Nanomechanical infrared spectroscopy system and method using gated peak force IR

Abstract

Apparatus and method for characterizing samples with an AFM and pulsed IR laser directed at the tip of the AFM probe. Gated laser pulsing and gated detection based on lock-in amplifiers, box-wheel integrators, or FFT may be used with peak-force tapping operation. Nano-spectroscopic measurements with resolutions of 20 nm or even 10 nm or less can be performed, along with nano-mechanical and other property measurements.

Description

[Technical Field]

[0001] Preferred embodiments relate to performing nano-mechanical and spectroscopy measurements, and more particularly to using peak force tapping (PFT) and gated multi-pulse excitation and detection in vibrational modes of AFM operation to obtain nanoscale localized spectroscopy and / or imaging of samples along with nano-mechanical and other properties. [Background technology]

[0002] Infrared spectroscopy and scanning probe microscopy (SPM) have been combined to perform spectroscopy by integrating an infrared light source, such as a tunable free electron laser, optical parametric oscillator, or quantum cascade laser, with an atomic force microscope (AFM) that has a sharp probe to measure the local absorption of infrared light by a sample. Previous techniques in this area are often based on contact-mode AFM, which extracts the absorption signal from contact resonant vibrations that occur when the sample expands (or contracts) while absorbing light. Recently, tapping-mode-based AFM techniques using IR illumination have been shown to achieve spatial resolutions down to 10 nm. Recently, PeakForce IR, an IR spectroscopy and imaging mode based on the Peak Force tapping AFM operational mode, has been developed with a resolution of 10 nm.

[0003] Generally, information about a sample can be obtained by monitoring the interaction between the sample under test and electromagnetic energy. In spectroscopy, the transmission of light through or reflection from a sample produces a sample characteristic plot of transmitted or reflected intensity as a function of wavelength. This spectroscopic information allows users to determine the physical properties of the sample, such as chemical composition or temperature.

[0004] In particular, spectroscopic measurements with nanoscale spatial resolution continue to improve, but despite the development of imaging techniques with spatial resolution beyond the diffraction limit, spectroscopy that simultaneously achieves chemical specificity and molecular-level sensitivity remains elusive.

[0005] SPM has facilitated improvements in this field. AFM is a typical instrument that includes a probe having a tip, and interacts with the surface of a sample with a force suitable for characterizing the surface of the sample down to the atomic level. Generally, the probe is guided to the surface of the sample to detect changes in the sample's properties. By providing relative scanning movement between the tip and the sample, surface property data can be obtained for specific regions of the sample, and a corresponding map of the sample can be generated.

[0006] A typical AFM system is shown schematically in FIG. 1. The AFM 10 uses a probe assembly 12 including a probe 14 with a cantilever 15. A scanner 24 generates relative motion between the probe 14 and a sample 22 while the probe-sample interaction is measured. In this manner, images or other measurements of the sample can be obtained. The scanner 24 typically consists of one or more actuators that generate motion in three orthogonal directions (X, Y, and Z). Often, the scanner 24 is a single integrated unit that includes one or more actuators, e.g., piezo-tube actuators, to move the sample or probe in all three axes. Alternatively, the scanner may be an assembly of multiple separate actuators. Some AFMs separate the scanner into multiple components, e.g., an XY scanner that moves the sample and a separate Z-actuator that moves the probe. Thus, the apparatus can generate relative motion between the probe and the sample while measuring the topography or some other surface characteristic of the sample, as described, for example, in U.S. Pat. No. 34,489 to Hansma et al.; U.S. Pat. No. 5,266,801 to Elings et al.; and U.S. Pat. No. 5,412,980 to Elings et al.

[0007] In a typical configuration, the probe 14 is often coupled to a vibration actuator or driver 16 that is used to drive the probe 14 at or near the resonant frequency of the cantilever 15. Another configuration measures the deflection, torsion, or other motion of the cantilever 15. The probe 14 is often a micromachined cantilever with an integrated tip 17.

[0008] Generally, an electronic signal from an AC signal source or driver 18 is applied under the control of an SPM controller 20 to drive the probe 14 to oscillate via an actuator 16 (and / or to oscillate the sample, e.g., via a scanner 24). The probe-sample interaction is typically controlled via feedback by the controller 20. In particular, the actuator 16 may be coupled to the scanner 24 and the probe 14, or may be formed integrally with the cantilever 15 of the probe 14 as part of a self-actuated cantilever / probe.

[0009] In many cases, a selected probe 14 is oscillated in contact with a sample 22, with sample characteristics being monitored by detecting changes in one or more characteristics of the probe 14 oscillation, as described above. In this regard, a deflection detection apparatus 25 is typically utilized to direct a beam toward the backside of the probe 14, where it is then reflected toward a detector 26. As the beam moves across the detector 26, an appropriate signal is processed in block 28 to determine, for example, the RMS deflection, which is transmitted to a controller 20, which processes the signal to determine changes in the probe 14 oscillation. Generally, the controller 20 generates control signals to maintain a relatively constant interaction between the tip and sample (or the deflection of the lever 15) to maintain a setpoint characteristic of the probe 14 oscillation. More specifically, the controller 20 can include a high-voltage amplifier 34 and a PI Gain Control block 32 that compares a signal corresponding to the probe deflection due to the tip-sample interaction with a setpoint to a circuit 30, conditioning the resulting error signal. For example, the controller 20 may set the vibration amplitude to a setpoint value A to ensure a constant force between the tip and the sample. S Alternatively, the phase or frequency of the setpoint may be used.

[0010] A workstation 40 may be provided within the controller 20 and / or within a separate controller or connected system or stand-alone controller that receives collected data from the controller and manipulates the data acquired during the scan to perform point selection, curve fitting and distance determination operations.

[0011] AFMs can be designed to operate in a variety of modes, including contact and vibration modes. Motion is achieved by moving the sample or probe assembly up and down perpendicular to the sample's surface in response to deflections of the probe assembly's cantilever as it is scanned across the surface. Scanning occurs in the "xy" plane, which is at least generally parallel to the sample's surface, and vertical motion occurs in the "z" direction, which is perpendicular to the xy plane. Note that the term "generally parallel" is used because many samples have roughness, curvature, and tilt that are out of plane. Data regarding this vertical motion can thus be stored and used to construct an image of the sample surface corresponding to measured sample characteristics, such as surface topography. Known as a TappingMode® AFM (TappingMode® is a trademark of the applicant), this is any mode of AFM operation in which the tip oscillates at or near the resonant frequency of the probe's cantilever. A feedback loop attempts to keep the amplitude of this oscillation constant to minimize the "tracking force," i.e., the force resulting from the tip / sample interaction. Another feedback device keeps the phase or oscillation frequency constant. As with contact mode, such feedback signals are collected, stored, and used with data to characterize the sample. Note that "SPM" and acronyms for specific types of SPM are used to refer to microscope equipment or related technology, such as "atomic force microscope." In a recent improvement on the ubiquitous Tapping Mode (TappingMode®), called Peak Force Tapping (PFT) Mode, discussed in U.S. Patents 8,739,309, 9,322,842, and 9,588,136, which are incorporated by reference, feedback is based on the force measured with each oscillation cycle (which is also informed by the transient probe-sample interaction force).

[0012] Regardless of its mode of operation, AFMs can achieve atomic-level resolution on a variety of insulating or conducting surfaces in air, liquid, or vacuum by using a piezoelectric scanner, a light lever deflection detector, and a miniature cantilever fabricated using photolithography techniques. Because of their resolution and versatility, AFMs have become important measurement tools in a variety of fields, from semiconductor manufacturing to biological research.

[0013] Infrared (IR) spectroscopy is a useful tool in many analytical fields, such as polymer science and biology. However, conventional infrared spectroscopy and microscopy have resolutions on the scale of a few microns, limited by optical diffraction. It has proven particularly useful to perform infrared spectroscopy on a very localized scale, at or below the size of a biological organelle, at various points on the sample surface. In this way, information about the sample composition, such as the location of different materials or molecular structures, can also be obtained.

[0014] Conventional far-infrared (IR) spectroscopy is a widely used technique for measuring the properties of materials. In many cases, the unique characteristics of the infrared spectrum can be used to identify unknown materials. While IR spectroscopy is performed on large samples, it can provide compositional information, but it cannot provide nanoscale structural information. As mentioned above, IR spectroscopy can collect IR spectra with a limited resolution of a few microns. Far-field localization techniques can achieve spatial resolution down to approximately 20 nm through point-spread function reconstruction, but chemically specific information is generally limited and rely on fluorescence from discrete molecules or quantum dot emitters.

[0015] Scattering scanning near-field optical microscopy (s-SNOM) has found some application in infrared spectroscopy and imaging. In scattering SNOM (s-SNOM), external illumination of a sharp (metal or semiconductor) AFM probe tip generates detectable light scattering in the near-field probe-sample interaction region (light scattering can vary depending on the material underneath the tip). Alternative access methods, such as Coherent Anti-Stokes Raman Spectroscopy (CARS) or Tip-Enhanced Raman Scattering (TERS), are also chemically sensitive but are also based on detecting scattered light from the sample.

[0016] Despite progress in the development of imaging techniques with spatial resolution beyond the diffraction limit, spectroscopy that offers chemical specificity and sensitivity at the molecular level remains elusive. The following is a discussion of techniques that rely on mechanical detection of IR absorption, as opposed to optical detection as embodied in s-SNOM, CARS, or TERS.

[0017] One technique based on the use of AFM to generate such local spectra is described in a publication entitled "Local Infrared Microspectroscopy with Sub-wavelength Spatial Resolution with an Atomic Force Microscope Tip Used as a Photo-thermal Sensor (PTIR)" (Optical Letters) (Vol. 30, No. 18, September 5, 2005). This technique is also discussed in U.S. Pat. No. 8,402,819 (hereinafter, the "'819 patent"). While those skilled in the art will understand the details of the disclosed technique, this technique will be briefly described here for clarity. A recent review of AFM-IR techniques based on photothermal detection is given by Mathurin et al. in Journal of Applied Physics, 2022, 131, 010901.

[0018] Referring to the '819 patent, in PTIR, infrared radiation is incident on a specific region of a sample. At wavelengths absorbed by the sample, absorption typically causes a localized increase in temperature and rapid thermal expansion in the sample. A probe is positioned to interact with the sample and a transducer to generate a signal related to the absorbed infrared energy in the region under the probe tip. "Interacting" refers to positioning the probe tip sufficiently close to the sample so that a probe response is detected in response to the absorption of infrared radiation. For example, interaction may be in contact mode, tapping mode, or non-contact mode. One or more probe responses to the absorbed radiation can be read using an associated detector. The induced probe response may be probe deflection, probe resonant vibration (bending, twisting, lateral, etc.), and / or probe thermal response (e.g., temperature change). In the case of probe deflection and / or probe resonant vibration, a suitable detector may include a light beam reflector with amplification and signal conditioning electronics for a split-segment photodiode. For thermal responses, suitable detectors may include, for example, a Wheatstone bridge, current and / or voltage amplifiers, and / or other associated electronics for detecting, amplifying, and conditioning the thermal signal from the probe. The probe response is measured as a function of the wavelength of the incident radiation to produce an absorption spectrum. From the spectrum, materials within the sample can be characterized and / or identified.

[0019] As described in the '819 patent, the AFM device was used with a bottom-up illumination scheme, where the sample was mounted on a ZnSe prism and light was transmitted from below. A pulsed IR source (in this case, a free-electron laser (FEL) beam) was directed toward the prism and struck the sample at an angle that caused total internal reflection, so that the beam propagated within the sample and was extinguished in the air. Thus, only the sample was exposed to the laser beam, while the AFM probe was minimally exposed to the beam. The probe was positioned at a point on the sample by a scanner and held at an average height by feedback electronics. Vertical and lateral deflection signals, as well as feedback signals, could all be monitored.

[0020] When the FEL is pulsed, the sample can absorb some of the energy, resulting in rapid thermal expansion of the sample, as shown in Figure 3 of the '819 patent. If the cantilever's ability to respond to the shock is slower than the shock, this has the effect of quickly shocking the cantilever arm, causing resonant oscillations in the cantilever. Since the absorbed energy is ideally conserved within the sample, only minimal infrared energy is absorbed by the cantilever itself; this shock results in rapid sample expansion. When the probe is brought into contact with the surface by the feedback electronics, the resonance signal moves too fast for the feedback electronics, but is directly observable by the photodetector. Thus, the cantilever moves while still in contact with the surface, creating an effect known as "contact resonance." The absolute deflection, amplitude, and frequency characteristics of the contact resonance vary with the amount of absorption, as do other characteristics, such as ringdown and / or local intensity in localized regions around the probe tip, by analyzing the Fourier transform (FFT) of the ringdown event. Also, depending on the direction of expansion, vertical, paralateral and / or torsional resonances can be excited.

[0021] Resonance-enhanced PTIR (PTIR) is a recent method that offers improved signal levels and spatial resolution, as described in U.S. Patent No. 8,869,602 and the publication "Tip-enhanced Infrared Nanospectroscopy through Detection of Molecular Expansion Forces" by Lu et al. (Nature Photonics 8, 307, 2014). It claims that sensitivity and spatial resolution can be improved by using both field enhancement at the AFM tip (also present in s-SNOM and TERS) and resonant excitation of the cantilever mode (e.g., bending mode or contact resonance mode). The latter can be achieved via IR laser pulsing at the same frequency as the secondary cantilever bending mode while the AFM is operated in contact mode and photoexpansion is detected. Spatial resolution was 25 nm, but film thickness was only about 2 nm. Furthermore, thin films were deposited on Au substrates, significantly enhancing the electric field in the substrate-tip cavity occupied by the sample. The need for substrate enhancement limits this method's applicability to thin films that can be deposited on such substrates. Furthermore, AFM contact modes have significant drawbacks compared to intermittent contact (e.g., tapping) and peak force tapping modes in the form of tip / sample contamination, tip or sample wear, and reduced performance for soft or sticky samples. Tip contamination and tip wear are particularly significant drawbacks because changes in tip shape or surface affect field distribution and field enhancement. Further improvements have been made by using bench-top QCLs instead of large user facility lasers, such as FELs. Other lasers, such as optical parametric oscillators operating in the 2-10 micron wavelength range, have also become available to complement the 5-12 micron range of typical QCLs.

[0022] Another recently developed technique is photo-induced force microscopy (PiFM), described in U.S. Patent No. 8,739,311, which is essentially identical to tapping-mode IR. Here, AFM operates in tapping mode at one mechanical resonance of the cantilever (typically in the 500–1500 kHz range), while PiFM detection is performed in a different cantilever mode by tuning an IR laser to the difference frequency. Like PTIR, PiFM detects mechanical motion, but while PTIR detects optical expansion, PiFM is claimed to be driven by the optically induced dipole-dipole force between the tip and sample. High sensitivity and spatial resolution of up to 10 nm have been observed. Currently, as with resonance-enhanced PTIR, the IR source is limited to QCL lasers because the laser frequency must be tuned to the difference frequency of a specific cantilever mode. Furthermore, because the cantilever mode frequency varies depending on the material underneath the tip, a frequency-shift tracking mechanism is required to appropriately adjust the laser frequency. Importantly, PiFM relies on resonant tapping, in contrast to methods based on the Peak Force Tapping mode of the presently preferred embodiment, as described below, where probe vibration occurs at a level much lower (minimum 5x) than the cantilever resonance.

[0023] In another technique known as Peak Force IR and described in U.S. Patent Nos. 8,955,161, 9,207,167, 9,719,916, and 10,520,426, all of which are incorporated herein by reference, Peak Force Tapping® mode AFM is combined with a light sample overhead to excite a localized photothermal response at the tip-sample interface. This improves resolution and minimizes sample preparation. This method identifies modulus changes based on directional steps to provide an IR absorption signature of the sample. Measurement techniques sensitive to modulus changes, such as Peak Force Tapping (PFT) AFM mode or contact resonance mode, can be employed.

[0024] The abbreviation Peak Force IR (PFIR) generally refers to more recent related technology. PFIR, a method based on peak force tapping in AFM-IR, detects sample reactions induced by an IR laser during a PFT cycle and can provide chemical and nanomechanical information about the sample. Its operating principle is described in U.S. Patent No. 10,845,382, or more recently in Wang et al., Chem. Soc. Rev., 2022, 51, 5268-5286, or Mathurin et al., Journal of Applied Physics, 2022, 131, 010901. In other words, if a laser pulse is absorbed by the sample during the tip-sample contact time within a PFT cycle, the cantilever deflection is modified, and this change is detected. Typically, this change is an oscillation or offset in the deflection signal. To remove the gradually changing cantilever deflection background by subtracting cycle 2 from cycle 1, a single laser pulse was used in the first PFT cycle, and subsequent cycles did not use IR laser illumination. In a subsequent realization (Wang et al., Nano Lett. 20, 3986, (2020)), a fitting procedure was used to remove the gradually changing background before the FFT for signal extraction. However, despite the successful demonstration of PFIR microscopy, its widespread adoption is limited by the complexity of customized signal generation and processing routines. Typical PFIR microscopes require hardware or software-level programming for signal acquisition and processing for background fitting, background subtraction, and FFT extraction, which often involves programming using LabVIEW or a comparable platform.

[0025] Continued improvements are required for wider adoption and better performance of IR techniques based on peak force tapping combined with nanoscale chemical discrimination and nano-mechanical characterization. Summary of the Invention [Means for solving the problem]

[0026] By using Peak Force Tapping (PFT) Mode AFM, the preferred embodiment overcomes the shortcomings of the prior art. In the preferred embodiment, complexity is reduced by gate pulsing during the contact time and gate detection using a lock-in amplifier without background subtraction.

[0027] In one preferred embodiment, an apparatus for performing spectroscopy on submicron regions of a sample using an atomic force microscope (AFM) includes a driver that generates an oscillatory drive signal that causes a probe of the AFM to interact with the sample during a plurality of probe-sample interaction cycles to generate a transient probe-sample interaction force, the oscillatory drive signal having a frequency less than the resonant frequency of the probe. The AFM also includes one or more controllers that control the transient probe-sample interaction force, and an adjustable light source that illuminates the tip-sample region with light pulses to induce deformation of the sample. A detector is provided that measures probe deflection due at least in part to the induced sample modification. At least one of a lock-in amplifier and a signal integrator is provided to extract the sample response to the light pulses from the measured probe deflection.

[0028] According to another aspect of this embodiment, at least one of the lock-in amplifier and the signal integrator is a lock-in amplifier, the sample responses are phase sensitive, and the phase-sensitive sample responses are averaged.

[0029] According to another aspect of this embodiment, the at least one controller uses the sample response to generate a spatially resolved map indicative of absorbed infrared radiation, wherein the map is generated for a sample area of ​​at least 100 x 100 pixels within 5 minutes.

[0030] In an additional aspect of this embodiment, the vibratory drive signal frequency is at least five times lower than the lowest resonant frequency of the probe.

[0031] According to a further aspect of this embodiment, the controller adjusts pulse times during a probe-sample interaction cycle to produce a 180° phase change in light-induced probe deflection over at least two cycles, subtracts the probe deflection corresponding to the at least two cycles, and then extracts the sample response from the subtracted probe deflection.

[0032] According to additional aspects of this embodiment, the at least one light pulse and the extracted sample response are gated during the probe-sample contact time. Further, the at least one light pulse and the extracted sample response are gated every cycle of the probe-sample interaction. The controller can also extract at least one of nano-mechanical and nano-electrical properties from the sample response.

[0033] In another aspect of this embodiment, the phase-sensitive lock-in amplifier output is averaged as opposed to amplitude averaging, resulting in lower noise and thereby an increased signal-to-noise ratio.

[0034] In particular, a method for performing spectroscopic measurements in the submicron region of a sample using an atomic force microscope (AFM) is provided. In a first step, the AFM probe interacts with the sample for several probe-sample interaction cycles to generate a transient probe-sample interaction force. In this case, the oscillatory drive signal has a frequency lower than the resonant frequency of the probe. Next, the method includes illuminating the tip-sample region with pulses of an adjustable light source to control the transient probe-sample interaction force and induce sample deformation. This is preferably performed during the tip-sample contact time. Thereafter, the method includes measuring probe deflection due at least in part to the induced sample deformation. Finally, the method includes extracting a sample response, which may or may not be determined by the tip-sample contact time.

[0035] The extracting step may also include providing one or more of a lock-in amplifier and a signal integrator to extract the sample response only during the tip-sample contact time. For example, if the extracting step uses a lock-in amplifier, the sample response is phase sensitive and the method may further include averaging the phase-sensitive sample response.

[0036] According to another aspect of this embodiment, the method further includes generating a spatially resolved map indicative of absorbed infrared radiation using the sample response, the map being generated for a sample area of ​​at least 100 x 100 pixels within 5 minutes.

[0037] In an additional preferred embodiment, a method for performing spectroscopic measurements on submicron regions of a sample using an AFM is again provided. The AFM probe interacts with the sample for several probe-sample interaction cycles, generating a transient probe-sample interaction force. In this case, the oscillatory drive signal has a frequency lower than the resonant frequency of the probe. The method then includes controlling the transient probe-sample interaction force and illuminating the tip-sample region with pulses of an adjustable light source to induce sample deformation, which occurs during the tip-sample contact time. The probe deflection due to the minimally induced sample deformation is then measured. Finally, the sample response to the light pulse is extracted from the measured probe deflection, but this extraction step is gated.

[0038] The extracting step also includes providing one or more of a lock-in amplifier, an FFT block, and a signal integrator to extract the sample response only during the tip-sample contact time, and since the sample response is phase sensitive, the phase-sensitive sample response is averaged.

[0039] The method of this additional preferred embodiment may further include the steps of (i) timing pulse times during a probe-sample interaction cycle to produce a 180° phase change in light-induced probe deflection during at least two cycles, (ii) subtracting the probe deflections corresponding to the two cycles, and (iii) extracting the sample response during the subtraction step. Furthermore, the sample response can be used to generate spatially resolved maps of absorbed infrared radiation. A map of 100x100 pixels or more for the sample area can be produced within 5 minutes, and the resolution of the sample response map may be less than 20 nm.

[0040] In a further preferred embodiment, a method for performing spectroscopic measurements using an AFM is provided. The method includes causing a probe of the AFM to interact with a sample for a plurality of cycles to generate a probe-sample interaction force with an oscillatory drive signal. The method also includes providing a pulsed light source to generate a plurality of pulses having a pulse width and directing the pulses at a sample in which the probe is located to produce an induced sample response. The pulses substantially coincide with a probe-sample contact time. A probe deflection that depends at least in part on the induced sample response is measured. At least one characteristic of the sample response is derived. Additionally, the directing step of the method is gated during the probe-sample contact time, and the measuring step is gated during the probe-sample contact time for all cycles of the causing step.

[0041] Different preferred embodiments may further include using the sample response to create a spatially resolved map showing absorbed infrared radiation. The map for the sample area may be created in less than 5 minutes with a size of at least 100x100 pixels. The resolution of the sample response may be less than 20 nm, or even less than 10 nm. According to other features, at least one of the illuminating and extracting steps is gated during the probe-sample contact time. Also, at least one of the illuminating and extracting steps is gated for every cycle of the causing step.

[0042] These and other features and advantages of the present invention will become apparent to those skilled in the art from the following detailed description and accompanying drawings. It should be understood, however, that the detailed description and specific examples, while indicating preferred embodiments of the invention, are given by way of illustration and not limitation. Many changes and modifications may be made within the scope of the invention without departing from the spirit thereof, and the invention includes all such modifications. [Brief explanation of the drawings]

[0043] Preferred embodiments of the present invention are illustrated in the accompanying drawings, in which like reference numerals refer to like parts throughout.

[0044] [Figure 1] FIG. 1 is a schematic diagram of a prior art atomic force microscope (AFM). [Figure 2] FIG. 2 is a schematic diagram for a gated Peak Force IR (PFIR) setup of the preferred embodiment. [Figure 3] FIG. 3 is a plot of PFT deflection versus time illustrating the laser-driven probe response during gate pulsing according to a preferred embodiment. [Figure 4] 4A-4D in Fig. 4 show a PFT deflection versus time pulsing and readout scheme for laser pulsing synchronized with the PFT deflection cycle. According to a preferred embodiment, no phase change of the pulsing sequence occurs between successive PFT cycles. [Figure 5] 5A-5D in FIG. 5 are plots of PFT deflection versus time showing a 180° phase change between laser pulses in successive PFT cycles according to a preferred embodiment, a measurement indicative of contact vibration that can be eliminated in a preferred embodiment. [Figure 6A] FIG. 6A is a plot illustrating an alternative gate detection scheme according to a preferred embodiment. [Figure 6B] FIG. 6B is a plot illustrating an alternative gate detection scheme according to a preferred embodiment. [Figure 6C]FIG. 6C is a plot illustrating an alternative gate detection scheme according to a preferred embodiment. [Figure 7] 7-10 show the variable effect of continuous pulsing and continuous detection compared to gated pulsing and detection of the preferred embodiment, with FIG. 7 being a graph of laser repetition rate versus IR signal. [Figure 8] Figure 8 is a schematic AFM block diagram of the setup for acquiring intensity maps with a typical alignment step of the PFIR. [Figure 9] Figure 9 depicts the image using the settings in Figure 8. [Figure 10] FIG. 10 is a graph of wavenumber versus normalized lock-in amplitude. [Figure 11A] FIG. 11A shows imaging, spectroscopy and correlation nanomechanical measurements taken using the setup according to a preferred embodiment. [Figure 11B] FIG. 11B shows imaging, spectroscopy and correlation nanomechanical measurements taken using the setup according to a preferred embodiment. [Figure 11C] FIG. 11C shows imaging, spectroscopy and correlation nanomechanical measurements taken using the setup according to a preferred embodiment. [Figure 11D] FIG. 11D shows imaging, spectroscopy and correlation nanomechanical measurements taken using the setup according to a preferred embodiment. [Figure 11E] FIG. 11E shows imaging, spectroscopy and correlation nanomechanical measurements taken using the setup according to a preferred embodiment. [Figure 11F] FIG. 11F shows imaging, spectroscopy and correlation nanomechanical measurements taken using the setup according to a preferred embodiment. [Figure 12A] FIG. 12A shows high-resolution nano-IR absorption imaging data with a resolution of less than 10 nm acquired using the setup according to a preferred embodiment. [Figure 12B] FIG. 12B shows high-resolution nano-IR absorption imaging data with a resolution of less than 10 nm acquired using the setup according to a preferred embodiment. [Figure 12C] FIG. 12C shows high-resolution nano-IR absorption imaging data with a resolution of less than 10 nm acquired using the setup according to a preferred embodiment. [Figure 13] FIG. 13 is a flow chart of a method according to a preferred embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0045] FIG. 2 shows an experimental setup 200 for an embodiment of the present invention. A probe 201, with a cantilever 202 terminated in a sharp tip 203, engages a sample of interest 204. The IR probe 201 is preferably coated with a metal such as Au or PtIr to achieve a lightning rod effect and localize the light under the tip 203. The sample 204 is mounted on an atomic force microscope stage 206 that includes a 3D piezo scanner. A piezo 208 can be attached to the cantilever 202. The AFM is capable of peak force tapping operation (e.g., Bruker's Dimension Icon® or Multimode® AFM). The sample stage 206 and / or piezo 208 provide relative vertical motion between the tip and sample, while stage 206 can impart in-plane XY motion for sample scanning. The vertical deflection of the probe 201 is detected using conventional beam-bounce optical detection with a diode laser 210 and a position sensor 212 (e.g., a 4-quadrant photodetector). The vertical deflection is measured and sent to a controller 214 for AFM feedback, which controls the z-position of the sample 204, for example, using a stage / xyz scanner 206 and / or a piezo 208. Of particular importance is that the atomic force microscope controller 214 is equipped with a Peak Force Tapping mode capability, as described, for example, in U.S. Pat. No. 10,845,382.

[0046] The controller 214 also controls a frequency- and wavelength-tunable light source 216. The light source 216 can provide a wide range of wavelengths, from ultraviolet to far-infrared. In one embodiment, the source 216 provides infrared radiation (IR) that matches the vibrational resonance of molecules in the substance under test, i.e., the sample 204. The laser 216, such as a quantum cascade laser (e.g., MIRcat, manufactured by Daylight Photonics) or an optical parametric oscillator (OPO), provides laser pulses 218 at a frequency dictated by the controller 214. The light beam 222 is focused onto the tip-sample region, i.e., the tip-sample interaction region, via a focusing element 220 (e.g., an off-axis parabolic type with a 25 mm focal length) or other optical focusing element, such as a lens. The resulting spatial scans 224 and wavelength-dependent nanoscale localized spectra 226 showing IR absorption at various wavelengths λ, λ, and λ are processed by the controller 214 or workstation and displayed on a workstation screen or stored with the data. Such IR imaging data can be obtained before, after, or during the collection of other sample attribute data (e.g., mechanical (elastic modulus, adhesion), electrical (surface potential or current in KPFM or TUNA), or other measurements that may be provided with the AFM operating mode of PFT).

[0047] Preferably, the relative position between the focal point of the infrared beam 222 and the tip 203 remains constant during IR data collection, i.e., the optical alignment with respect to the tip remains unchanged during IR absorption mapping across the sample and point spectroscopy at a fixed sample position. This allows quantitative comparison of the surface response to IR light at different locations, since the light intensity in the probe-sample interaction region where surface deformation occurs remains constant during IR scanning of the surface at a single IR wavelength.

[0048] In other embodiments, the IR laser spot may be much larger than the AFM scan area, and the light intensity variation during the scan of the probe relative to the IR illumination spot remains sufficiently constant (e.g., within 10%) during the scan. As a result, IR data at other positions on the probe 201 remain within 10% due to different laser powers. In further embodiments, the effects of relative motion between the probe and the IR illumination area can be compensated for. One method is to track the position of the probe with the IR illumination spot during the scan. Another method is to measure the spatial variation of the IR signal on a sample that exhibits a uniform IR response. If three-dimensional PFIR responses are acquired for various x, y, and z positions relative to the IR illumination spot while the probe 201 is in contact with the sample, measurements on other samples can be corrected for spatial IR light variations.

[0049] The controller 214 includes a frequency generator for pulsing the laser source 216. For example, a QCL allows pulsing according to an applied TTL signal. Alternatively, IR pulses may be selected from the laser output beam 222 via optical means, such as an acousto-optical modulator, an electro-optical modulator, or a Pockels cell. A mechanical pulse picker (chopper) or a rotating mirror can block unwanted pulses and allow only selected pulses to pass through to the chip. Such an element may be inserted at the IR output of the IR light source or may be part of the IR light generation process of the laser system itself. In this case, for example, a Pockels cell acts as a pulse selector to select pump laser pulses in an optical parametric oscillator or amplifier that drives the IR light generation process. Ultimately, what matters is that the laser pulses illuminate the chip 203 at a pulse repetition rate controlled by the controller 214. The IR light beam 222 is linearly polarized along the tip 203, and the field is enhanced at the apex of the tip 203, which is typically conductive or metal coated (e.g., PtIr, Pt, or Au), resulting in test conditions similar to those known for TERS or s-SNOM. A non-conductive tip and polarization of the light perpendicular to the tip 203 results in a reduced signal.

[0050] In PFT, the vertical position of the sample 204 on the stage 206 may be sinusoidally modulated with an appropriate drive signal provided by the controller 214 at a low frequency of several kilohertz, substantially (i.e., at least five times) lower than the cantilever resonance frequency. This allows the probe 201 to be fixed. Alternatively, the probe 201 may be sinusoidally modulated in vertical position using, for example, a piezo 208 (or an equivalent drive device that uses magnetic, electrostatic, thermal, or optical forces to drive the cantilever). In essence, relative vibration between the probe and sample is required. Assuming only sample vibration, around the upper pivot point of the vibration, the sample contacts the tip 203 of the probe 201 with a controlled radius of 1 to 50 nm. The maximum deflection of the cantilever during contact, i.e., the peak force, is used as a feedback setpoint by the controller 214 to maintain the average distance between the sample 204 and the tip 203. For example, mechanical properties such as modulus, dissipation, and adhesion can be extracted from the PFT by analyzing the time-varying trace of the cantilever's vertical deflection recorded by the deflection sensor 212.

[0051] FIG. 3 illustrates tip vertical deflection data 300 as a function of time during a peak-force tapping (PFT) cycle in a preferred embodiment. In this example, vertical deflection is given for the case of an IR laser pulse perturbation (deflection 302) and the case of no perturbation, i.e., no IR laser pulse or no IR pulse during the contact time (deflection 304—dotted curve, vertically offset for clarity). During the PFT cycle, the probe approaches the sample and makes contact with it at snap-in contact time 306, which is approximately 160 microseconds. Free-space oscillations 308 of the cantilever present prior to this moment are subsequently suppressed. Note, however, that snap-in contact can cause cantilever oscillation at one of the cantilever's contact resonances. This is evident in one or two oscillation cycles of deflection 304 during the first ~10 microseconds after snap-in contact point 306. After point 306, the tip 203 and sample 204 continue their relative motion, with the probe 201 and sample 204 approaching each other until the peak force reaches peak force tapping control point 310, i.e., the point at which feedback holds the force constant during the PFT cycle. The location of the feedback point is adjustable in software but preferably corresponds to the maximum value of the PFT deflection curve. The relative motion of the tip and sample then reverses, causing the tip to lift off the surface at adhesion point 312 (here, approximately 240 microseconds). This phenomenon again generates a strong free-space oscillation of the cantilever (245-270 microseconds), which decays until the probe again contacts the sample surface in the subsequent PFT cycle. In the presented example, laser pulse 316 is applied for the duration of the probe-sample interaction cycle, or contact time t c, i.e., only exists during the time interval (160-240 microseconds) during the PFT cycle when the tip contacts the sample, which is the time between the snap-in contact time 306 and the attachment point 312. The relative time markers at which the start and end of the IR pulse train are set can be the peak force tapping control point 310, or less preferably the snap-in contact point 306 or the attachment point 312, as will be described later, and the delay for the laser start is the contact time t along the pulse. c Note that contact times are typically between 1 ms and microseconds depending on the operating frequency of the peak force tapping, and IR laser pulse widths are typically between 1 picosecond and several hundred nanoseconds.

[0052] In this example, common polymer samples, such as PMMA and polystyrene (PS), induce various cantilever responses when the sample absorbs IR laser radiation during laser pulsing. Generally, samples are deformed in several ways by the material; for example, light-induced effects can cause surface motion, charge accumulation / displacement, and / or sample polarization, resulting in the generation of mechanical or electromagnetic surface pulse forces in response to the light. In the case of PS, the infrared surface pulse force arises from the sample expanding as light is absorbed, most significantly resulting in cantilever oscillation 314, which is synchronously averaged here over various PFT cycles for clarity. Such oscillations can arise at the cantilever's contact resonance, in which case the IR signal, i.e., the oscillation amplitude, is amplified by the q-factor of the contact resonance mode, similar to a resonantly driven oscillator. Because the force changes during the PFT cycle, the contact resonance and "Q" are expected to change slightly during the PFT cycle. The described vibrations can also occur off-resonant with all cantilever modes, and vibrations at frequency f can be induced while the laser repetition rate in pulse train 316 is a fraction of f (1 / n, where n = 1, 2, or 3). That is, a low resonant drive excites higher harmonics that coincide with the contact resonance of the probe. The most prominent laser-induced effects are vibrations, which are analyzed using a lock-in amplifier in preferred embodiments or other means such as an FFT block or a boxcar-like signal integrator. Typically, the contact resonance employed is in the 50-3000 kHz range, and the average laser power used to stimulate the probe resonance is typically less than 1 mW.

[0053] Here, a lock-in amplifier is a physical device and / or algorithm that demodulates the response of a system at a reference frequency. A lock-in amplifier may be an electronic assembly including analog electronics, digital electronics, and a combination thereof. It may also be a computational algorithm implemented by digital electronics such as microprocessors, field programmable gate arrays (FPGAs), digital signal processors (DSPs), and personal computers. A lock-in amplifier analyzes a vibrating system and outputs various signals, including amplitude, phase, in-phase (X), quadrature (Y) components, or a combination thereof. Here, a lock-in amplifier can also generate measurements at both the reference frequency and higher harmonics of the reference frequency.

[0054] In a preferred embodiment of the present invention, the controller 214 controls the contact time t c The iteration rate f defined between laser This triggers laser emission in the form of a laser pulse 316, and no emission occurs outside of this probe-sample interaction cycle. The peak force tapping control point 310 serves as a synchronization point in the PFT cycle, relative to which the start and stop of laser pulsing is selected.

[0055] In a preferred embodiment, detection of the laser-induced IR signal relies on gated detection 318 based on a lock-in amplifier, boxcar signal integration, or similar techniques. In lock-in detection, the vertical deflection signal 302 with the laser-induced deflection change 314 is demodulated at a reference frequency given by the laser repetition rate. The controller 214 that determines the laser repetition rate may have a built-in lock-in device to which the reference frequency is provided. Data collection and lock-in demodulation begins at the start of laser pulsing and ends after the last laser pulse, i.e., the collection is gated to prevent actual events of snap-in contact 306 and adhesion point 312, and the signal is demodulated at the contact time t c The contact time varies depending on the sample properties and the PFT operating settings. While the PFT properties remain fairly constant during imaging, the sample under the tip may experience significant changes in mechanical properties (e.g., adhesion strength) that can affect the time. c In this case, the gating length and the position relative to the PFT cycle are kept constant during the scan at various sample positions, and the smallest t c It is beneficial to keep the time less than 1 / s. In this manner, noise and instability due to PFT motion around the snap-in contact time and the attachment point can be avoided from affecting data collection. Using a larger gating window or performing a sustained lock-in over various PFT cycles to collect data beyond the contact time adds noise to the signal of interest because local information about the tip-sample interaction region is not available during the period when the probe separates from the sample.

[0056] To increase the signal-to-noise ratio, acquired data can be averaged within or between PFT cycles. Data averaging refers to summing or integrating data outputs from devices such as a lock-in amplifier, and normalizing the acquisition time to the number of collected analog-to-digital converter (ADC) samples or PFT cycles. In a preferred embodiment, averaging data over two or more PFT cycles involves averaging the phase-sensitive output of the lock-in acquired during a gated detection window. That is, the amplitude and phase outputs acquired in each PFT cycle are averaged as a complex value with the next PFT cycle. This contrasts with averaging only the amplitude output, ignoring the phase information. Phase-sensitive averaging results in lower signal noise because, for example, two complex values ​​(representing noise) with the same amplitude but opposite phase cancel each other out when added together, but the sum of the amplitudes does not. Thus, in this example, noise is suppressed in the first case but not in the second case. If increased noise and elevated baseline / offset are tolerated, in one embodiment phase-insensitive averaging of amplitude is also possible.

[0057] As previously mentioned, laser pulsing and detection are preferably limited to the probe-sample interaction cycle. The beginning and end of the detection and pulsing window may be selected by the user relative to the peak force control point 310 to restrict pulsing 316 within the boundaries given by the snap-in contact point 306 and the attachment point 312. Instead of point 310, the synchronization point may be the snap-in contact point 306 or the attachment point 312. However, it should be noted that the latter two may vary depending on the sample, tapping amplitude, peak force, and other effects, so the peak force tapping control point 310 is more preferred. The contact time t cSince t varies depending on the effects mentioned above, particularly sample position, it is beneficial to dynamically adjust the length of the laser pulsing window relative to the measured points 306 and 312. Alternatively, the pulsing length may be calculated from scan parametrics (e.g., peak force) or may be held constant to always remain within points 306 and 312 during the scan. When the laser pulsing window is t c Note that even larger values ​​than t are acceptable, and continuous pulsing over the entire PFT cycle also works well as discussed. However, limiting pulsing to only the contact time reduces heating of the sample and reduces undesirable effects such as induced free oscillation of the cantilever if the probe is detached from the surface after attachment point 312. Gating detection only during the contact time improves the signal-to-noise ratio, as otherwise only noise and artifacts would be present during the absence of tip-sample contact. c This will enter the outer detection channel.

[0058] In PFIR, a longer tip-sample contact time is desirable to increase the duty cycle and thus improve the signal-to-noise ratio. However, the tip must be able to move away from the surface without touching it. The contact time is controlled by the PFT tapping amplitude, which typically ranges from 30 to 150 nm for PFIR.

[0059] In the example vertical deflection data 300, other less pronounced features may also serve as signals indicative of IR absorption or IR-induced surface pulse forces. In addition to the strong oscillations in the deflection trace mentioned above, here vertical, but generally vertical and / or horizontal, IR laser absorption may alter the mechanical properties of the sample. For example, this may result in the adhesion point 312 being shifted in time (later or earlier) and / or size (i.e., to higher / lower deflection values). The same applies to the slope between the snap-in contact time 306 or the force maximum (which overlaps with the peak force setpoint 310 in the example data 300 of FIG. 3 ) and the adhesion point 312, which indicates a change in elastic modulus. Free-space oscillations 308 may also be modified as a result of IR-induced surface pulse forces, which may occur due to sample expansion or contraction. This may include changes in amplitude, frequency, or phase when compared to IR-unpulsed PFT cycles. Compared to the IR un-pulsed trace, these changes persist for several hundred microseconds after the IR pulse until the next probe-sample interaction stops the free-space cantilever oscillation at the snap-in contact time 306.

[0060] Other samples, especially inorganic samples, exhibit PFIR signal generation processes distinct from mechanical probe-sample forces due to surface expansion and contraction or shock waves. Inorganic materials with low thermal expansion coefficients exhibit electromagnetic probe-sample forces, such as charge accumulation, displacement, or polarization induced by IR light. For example, graphene supports surface plasmons and localized collective charge oscillations in IR, allowing the IR light to efficiently couple with momentum provided by the probe tip 203 (similar to s-SNOM). Surface and bulk plasmons are detected from PFIR with associated electromagnetic probe-sample forces. Similar forces can also be expected for other quasiparticle excitations, such as phonon-polaritons in boron nitride.

[0061] Note that in a preferred embodiment, to maximize the duty cycle and signal-to-noise ratio, no non-pulsed PFT cycles are used, and all cycles are exposed to IR laser pulses. Other sample attributes, such as electrical properties (current or surface potential) or nanomechanical properties (adhesion, elastic modulus) acquired during PFT operation, can be acquired during laser irradiation. Crosstalk or interference between IR absorption measurements and elastic modulus can occur, for example, when the sample softens or melts under IR irradiation. In such cases, it is beneficial to separate the PFT cycles used for IR absorption extraction from those used for other properties; for example, during nanomechanical data collection, the IR laser is not pulsed and no corresponding data is extracted during IR absorption measurements. Such separation can occur within an imaging scan line, alternating between scan lines, between trace and retrace lines, or alternating between PFT cycles. Even within a single PFT cycle, laser pulsing and signal detection can be separated from nanomechanical data extraction; for example, IR absorption can be acquired during the first half of a PFT probe-sample interaction cycle, and elastic modulus data can be acquired during the second half.

[0062] Interference can occur between the laser-induced vibrations and the PFT force feedback mechanism, which is necessary to ensure stable AFM operation. To prevent this, in a preferred embodiment, the force feedback in controlling peak force tapping at point 310 is based on a low-pass filtered deflection signal. For example, a 40 kHz low-pass filter suppresses laser-induced deflection oscillations in the absorbing sample due to laser pulsing from one of the contact resonances in the hundreds of kHz to several MHz range. Meanwhile, the low-pass filter must transmit low-frequency deflection changes so that the feedback maintains a constant peak force setpoint during AFM operation, either while scanning the sample or while stationary (e.g., during infrared absorption spectrum acquisition). Without the low-pass filter, the peak force setpoint feedback would be disrupted, causing increased peak force noise, topography height noise, and general noise for all AFM modes (including infrared absorption) that rely on stable AFM operation. If the interference is small, or if peak force tapping feedback is only applied during non-pulsed PFT cycles, the low-pass filter can be removed. Or alternatively, the time window around the peak force tapping control point 310 can be excluded from laser pulsing so that the AFM feedback is not disturbed.

[0063] Figure 4 illustrates an embodiment based on continuous laser pulsing. In this example, two consecutive PFT cycles are shown. Figure 4A shows the contact time t c is ~80 microseconds, and the total PFT cycle time t PFT The PFT cycle f is ~220 microseconds. PFT The vertical deflection signal for ∼4.5 kHz is shown as a function of time.

[0064] Figure 4B shows a well-defined repetition rate f laser =1 / T laserThe figure shows laser pulses 400 continuously running at a constant speed. The pulses are synchronized with the PFT cycle; that is, each laser pulse occurs at the same relative time point in the PFT cycle. For example, the vertical marker 402 for two consecutively displayed PFT cycles overlaps with the peak force tapping control point 310 for each cycle. That point 310 serves as a defined time marker within the PFT cycle and is synchronized with the relative PFT motion between the cantilever and the sample; that is, in the case of sinusoidal PFT motion, the maximum of the PFT oscillation occurs at a fixed phase or time delay relative to the peak force tapping control point 310. The snap-in contact time 306 or adhesion point 312 are not suitable as time stamps (peak force tapping control points) because they change dynamically with respect to point 310 and the resulting PFT motion. Due to these dynamic changes, points 306 and 312 must be derived through measurements based on sample attributes and AFM motion. Marker 402 indicates that the laser pulse 400 is synchronized with the PFT cycle, and in the given example, there is always a laser pulse that overlaps with the peak force tapping control point. In other words, the laser repetition rate f laser is an even integer multiple of the peak force tapping frequency, i.e., f laser =n*f PFT where n is an even integer.

[0065] In a typical example, the PFT frequency is f PFT = 2 kHz, and the laser repetition rate is f laser= 1300 kHz, 1302 kHz, or 1304 kHz. The laser pulse repetition rate and PFT frequency can be synchronized, for example, by sharing a common clock. As described with reference to Figure 3, the laser pulse generates IR-induced mechanical or electromagnetic surface pulse forces, which can be caused by surface motion (sample expansion or contraction), charge accumulation / displacement, and / or sample polarization. Such forces generate oscillations 404 (otherwise undisturbed tip deflection 406), as highlighted in Figure 4C. Here, only the deflection during the contact time tc, or probe-sample interaction cycle, is shown because this is the time window of interest. For illustrative purposes, the frequency and amplitude of the oscillations are exaggerated. Outside of the contact time, as will be discussed later, this is the contact time during which the laser does not induce significant cantilever oscillations unless the cantilever's free-space resonance is nearby. As previously mentioned, synchronizing the laser pulse repetition rate to the PFT cycle frequency also synchronizes the oscillations of the tip deflection during the contact time to the PFT cycle. That is, as indicated by vertical marker 402, the minimum and maximum values ​​of the laser-induced vibration occur at the same time points within the PFT cycle compared to a reference point such as peak force tapping control point 310.

[0066] FIG. 4D shows a lock-in reference signal 408. This reference signal is synchronized with the laser repetition rate since both originate from the controller 214. In one embodiment, the lock-in amplifier is gated to demodulate the laser-induced signal only during the probe-sample interaction cycle, as indicated by time window 410. This gating of the lock-in operation is triggered relative to a time marker of each PFT cycle, such as the peak force tapping control point 310, as described above. This lock-in-based signal demodulation is then repeated during the PFT cycle, as illustrated in PFT cycle window 412. The amplitude and phase (or in-phase and quadrature components) acquired by the lock-in amplifier for each cycle can be averaged as a complex value during successive cycles, as previously described. In a further embodiment, the lock-in device demodulates the laser-induced signal only during the entire PFT cycle time t PFT For example, between 150 and 370 microseconds as shown at 414. The disadvantage of this method is that the lock-in device demodulates the noisy deflection signal only during the portion of the PFT cycle where the tip and sample are not in contact. c The noise added to the actual signal, present only during the period, reduces the overall signal-to-noise ratio of the laser-induced signal. Furthermore, when the laser repetition rate is close to the free resonance of the cantilever (or when it is at a fraction of 1 / n of the cantilever resonance (n = 1, 2, 3... integers), and the cantilever resonance is close to a harmonic of the laser repetition rate), the cantilever is driven by at least two possible effects: (1) sample absorption, resulting in surface pulse forces generating acoustic waves (when the AFM is operating in air or other environments other than a vacuum) that can effectively drive the cantilever, typically several micrometers to tens of micrometers above the surface in PFT. For example, the tip 203 has a length of 3-18 micrometers, and the PFT amplitude is 10-200 nm, so that the entire cantilever is approximately 3-18 micrometers above the sample. (2) Laser absorption by the cantilever material or coating reduces the contact time tc Unwanted cantilever vibrations can occur outside of the tip. In either case, the cantilever vibrations can lead to significant laser-induced signals that are not from the localized surface pulse force directly beneath the tip, but are artifacts due to cantilever absorption unrelated to sample properties, or, depending on the sample, may be non-localized, originating from a large region of the sample and away from the tip. Such non-localized artifacts are not the desired laser-induced signal in the tip-sample region.

[0067] In summary, it is preferable not to detect outside the probe-sample interaction cycle. c Alternatively, laser pulsing can be achieved through gated lock-in detection only during t c The lock-in device may be limited or gated to occur only during the PFT cycle time t PFT The signal can be demodulated over the entire contact time t c Because there is no laser pulse outside of t, acoustic effects and cantilever absorption do not drive the cantilever modes. The lock-in amplifier continuously demodulates over various PFT cycles 416 until the desired integration time is achieved. However, even then, the lock-in device may still see strong unwanted signals if the free oscillations 308 (Figure 3) are close to the lock-in reference frequency, which is the laser repetition rate. Gated laser pulsing (i.e., pulsing only during t) and gated lock-in detection (i.e., pulsing only during t) are used. c Lock-in signal demodulation (with only lock-in signal demodulation during the cantilever contact) is preferred due to several advantages. When the cantilever is not in contact with the sample, noise or signal generated by artifacts such as acoustic effects or cantilever absorption does not enter the signal demodulation, so the signal-to-noise ratio is not reduced. Also, fewer laser pulses applied to the sample result in less heating of the sample, which is beneficial for some samples, such as biological or polymer samples.

[0068] In this embodiment, because the oscillations 404 between successive PFT deflection traces are synchronized with a fixed phase shift of 0, the deflection traces, such as PFT cycles 1 and 2 or 1 through 4, can be synchronously averaged in the time domain before further processing to extract the laser-induced sample response. A timestamp or marker, such as the peak force tapping control point 310, serves as a time synchronization point at which the deflection traces are averaged. Such averaging increases the signal-to-noise ratio before the oscillations are analyzed, for example, to extract the sample's IR absorption that caused the deflection oscillations. When using a lock-in on this averaged time-domain data, the lock-in amplitude typically indicates the IR absorption signal, although other channels, such as the phase and in-phase and quadrature components, can also indicate IR absorption.

[0069] In a further embodiment, the laser pulses 400 are triggered by an event within each PFT cycle. A timestamp / marker, such as the peak force tapping control point 310 within each PFT cycle, can act as a trigger to fire a specific number of laser pulses in the pulse train after an adjustable positive or negative time delay. During laser firing, the pulses in the pulse train are fired at a constant laser repetition rate f laser In this operation, the emission of the laser pulse is triggered in the first PFT cycle, and the contact time of that first cycle is defined as t c (e.g., time 160-375 μs in Figure 4A) or, in the case of gated laser pulse emission, t c The timestamp / marker of the second PFT cycle can trigger the next train of laser pulses, which at point 418 indicates that the time delay between the last laser pulse triggered by the first PFT cycle and the first laser pulse triggered by the second PFT cycle is equal to the laser pulse cycle T laser That is, two separate pulse trains triggered by the first and second PFT cycles can be initiated with an equidistant time delay T between all pulses.laser is similar to a continuous pulse train with

[0070] Alternatively, at point 418, the time delay between the last pulse and the first pulse of the pulse trains of the first and second PFT cycles is equal to the laser pulse cycle T laser However, the deflection oscillations 404 do not exhibit phase delays during different PFT cycles because the laser pulses are still synchronized to each PFT cycle. This synchronization and lock-in allows the deflection signal to still be averaged in the time domain, and other extraction methods can operate on the averaged data. Because the phase difference between the reference oscillation 408 and the laser-induced deflection oscillations 404 is always constant, a lock-in amplifier or box-wheel signal integrator can operate continuously on real-time data, even in the presence of discontinuities 418, and average over multiple cycles.

[0071] In Figure 5, the contact time t cFurther embodiments of the present invention are presented based on the PFT vertical deflection signal in Figure 5A. Figure 5A shows a situation similar to Figure 4. The PFT vertical deflection trace shown at 502 without laser illumination shows clear oscillations 500 when driven by laser pulses 504. Specifically, for the PFT cycle shown, vertical time markers 402 are shown overlapping the laser pulses at 504. In contrast, Figure 5B shows a situation with laser pulses 508, where the pulses are offset by half a time cycle between laser pulses, i.e., marker 402 does not overlap the current laser pulse but is centered between the pulses at 508. In other words, pulse trains 504 in Figure 5A and 508 in Figure 5B show a 180° phase shift from each other. Such a phase shift represents a 180° phase shift in the laser-induced oscillations 506 compared to 500. Such a phase shift is not observed in FIG. 4 , where the oscillations in all PFT cycles are synchronized to a defined time point within each PFT cycle (e.g., peak force tapping control point 310). However, in the embodiment of FIG. 5 , the laser pulse repetition rate is such that the oscillations in successive PFT cycles are 180° out of phase. The laser-guided real-time signal extraction methods described above with reference to FIGS. 3 and 4 remain valid. That is, gated or ungated lock-in demodulation or box-wheel integration for gated or ungated laser pulsing also apply to FIG. 5 .

[0072] However, signal averaging in the time domain is now different. If successive pulses are applied synchronously in the time domain to extract the laser-induced sample response in vertical or horizontal cantilever deflection, the laser-induced vibrations in successive PFT cycles 500 and 506, respectively, are perfectly out of phase, resulting in vibration cancellation. In this case, a subtraction of two consecutive PFT deflection traces is required. Figure 5C illustrates such an operation, i.e., subtracting trace 506 from trace 500, resulting in subtracted trace 510, which is called the subtrace. 1-2Specifically, after this subtraction process, the amplitude of the oscillation 510 doubles, and the slowly varying PFT deflection trace 502 is canceled out, i.e., without laser illumination, the background 512 has almost zero deflection. This operation effectively removes the slowly varying PFT deflection signal. Such slowly varying signal 502 may add noise to the data extraction method if the slowly varying deflection frequency is sufficiently close to the laser pulse repetition rate. Conventional methods for removing such slowly varying signals include subtracting consecutive PFT deflection traces between pulsed and non-pulsed PFT cycles, which implies a reduced duty cycle. The embodiment shown in FIG. 5 does not rely on non-pulsed PFT cycles, and this twice-higher duty cycle is expected to improve the signal-to-noise ratio by sqrt(2) when applying a laser pulse every PFT cycle compared to applying a laser pulse every two seconds as in some prior art techniques. Another conventional method for removing the gradually changing curvature 502 of the PFT cantilever deflection is to fit a polynomial function and then subtract it (H. Wang et al., Anal. Chem. 93, 3567 (2021)). The disadvantages of such fitting methods are that real-time fitting is computationally expensive and is only an approximation, whereas the embodiment of FIG. 5 simply removes the measured (unfitted) background through a subtraction process. After using the described method to arrive at the subtracted signal 510, the curve can be further processed to extract the intensity of the laser-induced signal, for example, using a lock-in or boxcar-based method. An FFT algorithm can also be applied to such time-domain data, and the integral of the spectral amplitude around the laser repetition rate, the integral around harmonics of the laser repetition rate, or the average of the fundamental and harmonics would serve as the laser-induced sample response signal, indicating IR absorption.

[0073] The advantage of the PFT cycle subtraction method is explained with respect to artifact and noise suppression based on Figure 5D. Random noise superimposed on the deflection signal is partially canceled when consecutive PFT cycles are subtracted. Artifacts synchronized with the PFT cycles are also reduced. Figure 5D shows deflection measured without laser illumination. In a single deflection trace 514, significant oscillations are excited after the snap-in contact time 306. These oscillations at the cantilever contact resonance decay in amplitude. Averaging 4800 PFT cycles reveals that these oscillations are still present at 516. They are synchronized to add up consistently when averaged over thousands of cycles (note: 516 is vertically offset compared to 514 for clarity). These oscillations are already present even without laser illumination and can disrupt the laser-induced signal. This can occur especially when frequencies are close to or overlap each other, since both are contact resonances. Furthermore, harmonics of the oscillations generated by the sudden snap-in contact can overlap in frequency space with the laser-induced changes. 516 is shown to be completely coherent over the artifact-rich PFT cycle, so the difference method described above removes such artifacts and reduces their effect on actual laser-induced deflection changes.

[0074] In Figure 5, a 180° phase shift between successive PFT cycles is illustrated. Such a phase shift can occur with either continuous pulsing or gated pulsing. For example, with continuous pulsing, a pulsing frequency of 1301 kHz or 1303 kHz will result in a 180° phase reversal every 2 seconds for a 2 kHz PFT frequency. At 1300 kHz or 1302 kHz, no such reversal occurs, and the deflection oscillations are in phase between PFT cycles. In general, for a 180° phase shift to occur between successive PFT cycles, the laser pulsing frequency must be an integer multiple of the PFT frequency plus or minus half the PFT frequency. That is, f laser =(n±1 / 2)*f PFTwhere n = 1, 2, 3, etc. In gated pulsing, a time delay corresponding to half a laser pulse cycle can be introduced between the first and second PFT cycles before the start of the pulse train. This creates a 180° phase shift between the two cycles, which can be repeated in subsequent cycles. Delaying the laser pulse cycle by 1.5x, 2.5x, 3.5x, etc. also works. However, the oscillations of the first PFT cycle do not correspond to the 180° shifted oscillations of the second cycle until the laser pulsing begins in the second cycle. This results in a rapid change in the oscillation amplitude after subtraction, further reducing the duty cycle by using fewer laser pulses.

[0075] 5 has been described based on a 180° phase shift of the laser pulse between two consecutive PFT cycles, other embodiments can be deduced in a straightforward manner. For example, such a phase shift may be configured to occur between PFT cycles 1 and 3. In this case, oscillations 506 and 500 between cycles 1 and 3 are 180° out of phase, which is referred to as sub- 1-3 where the gradually varying baseline is again removed. The same applies to cycles 2 and 4, 2-4 The resulting curve sub 1-3 and sub 2-4 etc. are synchronously averaged, and laser-guided data extraction is sub 1-3 and sub 2-4, etc., and the results are averaged. In this example, the phase shift of the laser-induced oscillations 500 or 506 between consecutive PFT cycles is changed not only by 180 degrees, but also by 90 degrees (or 270 degrees, or more generally, n × 90 degrees, where n = 1, 3, 5, ... odd integers). If the oscillations in cycles 1 and 3 are 180 degrees out of phase, the same applies to cycles 1 and 5, or 1 and 7, since, of course, the two related frequencies, the PFT frequency and the laser repetition frequency, are fixed in the preferred embodiment. Therefore, other background subtraction combinations can also be employed. However, for imaging involving time-dependent changes in XY sample position, it is preferable to combine and average only PFT cycles that are close in time so that inferred signals from laser-induced sample deformations (e.g., absorption) can be attributed to a small number of spatial XY pixels. Otherwise, the spatial resolution of the laser-induced sample deformations is low. For example, at a typical PFIR line scan rate of 1 Hz, a 256-pixel line corresponds to approximately 2 ms per pixel for the trace and 2 ms for the retrace. Typically, since only the trace or retrace is recorded, the maximum time over which the PFT cycles must be averaged over various spatial pixels is 2 ms. If the PFT frequency is 2 kHz, then only four PFT cycles occur over a 2 ms pixel time.

[0076] In imaging, it is desirable not to average over too many spatial pixels to avoid degrading spatial resolution below the desired minimum of 2, 5, or 10 nm, but similar problems arise in nano-spectroscopy at fixed spatial locations. In spectroscopy, the laser wavelength is swept and the sample response is recorded to arrive at a wavelength-dependent sample response curve, such as a nano-IR absorption spectrum. That is, the wavelength is held constant for a specific short time, e.g., 10 ms for swiping across a 1000 cm-1 wide range in 5 seconds with 2 cm-1 resolution. If the averaging time for extracting the laser-induced sample response is much longer than 10 ms, the spectral resolution will be reduced as the data is averaged over multiple wavelengths.

[0077] In the cases of Figures 4 and 5, the contact time t c We describe an embodiment in which the phase of the laser-induced oscillations is constant between successive PFT cycles (FIG. 4) or shifted by 180° (FIG. 5) within a given time. In a more general embodiment, the phase change between successive probe-sample interaction cycles is not 0° or 180°. In this case, only PFT cycles that are far apart from each other will exhibit a phase shift close to 180° in the deflection oscillations, which can be subtracted as shown in FIG. 5C by a substantially flat, i.e., zero, background 512 and a substantially constructive interference 510 of the two selected deflection traces. Because the PFT cycles are far apart from each other, averaging over multiple spatial pixels in the imaging or spectral step of nanoscale spectroscopy becomes problematic, resulting in reduced spatial or spectral resolution. Here, when gated, the contact time t c 4, lock-in demodulation or box-wheel signal integration based on individual PFT deflection traces, such as 404 shown in FIG. 4, is preferred during the entire PFT cycle 414 if not gated. That is, phase-sensitive averaging can be applied between the data of individual PFT cycles, or lock-in demodulation or box-wheel signal integration can be performed continuously over multiple cycles until the desired integration time is reached.

[0078] Note that in this mode of operation, the laser repetition rate is not synchronized with the PFT cycle frequency. For continuous pulsing, gating the signal extraction is preferred, and the gated extraction window must overlap the probe-sample interaction cycle for the highest duty cycle and signal-to-noise ratio. In other embodiments, the laser pulses may be gated for continuous or gated signal extraction. In one such embodiment, the laser receives a TTL signal at the reference frequency of a lock-in or box-wheel integrator, but this TTL signal occurs only during the tip-sample interaction cycle. This allows the laser to emit pulses synchronized to the reference frequency, while limiting the pulse train to the tip-sample interaction cycle. The lock-in amplifier or box-wheel signal integration may occur continuously over multiple PFT cycles. For example, a 10 ms integration time is applied, averaging 20 PFT cycles for a 2 kHz PFT frequency. Alternatively, signal extraction may be gated and limited to the probe-sample interaction cycle.

[0079] Detection gating can be achieved, for example, by continuously reading data from the lock-in amplifier and selecting which data to keep and which to discard in controller 214. Additionally, deflection data input to lock-in, FFT, or boxcar-like detection can be partially replaced with data that does not increase noise due to continuous detection. For example, lock-in integrated over time window 414 typically includes noise that occurs during the 250-375 microsecond time interval in FIG. 4, when the tip has already left the surface. To prevent this, deflection data within the corresponding range can be replaced with an appropriate constant value (e.g., the average of the lock-in output within window 410). In this case, calculating the average value for signal 414 as well as signal 410 does not alter the lock-in reading in terms of value or noise. Similar access methods can be applied to other extraction methods. For example, an FFT algorithm can operate on the subtracted deflection data 510 shown in FIG. 5, where deflections beyond the probe-sample interaction cycle are zeroed out. However, in the case of the FFT, it is more preferable to reduce the number of data points over which the FFT algorithm is performed in order to save computation time, which makes it advantageous to perform gated signal extraction only during the contact time.

[0080] While the laser-induced signal has been described above as an oscillation of the deflection channel, more complex signal forms are possible. This is illustrated in Figure 6, along with a signal extraction method optimized to obtain the best signal-to-noise ratio. Figure 6A shows the signal with a pulse length of D tlaser and the interval T laser At time t laser0 6 shows a typical laser pulse train 600 with laser pulses starting from D tlaser and f laser =1 / T laser Typical values ​​for are 1 picosecond to 1 microsecond and 50 to 3000 kHz, respectively. Note that the rectangular shape of the laser pulse is an approximation, and other rising and falling edges are possible.

[0081] In FIG. 6B, the contact time t c The laser-induced deflection trace 602 is shown in the figure, after subtracting the gradually varying contact PFT deflection. In this figure, the laser is not fired, resulting in a flat line. For example, the laser pulse D tlaser During t the sample absorbs and expands, resulting in a linear change in the deflection signal over time 604. The change is generally proportional to the laser pulse itself, i.e., t defl、0 =t laser、0 When the laser pulse ends, no more energy is delivered to the sample, and thermal diffusion causes the sample to cool and the deflection signal to decrease. This decrease undergoes an exponential decay 606 until the next laser pulse starts the heating and cooling cycle again. T laser Depending on the time scale of the pulse and the relaxation time of the sample, which is determined by sample attributes such as thickness and thermal coupling with the substrate, the sample may cool between pulses to a temperature close to the temperature it would have had in the absence of the laser pulse. In such a case, shown by curve 608, the sample exhibits a fast exponential decay towards the same "undisturbed" sample temperature as if the sample had not been pulsed. If the cooling time is T laser If the PFT contact window is longer than , the sample temperature increases, and the change in deflection due to sample absorption and thermal expansion also increases with the first laser pulse in the pulse train at the beginning of the PFT contact window. After the first laser pulse occurs, an equilibrium state is reached, meaning that the sample temperature cycles between minimum and maximum temperatures before and after each laser pulse, but the minimum and maximum temperatures for each pulse in the pulse train remain the same and continue to rise. Thus, after equilibrium is established, the deflection baseline or offset is higher than for a non-pulsed PFT cycle; i.e., curve 602 has a higher baseline and is vertically offset with a smaller deflection change due to laser pulsing. Depending on the sample's thermal properties, the laser repetition rate, its relationship to the probe's contact resonant frequency, and the probe's characteristics (e.g., q factor), laser pulsing can result in, for example, a softer increase 604, a decrease step 604 for negative thermal expansion samples, or a large deflection change tdefl、0 and the start of the laser pulse t laser、0 Such delays between the time t1 and t2 may result in a different probe response than 602 or a noticeable oscillation 314.

[0082] 6C illustrates a detection method other than the lock-in based detection method described above. Curve 610 shows an example of box-wheel type signal integration and averaging. g0 The deflection signal is g1 It is integrated during the dead time t g2 In a preferred embodiment, this method is applied to the deflection signal 602 after removing the slowly varying in-contact PFT deflections, acquired as described in FIG. 5 or using a non-pulsed PFT cycle with a reference for subtraction. In general, this approach works on raw PFT deflections as well as 304. Starting t g0 is the laser pulse t such that the integration gate coincides with the signal of interest (where the deflection increases 604 and decays 606). laser、0 Selected for t g1 The integrator detects the signal when the gate is open at t g2 When the gate is closed at t, noise and interference are not detected. This is because the input signal (i.e., the deflection) is multiplied by a boxcar function, t g1 This corresponds to integrating over . Boxcar averaging then occurs via gated integration of many laser pulses to perform potential averaging in the next PFT cycle. Curve 612 shows an example of a narrower signal integration gate suitable for detecting the faster decay of curve 608. Otherwise, if measured sequentially over time, the time interval after 608 has fully decayed would contribute to the captured noise but would not carry the laser-induced signal, resulting in a lower signal-to-noise ratio. Note that in a preferred embodiment, the signal integration or accumulation is normalized according to the acquisition time, the number of laser pulses, or the number of ADC samples being integrated.

[0083] Other gating functions may also be suitable. Curve 614 shows such an optimized gating function, where the deflection signal is integrated during the linear rise 616, but the results are weighted by the gating function. That is, data points acquired at and near the peak of the gating function have the highest weighting. An exponentially decreasing portion 618 of the gating or weighting function may also be applied. Such a gating function may be tailored to the particular sample and excitation conditions (e.g., laser pulse length) to optimize the rise time 616 and fall time 618 while minimizing noise integration.

[0084] While lock-in amplifier and FFT-based signal extraction require equidistant periods between laser pulses within a PFT probe-sample interaction cycle, i.e., a fixed laser repetition rate, box-wheel integration allows for non-equidistant pulsing by using a more general gating function. The laser pulse distance can vary from laser pulse shot to laser pulse shot within and between PFT probe-sample interaction cycles, as long as the gating for box-wheel integration is substantially consistent with the presence of laser-induced deflection changes. This is achieved by using a gate length t g1 and t g2 This means that even if the λ / 2 varies, or the cycle between pulses varies, it still must be synchronized to the laser pulse. This pulsing scheme does not allow for good control over the emission time, but it does require a light source whose emission time can be determined (e.g., via a photodiode) and used to trigger the boxcar detection.

[0085] It should be noted that there are other ways to obtain the laser-induced signal from the deflection curve 602 or the more oscillatory motion 314 or 510. For example, the rms of the signal can be determined, a sine or cosine curve representing the laser-induced signal can be fitted, and the minimum and maximum vibration amplitudes can be read out. A bandpass filter around the laser repetition rate or a highpass filter below it can be applied to the deflection signal before further data extraction.

[0086] Laser pulsing does not need to occur at the contact resonance of the probe when in contact with the sample. Pulsing can occur off-resonance. In particular, for continuous pulsing and continuous detection, it is beneficial not to excite and detect at the free cantilever resonance W or higher modes, ~6.3 W, or at 1 / n (n = 1, 2, 3... integer) fractions of such modes, in order to avoid driving the cantilever as it lifts off the surface. After lifting off the surface, the cantilever can oscillate at its free resonance frequency, and detecting from such frequencies would only result in more unwanted noise and background. In a specific implementation, laser pulsing can occur at approximately twice the free cantilever resonance W, i.e., 2 W. In that case, excitation at twice the free cantilever resonance will not excite the cantilever if the laser pulsing is continuous and not limited by the PFT contact time. Even if the sample absorbs or experiences acoustic waves, driving at 2 W effectively suppresses the vibrations by "kicking" the cantilever oscillation in phase with one another, then out of phase with another, etc. However, even after the cantilever leaves the surface after the adhesion point, it still exhibits the fundamental free resonance and harmonics, but is not actively driven by the laser.

[0087] If a contact resonance is selected for laser pulsing, it should be noted that the contact resonance generally shifts due to changes in the mechanical tip-sample interaction. This can occur due to changes in AFM parametrics (e.g., peak force setpoint) or due to changes in nano-mechanical sample properties, especially when scanning inhomogeneous samples. For example, a drop in the IR signal in the sample region can result from a decrease in IR absorption or a change in modulus that shifts the contact resonance, resulting in artifacts and ambiguities. To compensate for this, frequency tracking, e.g., via a phase lock-in loop of the IR signal, is preferred and is well compatible with signal processing based on lock-in using amplitude and phase channels. Such a tracking mechanism, for example, dynamically adjusts the laser repetition rate to overlap with the contact resonance. Alternatively, the effect of a contact resonance shift can be minimized by operating in an off-resonant manner, i.e., by pulsing and detecting the laser away from the cantilever resonance, where the laser-induced signal responds less sensitively to the mechanical tip-sample interaction.

[0088] Signal extraction from the deflection trace using a lock-in amplifier, a boxcar signal integrator, an FFT, or similar techniques can also generate signals at higher harmonics of the laser repetition rate. If the cantilever is excited by the sample at the laser repetition rate during the PFT contact time, the deflection response 602 shown in Figure 6B is asymmetric and not purely oscillatory, and therefore can generally transmit harmonics higher than the fundamental frequency. This is useful in embodiments where the cantilever itself absorbs laser energy, generating nonlocal laser-induced vibrations that compete with the vibrations generated by the localized surface pulse force in the tip-sample region. Cantilever heating due to such laser excitation typically results in cantilever motion with an approximately sinusoidal shape. By detecting even higher harmonics, it is possible to suppress the delocalized cantilever motion at the fundamental frequency and recover the localized signal under the tip. A disadvantage of this method is the reduced signal strength at higher harmonics.

[0089] 7 to 10 show that laser pulsing and signal detection are not gated, and the PFT contact time t c This highlights the problems that arise when the entire PFT cycle is evaluated and exposed to continuous pulsing. Figure 7 shows the IR response 700 to continuous pulsing acquired by a continuously operating lock-in amplifier. The IR wavelength was selected at 1730 cm-1 to match the carbonyl absorption peak of the irradiated PMMA film sample. While the IR laser repetition rate was adjusted, the lock-in extracted signal amplitude was recorded as signal maxima 702 and 704 at data points 1394.6 kHz and 1402.7 kHz. Note that while discrete points are shown at random frequencies, continuous repetition rate sweeps are possible with this system. Two Lorentzian curves have been applied to the two peaks. When the tip is removed from the sample by a few hundred micrometers, the broad peak disappears, while the narrow peak at 1394.6 kHz remains at the same frequency (2 Lorentzian feet). In a control test where the tip recontacts the sample in PFT mode, the IR laser pulse is applied for a contact time t c The narrow peak disappears, and only the broad peak remains visible after fitting (Lorentzian feet). This proves that the narrow peak originates from the free resonance of the cantilever, whereas the broad peak originates from the localized surface pulse force felt only when the tip is in contact with the sample. When measuring nanoscale IR absorption, the desired local sample signal of interest is contained only in the broad peak. As previously mentioned, the presence of an even narrower free cantilever resonance inhibits the detection of the desired nano-IR signal. Here, we provide more experimental evidence of the adverse effects of continuous laser pulses and ungated detection.

[0090] Figure 8 shows the setup for acquiring the intensity map shown in Figure 9, which was acquired during the alignment step of the nano-IR setup prior to IR absorption measurements. When installing a new AFM probe, the IR laser light focus of beam 222 must be realigned in x, y, and z coordinates to the new position of tip 203 on sample 204 so that the IR signal can be measured in the tip-sample region. To do this, a visible alignment laser collinear with the IR beam is first focused near the tip. Next, the IR beam focus is scanned over the tip and the IR signal is recorded; the signal is maximized only when the IR beam focus overlaps the tip-sample region. The beam focus can be scanned in various ways. One method is to scan the focusing element 220, such as an off-axis parabolic mirror or lens, in x, y, and z while collecting the IR signal. Another method is to scan the angle of the incident IR beam 222.

[0091] The resulting intensity map of this signal retrieval is shown in Figure 9. Successive pulsing and successive lock-in detection at the 1403 kHz contact resonance 704 results in a number of signal maxima at 706 related to the cantilever geometry. c When pulsing only during contact, such gating significantly reduces the artifact, but there is still a single bright spot at 708 corresponding to the tip-sample position. The remaining artifact is due to ungated detection, which is still sensitive to the free-space vibration of the cantilever at 702. As mentioned previously, this vibration occurs after the cantilever lifts off (even without laser pulsing), and the frequency of the vibration is close to the detection frequency. At 709, the detection is additionally blocked, and the previous cantilever-related artifact is absent. Clearly, pulsing and detecting only at contact at 709 is more suitable for this alignment. While the optimal alignment of the laser beam focus with respect to the tip-sample position is clear, with ungated pulsing and detection, various signal maxima in image 706 can obscure the optimal alignment and potentially optimize for the wrong signal, e.g., optimizing for overlap of the beam focus with the cantilever 202 (instead of the tip 203).

[0092] Figure 10 shows various IR responses with IR laser wavenumbers for different intensity maps in Figure 9 in the PFIR of a PMMA film sample. All curves are normalized to unity and vertically shifted for clarity. The bottom curve 710 shows the IR laser-induced signal acquired by a gated lock-in amplifier (i.e., limited to the probe-sample interaction cycle) demodulating the vertically polarized signal during gated laser pulsing. The extracted amplitude signal is proportional to the sample absorption and exhibits a carbonyl absorption peak near 1730 cm ~1. Note that the signals presented here are not normalized by the laser power spectrum, which is typically required to acquire nano-IR absorption spectra. Such normalization by the laser background removes the effect of a non-constant laser power spectrum on the absorption signal and eliminates the waterline. Deep dips in the lock-in amplitude signal 710, e.g., at 1716 cm ~1, are also observed. -1 or 1732 cm -1is due to water absorption in the IR beam path (which cancels out when the signal is normalized to the laser power spectrum). Importantly, this single absorption line 710 was acquired with the IR laser beam aligned with the single bright hotspot at 709 in FIG. 9 . In contrast, curve 712 was acquired with the IR beam aligned with the center of image 706 shown in FIG. 9 and continuously pulsed and detected. Again, the carbonyl absorption is visible, but it is superimposed on a broader background, resulting in a raised baseline. The spectrum measured at the more distinct hotspot in the center of 708 falls between the extremes of spectra 710 and 712 (not shown), i.e., it has a smaller but higher broad background. Depending on the selected laser repetition rate, other hotspots or bright spots in image 706 (other than the center) shown in FIG. 9 can exhibit signals similar to curve 714. The carbonyl resonance is visible or barely visible, but with a broad background (water absorption lines are present). Curves 712 and 714 (sequential pulsing and detection) arise from the effects described above: the cantilever is excited by absorption of IR radiation and / or acoustic, long-range waves drive the cantilever, starting from sample absorption below the cantilever. In either case, the feature of interest in nano-IR absorption, i.e., the absorption peak of the carbonyl resonance in this example, can be obscured or heavily contaminated by these artifacts.

[0093] The effect, which can be seen with reference to Figures 7-10, is independent of the detection method; i.e., FFT-based signal extraction exhibits the same behavior as lock-in-based signal extraction, and applies equally to boxcar integration. The size of the artifact varies depending on the test conditions (e.g., IR spot size) and AFM probe attributes. For example, the free-space resonance frequency of the cantilever may not overlap significantly with the contact resonance for other geometric features of the cantilever. Alternatively, the tip 203 may be spatially separated from the cantilever 202 so that the IR spot does not significantly overlap with the cantilever 202 when the tip-sample region is excited. Alternatively, the contact time may occupy a larger portion of the PFT cycle, obscuring artifacts that occur outside of the contact time.

[0094] This method is a true multimode spectroscopy technique capable of simultaneously acquiring IR and nanomechanical data, both of which require PFT AFM operation. Figures 11A-11D show the combination of PFIR with adhesion and elastic modulus measurements using gated IR pulsing and sequential lock-in detection. In this embodiment, the laser repetition rate was not selected at the contact resonance but set at 662 kHz, a 30% reduction in the IR signal compared to the 789 kHz peak. The advantage of such non-resonant driving is that it minimizes mechanical artifacts in the IR image, which can result in higher image contrast and more image artifacts due to shifts in the contact resonance. At selected frequencies away from the resonance peak, the IR signal remains constant with frequency. For example, PLL-based frequency tracking during scanning would follow such changes, but they are not observed here. The sample was a PMMA bead in epoxy, scanned at 0.5 Hz with a 1.5-micrometer scan at 512 x 128 pixels. All images were collected simultaneously to prevent thermal anomalies that can occur when taking sequential images. The height image of the panel (a) shows a central feature with lower adhesion (b) and even lower modulus (c) than the rest of the image. Panel (d) shows a 1730 cm -1The IR absorption measured at 1000 kJ / cm2 is shown. The PMMA beads are bright in the lower right corner, but the center is not as bright.

[0095] Figure 11E shows the PMMA carbonyl absorption at a wavelength of 1730 cm -1 A larger scale IR scan of such a PMMA bead in an epoxy matrix is ​​shown in Figure 11F. The round, absorbing PMMA bead is lit in a non-absorbing matrix. The scan is performed at 0.5 Hz for a 10 micrometer scan size with 512 pixels. For the other probes, a laser repetition rate of 1400 kHz at contact resonance with a 0.4 N / m spring constant and a force of 1.5 nN at a 30 nm PFT amplitude were used. In this embodiment, the laser pulsing and lock-in detection are limited to only the contact time within the PFT cycle. Lock-in is realized inside the FPGA. The spectra for the PMMA bead and epoxy matrix are shown in Figure 11F. 2 cm -1 Ten (10) spectra were averaged with a spectral resolution of 10 s, each measured for 10 seconds. Generally, to obtain the nano-IR absorption from the chip response with wavelength, the latter was normalized with the laser output power measured with a power meter under similar test conditions. The system was purged with dry air to remove water lines.

[0096] Figure 12 highlights the spatial resolution achievable with embodiments of the present invention. The PS-b-PMMA block copolymer exhibits a resolution of 1730 cm -1 and the PMMA domain absorbing at 1599 cm -1The image visualizes the distribution of PS domains absorbing at 1000 nm. Each 400 nm scan image was acquired at a 1 Hz scan rate with 512 × 256 pixels within a 5-minute scan time. A 5 N / m spring constant probe was used with a PFT tapping amplitude of 30 nm and a PFT force setpoint of 3 nN. The typical PFIR range is from tens of piconewtons to several nanonewtons. In Figure 12A, gate pulsing (close to the contact resonance at approximately 961 kHz) and gate lock-in detection were used only during the contact time, demonstrating a spatial IR resolution of 8 nm, as shown in the line cut in Figure 12B. In Figure 12C, after subtracting the deflection of successive PFT cycles as shown in Figure 5, an FFT readout routine was employed to remove the gradually changing background and enhance the signal-to-noise ratio. Here, the laser repetition rate was 961 kHz, close to the contact resonance of the probe. Images were acquired at a 1 Hz scan rate with 512 × 256 pixels. The FFT amplitude is integrated over a ±20 kHz window centered around the laser repetition rate; this routine is implemented by the controller DSP. The IR spatial resolution is the same as before.

[0097] The preferred embodiment of gated pulsing and detection (sample response extraction) via lock-in amplifiers, boxcar integration, FFT, or similar methods offers speed advantages over conventional techniques. While previous single-pulse or multi-pulse excitation PFIRs have reported scan rates of 0.1–0.2 Hz, the preferred embodiment allows scan rates in excess of 0.5 Hz or 1 Hz, achieving 3–10x improvements. Such improvements are possible through a larger pulsing duty cycle during the entire body contact time and the inclusion of a non-pulsed PFT cycle without IR illumination. Furthermore, noise and artifacts are reduced through synchronization between pulsing and phase-sensitive detection, enabling phase-sensitive averaging without signal reduction through partially destructive interference. Background subtraction, as shown in Figure 5, further reduces noise and also reduces the amplitude of contact resonance vibrations caused by the "shock" of a fast snap-in contact as the PFT probe-sample interaction cycle begins.

[0098] It will be appreciated that in alternative embodiments, the wavelength range can be extended beyond the infrared of the preferred embodiment, for example, into the ultraviolet, visible, near-infrared, and terahertz or far-infrared spectral regions. QCLs and optical parametric oscillators exist as pulsed and modulated light sources in the infrared. The ultraviolet, visible, and near-infrared regions are handled by laser sources such as solid-state lasers, fiber lasers, diode lasers, optical parametric oscillators, or gas lasers, as well as laser sources based on nonlinear frequency conversion, including optical parametric generation, sum-frequency generation, harmonic generation, frequency comb generation, and related methods. In the terahertz spectral region, terahertz quantum cascade lasers are emerging, but terahertz gas lasers, terahertz antennas, or free-electron lasers already exist to encompass this range. The extended wavelength range from UV to terahertz allows for the surface pulsing of laser light resulting from various effects. In the terahertz region, plasmon polaritons in graphene or Cooper pair polaritons in superconductors exist, which can induce electromagnetic forces between the probe and sample under light excitation due to charge redistribution and charge oscillation. Another example is phonon resonances, which lead to absorption and photoexpansion in the terahertz range. In the ultraviolet, visible, and near-infrared ranges, for example, plasmonic resonances exist in metal nanostructures, which absorb energy for photoexpansion or modify the electromagnetic field through charge oscillation or charge redistribution, applying a surface pulse force to the probe.

[0099] In a further embodiment, the sample is illuminated from below, instead of the top-down illumination shown in Figure 2. Bottom illumination requires a transparent sample or a thin enough film (within a few wavelengths) to transmit light in the wavelength range of interest to the illuminated area. Bottom illumination has the advantage of exposing the tip 203 and probe 201 less to the laser pulse, reducing artifacts that can occur when the probe itself absorbs and heats up. Another advantage is that bottom illumination allows for a higher numerical aperture than top illumination, because the probe blocks some of the light with top illumination, whereas bottom illumination allows for the entire lower half of the probe to be used for focusing light. Therefore, a smaller focal spot reduces the laser power required and the heating of the sample. The main advantage of bottom illumination is that it enables PFIR of samples in liquid environments, as described below.

[0100] For bottom illumination, the sample may be placed on or spin-coated with a prism of a transparent material suitable for the wavelength range of interest (e.g., ZnSe or ZnS or diamond or germanium). Since the laser light is extinguished in air but continues to propagate inside the sample, total internal reflection occurs. In this way, only the sample is exposed to the radiation, resulting in a strong light-matter interaction.

[0101] This bottom-up configuration is most useful for measuring liquids. The chip and sample area are then surrounded by a fluid, for example, to study biological samples or electrochemical reactions in their natural environment. Water can be used as the liquid to minimize water absorption in the ultraviolet to near-infrared spectral range compared to the infrared range, allowing for the study of near-infrared absorption in biological materials in their natural environment. The wavelength range can be extended by using other suitable liquids, such as heavy water, which have no or minimal absorption in the wavelength range of interest. Compared to top-down illumination, where light travels longer through the liquid, bottom illumination minimizes water absorption.

[0102] In bottom-up illumination using a prism, total internal reflection of light creates an evanescent field, which induces light absorption from IR modes or electronic transitions in the sample within the evanescent field. The electronic transitions can be ultraviolet, visible, or near-infrared, for example, using fluorophores used in fluorescence microscopy. Absorption and the resulting non-radiative conversion of electronic transitions leads to thermal effects or pulsed forces due to mechanical responses, typically detected by PFIR. In other embodiments, prisms and total internal reflection are not required, but light is focused directly onto the tip-sample interaction region from below. The sample and substrate (e.g., ZnSe) must be transparent to allow light to penetrate and reach the tip-sample interaction volume.

[0103] A method 1300 of some preferred embodiments is shown in the flowchart of FIG. 13. Method 1300 includes oscillating the probe-sample distance in step 1302, for example, with peak-force tapping (PFT mode) at a frequency below the probe's resonance frequency (e.g., 2 kHz). In step 1304, the probe interacts with the sample during a tip-sample contact time at a specific x-y location of the sample. The maximum probe-sample interaction force at peak-force tapping is controlled in step 1306 to ensure good tip-sample contact and prevent sample damage for detecting laser-induced sample reactions in the next step of method 1300. Next, in step 1308, the tip-sample region where the tip contacts the sample is illuminated with pulses from a light source at the wavelength of interest, for example, an infrared laser in preferred embodiments. The pulse repetition rate is preferably set at or near the probe's contact resonance, although non-resonant pulsing is acceptable at the expense of signal-to-noise ratio.

[0104] In step 1310, laser pulsing is selected to occur continuously or gated, e.g., limited to a variable pulse train that substantially overlaps a probe-sample interaction cycle limited to the tip-sample contact time. In the latter "gated" case, the laser pulses are preferably synchronized to the probe-sample interaction cycle, arriving at the same relative time point in each cycle, but this is not required. In step 1312, continuous or gated detection is selected during the probe-sample interaction cycle. The gradually changing background of probe deflection, which repeats with each probe-sample interaction cycle, is subtracted in step 1314. To this end, the laser pulse repetition rate should be set in step 1316 so that the deflection changes induced by the optical pulses are out of phase with each probe-sample interaction cycle, preferably by 180° during successive cycles. For example, for a 2 kHz PFT frequency, a laser repetition rate of 751 kHz ensures phase reversal between successive cycles. In step 1318, the deflections of cycles with a 180° phase shift are synchronously subtracted (i.e., subtracting the same time points relative to a synchronized timestamp, such as the peak force tapping control point). Alternatively, the background can be removed using a non-pulsed probe-sample interaction cycle or fitted to a slowly varying background as a reference for subtraction. If background subtraction is not required, the probe deflection in step 1320 is used for further processing instead of the deflection subtracted in step 1318. In step 1322, the sample response is extracted from the deflection acquired in step 1320 or the processed deflection 1318. Synchronized averaging of the time-domain deflection data in steps 1318 or 1320 is an option to increase the signal-to-noise ratio as part of or before this extraction step 1322.

[0105] Extracting the sample response in step 1322 preferably uses a lock-in amplifier, a boxcar signal integrator, or an FFT routine / algorithm. In other words, averaging is optional; if the output of the extraction step provides phase information (e.g., in the case of a lock-in amplifier or FFT implementation), averaging of phase-sensitive responses is preferred. Note that the sample response is extracted from the deflection after measuring the deflection of at least one entire probe-sample interaction cycle, particularly when using a lock-in or signal integrator, i.e., extracted in real time as part of the probe-sample interaction cycle, even before the cycle is completed. Once the sample response is extracted, step 1322 is repeated in step 1324 to collect sample responses with more wavelengths of the light source. A resulting spectrum of the sample response as a function of wavelength is generated in step 1326, which in a preferred embodiment represents the absorption spectrum after normalization with laser power, which varies with wavelength. Alternatively, in step 1328, the wavelength is held constant while changing the sample position, and step 1322 is repeated. In such cases, a spatial map is generated in step 1330, for example to show infrared absorption by position. Steps 1326 and 1330 can also be combined to generate hyperspectral data, i.e., a spatial map containing spectra that vary with position. In further embodiments, when changing xy position during a spatial scan or changing wavelength during spectral collection, the laser repetition rate may have to be adjusted in step 1308 to follow contact resonances in a frequency-tracking step.

[0106] While the best mode contemplated by the inventors for carrying out the invention is disclosed above, the practice of the invention is not limited thereto, and it will be apparent that various additions, modifications, and rearrangements of the features of the invention may be made without departing from the spirit and scope of the inventive concept.

Claims

1. An apparatus for performing spectroscopy on a submicron region of a sample using an atomic force microscope (AFM), comprising: The device comprises: a driver that generates an oscillatory drive signal that causes the probe of the AFM to interact with the sample and generate a transient probe-sample interaction force during a plurality of probe-sample interaction cycles, the oscillatory drive signal having a frequency that is lower than a resonant frequency of the probe; at least one controller that controls the transient probe-sample interaction force; a tunable light source that illuminates the tip-sample region with light pulses to induce deformation of the sample; a detector that measures probe deflection due to minimally partially induced sample modifications; The apparatus includes one or more of a lock-in amplifier and a signal integrator for extracting a sample response to the light pulse from the measured probe deflection.

2. at least one of the lock-in amplifier and the signal integrator is a lock-in amplifier; the sample response is phase sensitive; and The apparatus of claim 1 , wherein the phase-sensitive sample responses are averaged.

3. the at least one controller uses the sample response to generate a spatially resolved map indicative of absorbed infrared radiation; The apparatus of claim 1 , wherein the map is generated for the sample area of ​​at least 100×100 pixels within 5 minutes.

4. 10. The apparatus of claim 1, wherein the vibratory drive signal frequency is at least five times lower than the lowest resonant frequency of the probe.

5. The at least one controller adjusting pulse times during a probe-sample interaction cycle such that a 180° phase change occurs in said optically induced probe deflection during at least two cycles; subtracting the probe deflection corresponding to the at least two cycles; The apparatus of claim 1 , wherein the sample response is extracted from the subtracted probe deflection.

6. The apparatus of claim 5 , wherein at least one of a lock-in amplifier, a signal integrator, and an FFT algorithm extracts the sampled response.

7. 10. The apparatus of claim 1, wherein at least one of the light pulses and extracted sample response is gated during the probe-sample contact time.

8. 8. The apparatus of claim 7, wherein the at least one light pulse and extracted sample reaction is gated every cycle of probe-sample interaction.

9. The apparatus of claim 1 , wherein the controller extracts at least one of a nano-mechanical property and a nano-electrical property from the sample response.

10. 1. A method for performing spectroscopic measurements on submicron regions of a sample using an atomic force microscope (AFM), comprising: The method comprises: causing a probe of the AFM to interact with the sample during a plurality of probe-sample interaction cycles to generate a transient probe-sample interaction force with an oscillatory drive signal having a frequency lower than a resonant frequency of the probe; controlling the transient probe-sample interaction force; illuminating the tip-sample region with light pulses from an adjustable light source to induce deformation of the sample during the tip-sample contact time; measuring a probe deflection due at least in part to the induced sample deformation; adjusting the pulse time during a probe-sample interaction cycle so that a 180° phase change occurs in the optically induced probe deflection during at least two cycles; subtracting the probe deflection corresponding to the at least two cycles; extracting a sample response in the subtracting step; A method comprising:

11. The method of claim 10, wherein at least one of the irradiating and extracting steps is gated during the probe-sample contact time.

12. The method of claim 11 , wherein the at least two cycles are consecutive cycles.

13. The method of claim 10 , wherein the sample response is extracted from at least one of a lock-in amplifier, a signal integrator, and an FFT algorithm.

14. The method of claim 10 , wherein the extracting step further comprises averaging the phase sensitive output using at least one of a lock-in amplifier and an FFT algorithm.

15. The method of claim 10 , wherein the vibratory drive signal frequency is at least five times lower than the lowest resonant frequency of the probe.

16. 1. A method for performing spectroscopic measurements using an atomic force microscope (AFM), comprising: The method comprises: causing the probe of the AFM to interact with the sample during a plurality of cycles to generate a probe-sample interaction force with an oscillatory drive signal; controlling the probe-sample interaction force; providing a pulsed light source to generate a plurality of light pulses, each having a pulse width; directing the pulse into the sample in which the probe is located, causing an induced sample response; measuring a probe deflection due at least in part to the induced sample response; extracting a sample response to the light pulse from the measured probe deflection; Including, The method, wherein the extracting step uses at least one of a lock-in amplifier and a signal integrator.

17. 17. The method of claim 16, wherein at least one of the directing and extracting steps is gated during the probe-sample contact time.

18. 18. The method of claim 17, wherein at least one of the directing step and the extracting step is gated every cycle of the causing step.

19. The method of claim 16 , wherein the causing step is performed in PFT mode.

20. 17. The method of claim 16, wherein the resolution of the sample response is less than 20 nm.

Images