plated mirror was used as the experimental sample. The objective was driven by the piezo-actuated objective scanner to move at equal intervals along the axial direction. The minimum distinguishable interval was used as the axial focusing capability of DCCRM. Figure 6 shows the measured axial response curve in DCCRM when the nanometer-precision objective scanning system moves in a series of 1 nm axial steps. As the reflected light from the silver-plated mirror becomes defocussed, a clear step profile was formed. Therefore, the axial focusing capability of the DCCRM system can reach a precision of 1 nm, offering an accurate focus tracking capability during long-time scanning imaging.
3.2 Axial resolution of Raman performance
To evaluate the axial resolution of DCCRM, bilayer graphene was used as the sample. Again, the objective was driven by a piezo-actuated objective scanner to move at a series of equal intervals along the axial direction. The FWHM of the spectral axial intensity response curve with the Raman confocal pinhole in a fixed position was used to measure a CRM system axial resolution [28]. Here, the spectral axial response curve was drawn using the G-band intensity at 1590 cm−1 [29]. For comparison, Fig. 7 also shows the axial intensity response for the DCCRM measurement. The FWHMs of these curves are 1.54 µm and 1.03 µm for the CRM and DCCRM measurements respectively. Therefore, DCCRM improves the axial resolution by 33.1% compared with CRM.
Fig. 7. Spectrum axial intensity responses of Graphene. The z scanning range is 9 µm, with a scanning step of 30 nm and exposure time 2s.
3.3 Lateral resolution of Raman performance
To evaluate the lateral resolution of DCCRM, a single multiwalled carbon nanotube (SMWCNT) was used as the sample for Raman mapping. The FWHM of the spectrum map cross-sectional line profiles was used as the lateral resolution of the system [30]. Again, Raman mapping employed the G-band intensity at 1600 cm−1. As shown in Fig. 8, the FWHMs of the spectrum cross-sectional line profiles are 332.7 nm and 255.8 nm for CRM and DCCRM, respectively. Therefore, DCCRM improves the lateral resolution by 23.1% compared with CRM, and the SNR of the image is also greatly improved.
Fig. 8. Mapping comparisons between CRM and DCCRM of single carbon nanotube. The x, y scanning range is 3.2 × 1.6 µm2, with a scanning step of 25 nm and exposure time 2s. (a) Mapping of CRM; (b) Mapping of DCCRM; (c) Cross-sectional line profiles.
3.4 High spatial resolution topographic imaging and Raman mapping performance
To further illustrate the improvements and utility of high spatial resolution topographic and Raman mapping offered by DCCRM, a silicon carbide crystal was used. Silicon carbide (SiC) has good physical and chemical properties, such as high mechanical robustness, high chemical inertness, and high breakdown field. Thus, it has a wide application prospect in aerospace equipment, high voltage switch applications, high-temperature electronic equipment, and other extreme environments [31,32]. However, traditional material processing methods have poor reproducibility due to the high stability of SiC, leading to difficulties in manufacturing of SiC based micro and nano scale structures and devices. Nevertheless, femtosecond laser micromachining is potentially a highly suitable method for the processing of SiC microdevices due to its transient ultrahigh-energy feature. Consequently, measuring the changes in the geometrical morphology and chemical composition of SiC after femtosecond irradiation will provide a robust means to optimize and promote the manufacturing method [33,34].
As suggested above, DCCRM should be able to achieve high precision in both in situ topographic imaging and Raman mapping. To verify this capability, a 4H–SiC (0001) crystal irradiated by a femtosecond laser (model: FemtoYL 20, wavelength: 1033 nm, energy: 40 mW) was used as the experimental sample.
The microscopic image of the 4H–SiC measurement area with a size of 30 × 30 µm2 is shown in Fig. 9. It can be seen that the simple microscopic image does not contain height or material distribution information. To address this, initially we obtained single point spectrum of points A, B and C in Fig. 9, where point A represents the femtosecond laser irradiation center, point B represents the edge of the irradiation area, and point C represents the edge of the nonirradiated area.
The Raman spectra of the three points are shown in Fig. 10(a). It can be seen that the characteristic peaks are different at different positions. The Raman peaks of 778 cm−1 and 967 cm−1 appear at all three points. The Raman peak of 520 cm−1 from the first-order Raman peak of crystalline silicon only appears at point B, and the Raman peak of 876 cm−1 representing amorphous SiC, specifically appears at point A. The results show that part of 4H–SiC transforms into amorphous SiC due to femtosecond laser irradiation. Since amorphous SiC is a metastable state, a part of the molecular bonds between Si–C is broken, resulting in the formation of monocrystalline Si.
Fig. 10. Topographic imaging and Raman mapping of 4H–SiC with femtosecond irradiated. (a) Single Raman spectrum at point A, B, C (440–1000 cm−1); (b) The topographic image; (c) Raman intensity mapping of 779 cm−1; (d) Fusion image of topographic image and Raman intensity mapping of 779 cm−1; (e) Raman intensity mapping of 520 cm−1; (f) Fusion image of topographic image and Raman intensity mapping of 520 cm−1.
A large area scan of the sample was used to further study the topography and chemical composition distribution of the 4H–SiC irradiated area. The scanning interval was 0.2 µm, the scanning points were 150 × 150, the image size was 30 × 30 µm2, the exposure time for each point was 0.5 s, and the spectral range was 440–1000 cm−1.
Figure 10(b) shows the topographic imaging of the 4H–SiC crystal irradiated by femtosecond laser. It is clear that the crystal irradiated shows obvious ridges with a height of 2.1 µm. The 4H–SiC (778 cm−1) and Si (520 cm−1) Raman peaks were used for mapping, and the results are shown in Fig. 10(c) and (e), respectively. As DCCRM uses the same excitation spot for reflected light and Raman scattered light, the topographic image precisely matches the Raman mapping point by point. Figure 10(b) was merged with Fig. 10(c) and (e), respectively, to get two composite images shown in Fig. 10(d) and (f). Figures 10(d) and (f) contain topographic information (in X, Y and Z), and material concentration (in colors). According to Fig. 10(d) and (f), the intensity of Raman peak at 778 cm−1 (4H–SiC) decreased significantly in the irradiated (uplifted) region of the sample after femtosecond laser irradiation, and crystalline silicon (520 cm−1) could be clearly seen at the edge of this region.
Thus the mechanism and results of femtosecond laser irradiation on 4H–SiC can be revealed by studying the Raman spectrum and intensity distribution using the DCCRM method and so guide femtosecond laser micromachining processes by correlating with performance analyses.
4. Conclusion
In conclusion, we proposed a high spatial resolution DCCRM method that achieves in situ topographic imaging and Raman mapping with high spatial resolution under homologous excitation. The axial focusing capability of DCCRM system was approximately 1 nm, which provided high spatial resolution topographic imaging and accurate focus tracking capability. DCCRM improved lateral and axial resolutions by 23.1% and 33.1%, respectively, compared to CRM, and the SNR of the image was improved. In addition, the method achieved a high spatial resolution fusion image of irradiated SiC by combining topographic imaging and Raman mapping, accurate and rich in information. DCCRM thus provides a new approach for the further applications of CRM in laser irradiation, materials science, and other fields. However, the two spectra collection makes DCCRM not suitable for application in the field of real-time spectroscopic imaging, e.g., spectroscopic imaging of living tissue/cells and alike. In the future, we will focus on improving detecting speed.
Funding
National Key Research and Development Program of China (2018YFF01012001); National Science Fund for Distinguished Young Scholars (51825501); The Civil Aerospace Technology Advance Research Project (D030207).
Disclosures
The authors declare no conflicts of 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.
References
1. X. Zhou, X. Hu, B. Jin, J. Yu, K. Liu, H. Li, and T. Zhai, “Highly Anisotropic GeSe Nanosheets for Phototransistors with Ultrahigh Photoresponsivity,” Adv. Sci. 5(8), 1800478 (2018). [CrossRef]
2. P. Kun, G. Kukucska, G. Dobrik, J. Koltai, J. Kurti, L. P. Biro, L. Tapaszto, and P. Nemes-Incze, “Large intravalley scattering due to pseudo-magnetic fields in crumpled graphene,” npj 2D Mater. Appl. 3(1), 7 (2019). [CrossRef]
3. C. Zhang, Q. Chen, W. Liao, R. Dai, L. Zhang, X. Jiang, J. Chen, Z. Zhang, and X. Jiang, “Analysis of residual stress fields from fictive temperature distributions within heat-affected zones of fused silica,” Opt. Express 29(26), 42511–42522 (2021). [CrossRef]
4. S. Pal, A. Ray, C. Andreou, Y. Zhou, T. Rakshit, M. Wlodarczyk, M. Maeda, R. Toledo-Crow, N. Berisha, J. Yang, H.-T. Hsu, A. Oseledchyk, J. Mondal, S. Zou, and M. F. Kircher, “DNA-enabled rational design of fluorescence-Raman bimodal nanoprobes for cancer imaging and therapy,” Nat. Commun. 10(1), 1926 (2019). [CrossRef]
5. N. Iturrioz-Rodriguez, R. Martin-Rodriguez, C. Renero-Lecuna, F. Aguado, L. Gonzalez-Legarreta, J. Gonzalez, M. L. Fanarraga, and A. C. Perdigon, “Free-labeled nanoclay intracellular uptake tracking by confocal Raman imaging,” Appl. Surf. Sci. 537, 147870 (2021). [CrossRef]
6. N. Altangerel, G. O. Ariunbold, C. Gorman, M. H. Alkahtani, E. J. Borrego, D. Bohlmeyer, P. Hemmer, M. V. Kolomiets, J. S. Yuan, and M. O. Scully, “In vivo diagnostics of early abiotic plant stress response via Raman spectroscopy,” Proc. Natl. Acad. Sci. U.S.A. 114(13), 3393–3396 (2017). [CrossRef]
7. S. Jiang, Y. Zhang, R. Zhang, C. Hu, M. Liao, Y. Luo, J. Yang, Z. Dong, and J. G. Hou, “Distinguishing adjacent molecules on a surface using plasmon-enhanced Raman scattering,” Nat. Nanotechnol. 10(10), 865–869 (2015). [CrossRef]
8. D. R. Klein, D. MacNeill, R. A. Ribeiro, Q. Song, D. T. Larson, P. C. Canfield, E. Kaxiras, R. Comin, and P. Jarillo-Herrero, “Enhancement of interlayer exchange in an ultrathin two-dimensional magnet,” Nat. Phys. 15(12), 1255–1260 (2019). [CrossRef]
9. K. C. Polavaram and N. Garg, “High-fidelity and high-resolution phase mapping of granites via confocal Raman imaging,” Sci. Rep. 11(1), 8022 (2021). [CrossRef]
10. M. S. Mu’min, T. Bohm, R. Moroni, R. Zengerle, S. Thiele, S. Vierrath, and M. Breitwieser, “Local hydration in ionomer composite membranes determined with confocal Raman microscopy,” J. Membr. Sci. 585, 126–135 (2019). [CrossRef]
11. N. Jiang, E. T. Foley, J. M. Klingsporn, M. D. Sonntag, N. A. Valley, J. A. Dieringer, T. Seideman, G. C. Schatz, M. C. Hersam, and R. P. Van Duyne, “Observation of multiple vibrational modes in ultrahigh vacuum tip-enhanced Raman spectroscopy combined with molecular-resolution scanning tunneling microscopy,” Nano Lett. 12(10), 5061–5067 (2012). [CrossRef]
12. A. P. Olson, C. T. Ertsgaard, S. N. Elliott, and N. C. Lindquist, “Super-Resolution Chemical Imaging with Plasmonic Substrates,” ACS Photonics 3(3), 329–336 (2016). [CrossRef]
13. P. Liu, X. Chen, H. Ye, and L. Jensen, “Resolving Molecular Structures with High-Resolution Tip-Enhanced Raman Scattering Images,” ACS Nano 13(8), 9342–9351 (2019). [CrossRef]
14. C. Chen, N. Hayazawa, and S. Kawata, “A 1.7 nm resolution chemical analysis of carbon nanotubes by tip-enhanced Raman imaging in the ambient,” Nat. Commun. 5(1), 3312 (2014). [CrossRef]
15. S. Rieger, M. Fischedick, K. J. Boller, and C. Fallnich, “Suppression of resonance Raman scattering via ground state depletion towards sub-diffraction-limited label-free microscopy,” Opt. Express 24(18), 20745–20754 (2016). [CrossRef]
16. S. Rieger, T. Wurthwein, K. Sparenberg, K.-J. Boller, and C. Fallnich, “Density matrix study of ground state depletion towards sub-diffraction-limited spontaneous Raman scattering spectroscopy,” J. Chem. Phys. 148(20), 204110 (2018). [CrossRef]
17. H. K. Chen, X. J. Wu, Y. Q. Zhang, Y. Yang, C. J. Min, S. W. Zhu, X. C. Yuan, Q. L. Bao, and J. Bu, “Wide-field in situ multiplexed Raman imaging with superresolution,” Photonics Res. 6(6), 530–534 (2018). [CrossRef]
18. K. Watanabe, A. F. Palonpon, N. I. Smith, L. D. Chiu, A. Kasai, H. Hashimoto, S. Kawata, and K. Fujita, “Structured line illumination Raman microscopy,” Nat. Commun. 6(1), 10095 (2015). [CrossRef]
19. S. C. Li, L. R. Qiu, Y. Wang, H. Cui, and W. Q. Zhao, “Super-resolution radially polarized pupil-filtering confocal Raman spectroscopy technology,” Meas. Sci. Technol. 31(3), 035903 (2020). [CrossRef]
20. J. Yan, F. Shi, M. Zhao, Z. Wang, Y. Yang, and S. Y. Chen, “Confocal Raman Sensing Based on a Support Vector Machine for Detecting Lung Adenocarcinoma Cells,” IEEE Sens. J. 19(21), 9624–9633 (2019). [CrossRef]
21. L. Ahlinder, S. Wiklund Lindstrom, C. Lejon, P. Geladi, and L. Osterlund, “Noise Removal with Maintained Spatial Resolution in Raman Images of Cells Exposed to Submicron Polystyrene Particles,” Nanomaterials 6(5), 83 (2016).6(5):83. [CrossRef]
22. B. Manifold, E. Thomas, A. T. Francis, A. H. Hill, and D. Fu, “Denoising of stimulated Raman scattering microscopy images via deep learning,” Biomed. Opt. Express 10(8), 3860–3874 (2019). [CrossRef]
23. D. J. Winterauer, D. Funes-Hernando, J. L. Duvail, S. Moussaoui, T. Batten, and B. Humbert, “Sub-Micron Spatial Resolution in Far-Field Raman Imaging Using Positivity-Constrained Super-Resolution,” Appl. Spectrosc. 73(8), 902–909 (2019). [CrossRef]
24. L. Duponchel, P. Milanfar, C. Ruckebusch, and J. P. Huvenne, “Super-resolution and Raman chemical imaging: from multiple low resolution images to a high resolution image,” Anal. Chim. Acta 607(2), 168–175 (2008). [CrossRef]
25. K. Matsuzaki, R. Shimada, and H. O. Hamaguchi, “Superresolution vibrational imaging by simultaneous detection of Raman and hyper-Raman scattering,” Opt. Lett. 36(13), 2545–2547 (2011). [CrossRef]
26. W. Zhao, J. Tan, and L. Qiu, “Bipolar absolute differential confocal approach to higher spatial resolution,” Opt. Express 12(21), 5013–5021 (2004). [CrossRef]
27. H. Cui, Y. Wang, L. R. Qiu, S. C. Li, J. M. Cooper, and W. Q. Zhao, “Synchronous nanoscale topographic and chemical mapping by differential-confocal controlled Raman microscopy,” Photonics Res. 8(9), 1441–1447 (2020). [CrossRef]
28. C. Korzeniewski, J. P. Kitt, S. Bukola, S. E. Creager, S. D. Minteer, and J. M. Harris, “Single Layer Graphene for Estimation of Axial Spatial Resolution in Confocal Raman Microscopy Depth Profiling,” Anal. Chem. 91(1), 1049–1055 (2019). [CrossRef]
29. D. L. Duong, G. H. Han, S. M. Lee, F. Gunes, E. S. Kim, S. T. Kim, H. Kim, Q. H. Ta, K. P. So, S. J. Yoon, S. J. Chae, Y. W. Jo, M. H. Park, S. H. Chae, S. C. Lim, J. Y. Choi, and Y. H. Lee, “Probing graphene grain boundaries with optical microscopy,” Nature 490(7419), 235–239 (2012). [CrossRef]
30. T. Dieing, “Resolution and Performance of 3D Confocal Raman Imaging Systems,” in Confocal Raman Microscopy, T. D. Jan Toporski and O. Hollricher, eds. (Springer, 2018), pp. 121–153.
31. C. Babin, R. Stöhr, N. Morioka, T. Linkewitz, D. L. Timo Steidl, E. Hesselmeier, V. Vorobyov, A. Denisenko, M. Hentschel, C. Gobert, P. Berwian, G. V. Astakhov, W. Knolle, R. Wörnle, S. Majety, P. Saha, M. Radulaski, N. T. Son, J. Ul-Hassan, F. Kaise, and J. Wrachtrup, “Fabrication and nanophotonic waveguide integration of silicon carbide colour centres with preserved spin-optical coherence,” Nat. Mater. 21(1), 67–73 (2022). [CrossRef]
32. B.-S. Song, T. Asano, S. Jeon, H. Kim, C. Chen, D. D. Kang, and S. Noda, “Ultrahigh-Q photonic crystal nanocavities based on 4 H silicon carbide,” Optica 6(8), 991–995 (2019). [CrossRef]
33. C. Kerse, H. Kalaycıoğlu, P. Elahi, B. Çetin, D. K. Kesim, Ö Akçaalan, S. Yavaş, M. D. Aşık, B. Öktem, H. Hoogland, R. Holzwarth, and FÖ Ilday, “Ablation-cooled material removal with ultrafast bursts of pulses,” Nature 537(7618), 84–88 (2016). [CrossRef]
34. Y. Huang, F. Tang, Z. Guo, and X. Wang, “Accelerated ICP etching of 6H-SiC by femtosecond laser modification,” Appl. Surf. Sci. 488, 853–864 (2019). [CrossRef]
