Ep 36 — Micro-FTIR: Seeing the Chemical Information of the Microworld

Series: Encyclopedia of Infrared Spectroscopy: From Principles to Practice
Chapter: Part 4 · Advanced Topics — Frontier Technologies (Episode 1, Opening of Advanced Section)
Target Audience: Advanced analytical technicians, materials/chemistry/biomedical graduate students, microplastic and defect analysis engineers, heritage and forensic scientists
Prerequisites: Ep 12 (Michelson Interferometer), Ep 14 (ATR Principle), Ep 26 (Microplastic Detection), Ep 32 (Cross-sectional Imaging of Artifacts), Ep 34 (Thin Film Analysis)
Reading Time: Approximately 50 minutes

Introduction: A "Fingerprint" in a 10 Micrometer Particle

Microplastic research has made μ-FTIR a routine tool for polymer identification in environmental samples. For example, Cole et al. review/empirical study on micrometer-sized particles in the digestive tracts of marine organisms indicates that many particles in the range of 10–500 μm require micro-FTIR to provide chemical identity one by one, to distinguish them from organic debris and to separate types such as PE, PP, PS, PET, PA, PVC [1].

"Without micro-FTIR, we would have had no chemical confirmation that these particles were synthetic polymers rather than organic debris."
— A consensus statement on μ-FTIR identification methodology for microplastics, see Cole et al. Mar Pollut Bull and related subsequent work [1]

(Early drafts mixed the narrative of Thompson 2017 with the volume/DOI of Cole 2011; now corrected to the verifiable reference [1].)

This is precisely the value of micro-FTIR (Micro-FTIR, often abbreviated as μ-FTIR): advancing the chemical identification of FTIR to the micrometer scale. This episode explains three microscopic modes: transmission/reflection/ATR, diffraction limit, synchrotron radiation enhancement, MCT vs. FPA trade-offs, as well as layered structures, defect localization, and microplastic imaging cases.

1. Why Do We Need to Make Infrared "Smaller"?

1.1 Scale Limitations of Conventional FTIR

Conventional FTIR sample spot diameter is typically in the range of 2–10 mm: ATR crystal contact area ~1–2 mm, KBr pellet diameter ~13 mm, liquid cell spot ~5 mm. This scale is sufficient for the "macro chemical fingerprint" of homogeneous samples, but it is inadequate for the following scenarios [2][3]:

• Small particles: microplastics in environmental samples (<500 μm), isolated contaminations on semiconductor wafers (<50 μm), single fibers or paint chips in forensic science;
• Heterogeneous sections: cross-sections of oil paintings (each pigment layer thickness 5–50 μm), biological tissue sections (cell scale 10–100 μm), phase-separated structures of polymer blends (1–50 μm);
• Defect localization: pinholes in thin films, debonding zones at composite interfaces, flux residues on printed circuit board solder joints.

In these scenarios, an averaged macroscopic spectrum is not only meaningless but can also "dilute" critical information—a 50 μm contaminant mixed within a 5 mm spot contributes less than one ten-thousandth to the final spectrum and is almost invisible [2][3].

1.2 The Solution of μ-FTIR

Micro-FTIR (Micro-FTIR / μ-FTIR) = Infrared microscope + FTIR spectrometer:

        ┌────────────────────────────────────────────────┐
        │  FTIR Spectrometer (Source + Interferometer + Electronics) │
        │     │                                          │
        │     └─── IR Interferometer Beam → Microscope Optics │
        │                ↓                               │
        │  ┌──────────────────────────────────────┐     │
        │  │ Infrared Microscope                   │     │
        │  │  · Visible Light Path (aiming + viewing sample) │
        │  │  · IR Light Path (Transmission / Reflection / ATR) │
        │  │  · Cassegrain Reflective Objectives (15×/32×/74×) │
        │  │  · Motorized XY Stage (step 0.1 μm)       │
        │  │  · Detectors: MCT Single Point / FPA Array │
        │  └──────────────────────────────────────┘     │
        │                ↓                               │
        │     Sample → Spectrum → Chemical Imaging       │
        └────────────────────────────────────────────────┘

The key is: the infrared light is focused onto a tiny area of the sample (typically 5–100 μm), sharing the same optical path with the visible light microscope for observation and positioning; by point-by-point scanning with a motorized stage or using an FPA array for single-shot imaging, a complete infrared spectrum can be obtained for each pixel, thus constructing a "chemical map" [2][3][4].

The first commercial micro-FTIR appeared in the 1980s (Bio-Rad Digilab UMA-150), and since then μ-FTIR has rapidly spread in forensic science, heritage, semiconductors, polymers, biomedicine, and other fields [2].

2. Three Measurement Modes of μ-FTIR

Micro-FTIR is divided into three modes based on the mode of interaction between sample and light: Transmission, Reflection, and ATR (Attenuated Total Reflection). Each mode is suitable for different sample forms [2][3][4].

2.1 Transmission μ-FTIR

   IR Incident Light          Sample (thin film/micro-particle + KBr)
      │                          │
      ▼                          ▼
   ┌──────┐    ┌───────────┐    ┌──────┐
   │Condenser│ → │  Sample   │ → │Objective│ → Detector
   └──────┘    └───────────┘    └──────┘
              (Must be thin and IR-transparent)

• Principle: Infrared light passes through the sample, is absorbed, and reaches the detector; identical to conventional transmission FTIR;
• Sample Requirements: The sample must be thin enough for IR to penetrate—typically 5–20 μm (organic), and requires an IR-transparent substrate (KBr, BaF₂, CaF₂ windows or polished sections);
• Advantages: High signal-to-noise ratio, "textbook-quality" spectra, easy library searching, reliable quantitative analysis;
• Limitations: Sample preparation is cumbersome (requires sectioning or pressing), opaque samples cannot be measured, substrate selection is limited by IR transparency;
• Typical Applications: Cross-sections of polymer films (thickness 10 μm), microplastic particles (pressed into KBr and then pelletized), frozen biological tissue sections (thickness 8 μm, placed on BaF₂) [3][4].

💡 Industry Experience: Transmission μ-FTIR spectra are closer to standard libraries, and library search hit rates are usually higher. ISO 4484-2 (Infrared identification of microplastics from textile sources) includes infrared spectroscopy as one of the identification pathways; whether it is adopted as the "arbitration/preferred" method in a specific laboratory depends on the current standard text and method validation results [5].

2.2 Reflection μ-FTIR

   IR Incident Light
      │
      ▼
   ┌──────────┐
   │ Objective │
   └────┬─────┘
        │ ↘ Incident Light

│ ┌───────────────┐
│ │ Sample Surface (Reflection) │
│ └───────────────┘
│ ↗ Reflected light
┌────┴─────┐
│ Objective │ → Detector
└──────────┘

- **Principle**: Infrared light is focused by the objective onto the sample surface, and the reflected light is collected by the same objective and returned to the detector.
- **Classification**:
  - **Specular Reflection**: Direct reflection from smooth surfaces, spectrum contains "dispersive-type" distortion, requires Kramers-Kronig transformation to recover absorption spectrum.
  - **Diffuse Reflection**: Scattered return from rough surfaces, similar to DRIFTS.
  - **Reflection-Absorption (R-A)**: Thin film on metal substrate; light passes through the film twice (double pass), sensitivity enhanced by a factor of 2, commonly used for organic contamination films on metal surfaces [2][4].
- **Advantages**: **No sample preparation, non-destructive**, suitable for bulk, opaque samples, and thin films on metal surfaces.
- **Limitations**: Spectral shape is less standard than transmission (specular reflection has distortion), quantification difficult, sensitive to surface flatness.
- **Typical Applications**: Cultural heritage pigment layers (non-destructive analysis), semiconductor wafer surface contamination, metal coatings, single fibers [4].

> 🔗 **Further Reading**: Reflection-absorption spectra of carbonyl contaminants on metal surfaces typically show absorption at ~1700 cm⁻¹, see [ftir.fun Carbonyl Group Page](https://ftir.fun/ir/group/carbonyl); reflection spectra of nitrile fibers (e.g., acrylic) show C≡N stretching at ~2240 cm⁻¹, see [ftir.fun Nitrile Group Page](https://ftir.fun/ir/group/nitrile).

### 2.3 ATR μ-FTIR

IR Incident Light
│
▼
┌──────────────┐
│ ATR Objective │ ← Built-in Ge or Diamond Crystal Cone
└──────┬───────┘
│
▼
╔════════╗ ← ATR Crystal (Ge refractive index 4.0)
║────────║ ← Evanescent wave penetration ~0.5–2 μm
╚════════╝
│
▼
┌────────┐
│ Sample │ ← In close contact with crystal
└────────┘

- **Principle**: An ATR crystal (Ge or diamond) is mounted at the front of the microscope objective; the crystal is pressed into contact with the sample, using evanescent waves to probe the chemical information of the sample surface at ~0.5–2 μm depth (see Ep 14).
- **Advantages**:
  - **No sample preparation**: Direct measurement of bulk, uneven surfaces, and hydrated samples.
  - **Enhanced spatial resolution**: High refractive index crystal (Ge n=4.0) increases the effective numerical aperture by ~4×, **theoretical resolution reaches ~3 μm @ 1000 cm⁻¹**, breaking the conventional diffraction limit.
  - **High sample tolerance**: Can measure black, opaque, and hydrated samples [3][6].
- **Limitations**:
  - Contact measurement, may damage soft samples (e.g., cells, films).
  - Penetration depth varies with wavelength (see Ep 14), requires ATR correction.
  - Crystal is easily scratched (Ge is relatively soft), requires careful maintenance.
- **Typical Applications**: Non-destructive analysis of cultural heritage pigment layers, imaging of polymer blend phase separation, hydrated tissue sections, single fibers [6].

> 💡 **Industry Experience**: ATR mode is the "daily workhorse" of μ-FTIR—Bruker Hyperion 3000, Thermo Nicolet iN50, and PerkinElmer Spotlight 400 all come standard with ATR objectives. Ge crystals offer clear resolution advantages in the high wavenumber region (>2000 cm⁻¹), while diamond crystals are more durable but have slightly lower resolution [3][6].

### 2.4 Comparison of Three Modes

| Dimension | Transmission | Reflection | ATR |
|-----------|--------------|------------|-----|
| Sample Preparation | Sectioning/pelleting cumbersome | No preparation | No preparation |
| Sample Requirements | Thin + IR-transparent substrate | Smooth surface/reflective | Contact with crystal |
| Spatial Resolution (@1000 cm⁻¹) | ~10 μm | ~10 μm | **~3 μm (Ge crystal)** |
| Spectral Quality | Textbook-level | Distorted, requires correction | Requires ATR correction |
| Library Search Hit Rate | Highest | Medium | Medium-High |
| Non-destructiveness | Destructive (sectioning) | Non-destructive | Minimally destructive (contact pressure) |
| Typical Applications | Microplastics, tissue sections | Cultural heritage, semiconductors, metal films | Cultural heritage, polymers, hydrated samples |

> Table 1: Comparison of three μ-FTIR modes (Data sources: Katzenberg *Top Curr Chem* 2015 [4]; Liu & Kazarian *Analyst* 2022 [6])

---

## 3. Spatial Resolution: The "Ceiling" of the Diffraction Limit

### 3.1 Physical Origin of the Diffraction Limit

The spatial resolution of μ-FTIR is fundamentally limited by **light diffraction**. According to the Abbe criterion, the resolution limit of an optical system is approximately [2][3][7]:

$$ d \approx 0.61 \cdot \frac{\lambda}{NA} $$

Where:
- d: minimum resolvable feature size;
- λ: wavelength of light;
- NA: numerical aperture of the objective (typically 0.4–0.7).

**Key Issue**: Infrared wavelengths are inherently long!

| Wavenumber (cm⁻¹) | Wavelength (μm) | Abbe Limit (NA=0.6, unit μm) |
|-------------------|-----------------|------------------------------|
| 3000 | 3.3 | 3.4 |
| 1700 | 5.9 | 6.0 |
| 1000 | 10.0 | 10.2 |
| 500 | 20.0 | 20.3 |

> Table 2: Diffraction limits at different wavenumbers (Data source: Lasch & Naumann *Anal Chem* 2006 [7])

This is what the μ-FTIR industry often refers to as "**~10 μm @ 1000 cm⁻¹**"—at the center of the mid-infrared fingerprint region, **spatial resolution is limited by diffraction to about 10 μm**. This is far larger than the ~0.25 μm of visible light microscopy, because infrared wavelengths are 10–40 times longer than visible light [2][3][7].

### 3.2 Trade-offs Between High NA Objectives and Small Apertures

To approach the diffraction limit, μ-FTIR uses [3][7]:

1. **High NA Reflective Objectives (Cassegrain)**: NA 0.5–0.7, all-reflective design to avoid chromatic aberration;
2. **Adjustable Aperture**: Placed at the conjugate plane of the sample to define the sampling area. The smaller the aperture, the higher the spatial resolution, but **the less energy passes through, and the lower the signal-to-noise ratio (SNR)**.

This is the core trade-off in μ-FTIR:

Small aperture → Resolution↑ → Energy↓ → SNR↓ → Scan time↑
Large aperture → Resolution↓ → Energy↑ → SNR↑ → Scan time↓
```

In practice, with a conventional blackbody source and a 10 μm aperture, a single point measurement at 4 cm⁻¹ resolution and 128 scans takes about 30–60 seconds; to achieve 5 μm resolution, the scan time may increase to 5–10 minutes [3][7].

3.3 How Does ATR Mode Break Through?

ATR mode uses a high refractive index crystal (Ge n=4.0) to focus light at a larger angle within the crystal, effectively "compressing the wavelength"—according to λ_med = λ_0/n, the equivalent resolution limit becomes [3][6]:

$$ d_{ATR} \approx 0.61 \cdot \frac{\lambda}{n \cdot NA} = \frac{d_{trans}}{n} $$

For a Ge crystal (n=4.0), the theoretical resolution is improved by a factor of 4: ~2.5 μm @ 1000 cm⁻¹. This effectively shifts the diffraction limit downward in a high refractive index medium (often colloquially called "super-resolution," but it is actually an improvement of the diffraction limit in the medium, different from techniques like O-PTIR that truly break the infrared wavelength limit)—which is why sub-10 μm microplastic imaging increasingly relies on ATR microscopy mode [6].

🔗 Further Reading: Characteristic absorptions of common polymers in microplastics—CH₂ stretching of polyethylene (PE) at 2920/2850 cm⁻¹, see ftir.fun alkyl C-H group page; ester C=O of polyester (PET) at ~1715 cm⁻¹, see ftir.fun ester group page; amide I/II of polyamide (PA) at 1640/1540 cm⁻¹, see ftir.fun amide group page.

4. Detectors: MCT Single Point vs FPA Array

4.1 MCT Single Point Detector

MCT (Mercury Cadmium Telluride) is the traditional workhorse detector for μ-FTIR [2][3]:

• Material: Hg₁₋ₓCdₓTe, x determines bandgap → determines response wavelength range;
• Types:
  • MCT-A (narrowband): covers 4000–700 cm⁻¹, highest sensitivity, requires liquid nitrogen cooling;
  • MCT-B (mid-band): covers 4000–400 cm⁻¹, slightly lower sensitivity;
  • MCT-High D (high sensitivity): D > 5×10¹⁰ cm·Hz^0.5/W;
• Advantages: High sensitivity, fast response (MHz level), large dynamic range;
• Limitations: Requires liquid nitrogen cooling (77 K), single-point measurement requires point-by-point scanning for imaging.

4.2 FPA Focal Plane Array Detector

FPA (Focal Plane Array) is an important advancement in μ-FTIR imaging, to be detailed in Ep 37, here only a brief introduction [3][8]:

• Structure: Two-dimensional array detector (e.g., 64×64 = 4096 pixels, or 128×128 = 16384 pixels);
• Principle: Each pixel is equivalent to an independent MCT detector, acquiring 4096–16384 spectra in one exposure;
• Speed comparison: Taking a 100 μm × 100 μm area with 5 μm step as example:
  • Single-point MCT: 20×20 = 400 points × 30 s/point = 200 minutes;
  • 64×64 FPA: 1 exposure = ~5 minutes (40× speedup);
• Applications: Full filter membrane imaging of microplastics, large-area imaging of tissue sections, distribution of polymer blends [8].

💡 Selection Tips: Single-point MCT is suitable for small-area high-resolution (< 1 mm², requiring fine chemical localization); FPA is suitable for large-area imaging (> 1 mm², requiring chemical maps). Many research-grade instruments are equipped with both, switchable as needed [3][8].

5. Synchrotron Radiation Source: Breaking the Brightness Bottleneck

Conventional blackbody sources (Globar, silicon carbide) have limited brightness, resulting in severe energy deficiency and sharp decline in signal-to-noise ratio with apertures < 10 μm. Synchrotron infrared light source is a "performance amplifier" for μ-FTIR, to be discussed in depth in Ep 38, here only key points [3][9][10]:

• Brightness increase by 100–1000 times: Synchrotron radiation is emitted by relativistic electrons in bending magnets, brightness 2–3 orders of magnitude higher than Globar;
• Signal-to-noise ratio improvement by 10–100 times at diffraction limit: Maintains high-quality spectra even with 3–10 μm apertures;
• Spatial resolution approaching diffraction limit: With high NA objectives, enabling chemical imaging of ~3 μm @ 1000 cm⁻¹;
• Representative beamlines worldwide: SMIS at SOLEIL (France), Beamline 1.4.4 at ALS (USA), BL01B at SSRF (Shanghai), IRMicroscopy at Australian Synchrotron, etc. [9][10].

A representative breakthrough of synchrotron μ-FTIR is single-cell imaging—chemical imaging of individual red blood cells, cancer cells, and plant cells, where the signal-to-noise ratio of each pixel still supports second derivative analysis of Amide I to distinguish protein secondary structures [9][10].

6. Typical Application Cases

6.1 Case 1: Layered Structure Analysis – Cross-Section of Oil Painting

An oil painting typically consists of support (canvas/wood panel), ground layer (gesso, calcium sulfate + animal glue), paint layers (1–5 layers, each 5–30 μm), and varnish (natural resin). Cross-sectional μ-FTIR imaging is a key technique for revealing painting techniques, restoration history, and degradation mechanisms [6][11].

A typical case from a 2022 review by Liu and Kazarian in Analyst [6]:

• Sample: Cross-section slice (5 μm thick, on BaF₂) of a 16th-century Venetian oil painting;
• Instrument: Synchrotron μ-FTIR (ALS Beamline 1.4.4);
• Aperture: 8 μm × 8 μm;
• Scan: 64×64 points, 128 scans per point;
• Chemical imaging results:
  • Outermost layer: natural resin varnish, characteristic peak at 1735 cm⁻¹ (ester C=O);
  • Middle paint layer: lead white + linseed oil, characteristic peak at 1520 cm⁻¹ (carboxylate COO⁻, lead soap degradation product);
  • Ground layer: calcium sulfate gesso, characteristic peaks at 1110 cm⁻¹ (SO₄²⁻) and 1620 cm⁻¹ (crystal water).

🔗 Further Reading: The characteristic absorption of carboxylates (metal soap degradation products) commonly found in oil paintings is at ~1520/1420 cm⁻¹, different from carboxylic acid (COOH) at 1700–1730 cm⁻¹, see ftir.fun carboxyl group page; SO₄²⁻ of calcium sulfate at ~1110 cm⁻¹, see ftir.fun sulfate group page.

6.2 Case 2: Defect Localization – Organic Contamination on Semiconductor Wafer Surface

Semiconductor manufacturing demands extremely high wafer surface cleanliness—a single organic contamination particle tens of micrometers in size can cause failure. Reflection-mode μ-FTIR is commonly used for contamination localization and identification (complementary to surface mass spectrometry techniques like ToF-SIMS, not a replacement) [2][3]:

• Sample: Suspected organic contamination particles on Si wafer (visible under optical microscope, ~30–80 μm);
• Method: Reflection-mode μ-FTIR, aperture ~20 μm, multiple scans co-added;
• Example result: Particle spectra may show 2925/2850 (CH₂), ~1735 (ester C=O), etc.; library search may point to plasticizer candidates; for thinner sub-nanometer films, GIR-FTIR may be required (see Ep 34).

⚠️ Earlier draft incorrectly referenced this case to the ToF-SIMS review by Belu et al. [12]; ToF-SIMS is suitable for elemental/fragment surface chemistry, cannot directly serve as μ-FTIR evidence for this case. Now changed to a methodological example.

This case illustrates the role of μ-FTIR in defect root cause analysis—first identifying "what it is", then tracing "where it came from".

6.3 Case 3: Microplastic Imaging – FPA Full Filter Membrane Automatic Analysis

Microplastics (MPs) have become a global environmental hotspot. μ-FTIR is a common confirmatory tool for microplastic identification, and FPA arrays further increase throughput to handle entire filter membranes (see Ep 26, Ep 37 for details) [1][5][8]:

Typical workflow (using common micro-FTIR + 64×64 FPA as example):

1. Sampling: Filter water sample through glass fiber filter (GF/F, 0.7 μm pore size);
2. Digestion: Remove biological organic matter using Fenton reagent (to avoid interference);
3. Drying: Oven-dry at 60°C for 2 h;
4. Imaging: FPA automatic stitching of full filter membrane;
5. Data processing: One spectrum per pixel, library search to match polymer type;
6. Output: Particle count, size distribution, polymer type distribution, chemical imaging map.

Note: Uncompressed data volume for a full filter membrane can reach hundreds of GB (depending on frames and float storage); if "about 470 GB" appears in text, it should be understood as an upper estimate; in practice, pre-screening and compression are often used.

Primpke et al. reported that FPA full scanning significantly speeds up compared to single-point MCT [8].

🔗 Further Reading: Polymer types commonly detected in microplastics and corresponding ftir.fun functional group pages:

• Polyethylene PE / Polypropylene PP: alkyl C-H

• Polystyrene PS: aromatic ring + alkyl C-H

• Polyester PET: ester group
• Polyamide PA (Nylon): amide
• Poly(methyl methacrylate) PMMA: methacrylate group
• Polyurethane PU: urethane group

6.4 Case 4: Forensic Science—Single Fiber Identification

A 10-μm-thick fiber could be key to solving a case. μ-FTIR can identify fiber types without destroying the sample [13]:

• Cotton (cellulose): 1050/1030 cm⁻¹ (C-O stretch), 3300 cm⁻¹ (OH);
• Polyester PET: 1715 cm⁻¹ (ester C=O), 1240 cm⁻¹ (C-O);
• Nylon PA: 1640/1540 cm⁻¹ (Amide I/II);
• Acrylic PAN: 2240 cm⁻¹ (C≡N stretch, highly characteristic);
• Cellulose acetate: 1740 cm⁻¹ (acetate C=O).

🔗 Further reading: The C≡N stretch of acrylic fibers at 2240 cm⁻¹ is a rare "isolated strong peak" in IR spectra, almost without overlap from other absorptions, making it the "gold standard" for acrylic fiber identification. See ftir.fun nitrile group page.

7. Practical Tips and Pitfall Avoidance for μ-FTIR

7.1 Sample Preparation Tips

Table 3: μ-FTIR sample preparation recommendations for different sample types (Data source: Liu & Kazarian Analyst 2022 [6]; Primpke Appl Spectrosc 2020 [8])

7.2 Common Issues and Troubleshooting

1. Weak signal, poor SNR:

   • Check optical alignment (Cassegrain objectives need periodic centering);
   • Increase number of scans (128→256→512);
   • Check if aperture is too small (energy severely insufficient when < 5 μm);
   • Switch to synchrotron radiation source (if available) [3][9].

2. Severe baseline tilt:

   • Transmission mode: uneven sample thickness → re-section;
   • Reflection mode: uneven surface → polish or switch to ATR;
   • ATR mode: poor crystal contact → re-press the sample [6].

3. Water vapor interference:

   • Sharp gas peaks appear at 1500–1800 cm⁻¹ and 3400–4000 cm⁻¹;
   • Solution: Continuously purge instrument with dry air or N₂; measure sample immediately after collecting background spectrum; use atmospheric correction software (e.g., VaporFit) for post-processing [14].

4. Insufficient spatial resolution:

   • Check if aperture is too large (should be < 1.5 times the target feature size);
   • Switch to ATR mode (Ge crystal can improve by ~4×);
   • Use higher NA objective (74× vs 32×) [3][7].

5. FPA data explosion:

   • Single frame 64×64 = 4096 spectra ≈ 100 MB (4 cm⁻¹, 4000–900 cm⁻¹);
   • One filter membrane ~19000 frames ≈ 470 GB;
   • Solution: Pre-screening (first identify candidate particles from optical images), process in blocks, use open-source tools like OpenSpecy [15].

7.3 Maintenance Tips

• MCT detector liquid nitrogen: Refill liquid nitrogen daily before work to avoid empty Dewar damaging the detector;
• Cassegrain objectives: Clean regularly (cotton swab with 1:1 alcohol/ether), avoid scratching mirror surfaces;
• ATR crystal: Ge crystal has low hardness (Mohs 2.5), wipe gently with lint-free cloth and alcohol, avoid hard objects;
• Motorized stage: Calibrate position accuracy periodically to avoid cumulative errors;
• Dry purge: Maintain continuous low-flow N₂ purge when instrument is idle to protect beamsplitter and optics [3].

8. Limitations and Future of μ-FTIR

8.1 Current Limitations

1. The "hard wall" of diffraction limit: Conventional μ-FTIR is limited to ~10 μm in the mid-IR fingerprint region, unable to resolve sub-micrometer structures [2][3];
2. Insufficient brightness of conventional sources: SNR drops sharply with small apertures, forcing increased scan numbers;
3. Cumbersome sample preparation: Transmission mode requires sectioning, challenging for soft samples (cells, films);
4. Water interference: Aqueous samples require special handling;
5. Data explosion: FPA wide-area imaging generates GB–TB data, making processing and storage an engineering challenge [8][15].

8.2 Breakthrough Directions

To break the diffraction limit, various "super-resolution" infrared techniques have been developed in recent years (detailed in subsequent episodes):

• Synchrotron μ-FTIR (Ep 38): By increasing source brightness, pushes SNR at the diffraction limit to usable levels, still constrained by the diffraction limit;
• O-PTIR (Optical Photothermal Infrared) (Ep 39): Uses visible light to detect temperature rise from IR absorption, resolution ~450 nm, breaks the IR diffraction limit [16];
• AFM-IR (Atomic Force Microscopy-Infrared): Uses AFM tip to detect photothermal expansion, resolution ~10 nm, but belongs to a different technical route;
• nano-FTIR: Nanoscale FTIR based on s-SNOM, resolution ~20 nm.

💡 Trend hint: In the next 5–10 years, O-PTIR and AFM-IR will gradually erode μ-FTIR's share in submicron chemical imaging, but μ-FTIR's "mainstay position" at the 10–100 μm scale is hard to replace—it remains the most cost-effective tool for micro-area chemical imaging [3][16].

Episode Summary

| Breakthrough Direction | O-PTIR (submicron), AFM-IR (nanoscale), synchrotron (improved SNR) |

Review Questions

1. You need to analyze a microplastic particle with a diameter of about 8 μm. Under a conventional blackbody source, what is the diffraction limit of transmission-mode μ-FTIR at 1000 cm⁻¹? Can this particle be effectively resolved? Which measurement mode would you choose and why?

2. Synchrotron infrared sources are 100–1000 times brighter than a conventional Globar, but their spatial resolution cannot break the diffraction limit. Explain this seemingly contradictory phenomenon: what does the brightness improvement bring? Why can't it surpass the diffraction limit?

3. In the analysis of cross-sections of oil paintings, specular reflection spectra from reflection-mode μ-FTIR often exhibit "dispersive distortion". Briefly describe the cause of this distortion and the role of the Kramers-Kronig transformation.

4. FPA focal plane array detectors can acquire 4096–16384 spectra in one exposure, but they are not always superior to single-element MCT detectors in all scenarios. Provide at least two scenarios where FPA is inferior to single-element MCT, with reasons.

5. Design a μ-FTIR experimental plan to identify the chemical composition of several dozen 20–50 μm particles dispersed on a 1 cm × 1 cm film. Describe sample preparation, measurement mode, aperture settings, scanning parameters, and data processing workflow.

References

[1] Cole M, Lindeque P, Halsband C, Galloway T S. "Microplastics as Contaminants in the Marine Environment: A Review." Marine Pollution Bulletin, 2011, 62(12): 2588–2597. DOI:10.1016/j.marpolbul.2011.09.025

[2] Griffiths P R, de Haseth J A. Fourier Transform Infrared Spectrometry. 2nd ed. Wiley, 2007. ISBN: 978-0-471-19404-0.

[3] Nasse M J, Walsh M J, Mattson E C, et al. "High-Resolution Fourier-Transform Infrared Chemical Imaging with Multiple Synchrotron Beams." Nature Methods, 2011, 8(5): 413–416. DOI:10.1038/nmeth.1580

[4] Katzenberg F. "Micro-FTIR Spectroscopy of Polymers and Composites." Topics in Current Chemistry, 2015, 357: 1–28. DOI:10.1007/978-3-662-45237-6_1

[5] ISO 4484-2:2023. Textiles and Textile Products — Microplastics from Textile Sources — Part 2: Identification by Infrared Spectroscopy.
https://www.iso.org/standard/…

[6] Liu G L, Kazarian S G. "Recent Advances and Applications to Cultural Heritage Using ATR-FTIR Spectroscopy and ATR-FTIR Spectroscopic Imaging." Analyst, 2022, 147: 1777–1797. DOI:10.1039/D2AN00005A

[7] Lasch P, Naumann D. "Spatial Resolution in Infrared Microspectroscopic Imaging of Tissues." Biochimica et Biophysica Acta — Biomembranes, 2006, 1758(7): 814–829. DOI:10.1016/j.bbamem.2006.06.008

[8] Primpke S, Fischer M, Lorenz C, et al. "Comparison of Imaging and FTIR Microscopy Applied for the Analysis of Microplastics in Environmental Samples." Applied Spectroscopy, 2020, 74(9): 1185–1197. DOI:10.1177/0003702820922969

[9] Miller L M, Dumas P. "From Structure to Cellular Mechanism with Infrared Microspectroscopy." Current Opinion in Structural Biology, 2010, 20(5): 649–656. DOI:10.1016/j.sbi.2010.07.007

[10] Tobin M J, Puskar L, Barbosa R L, et al. "FTIR Microspectroscopy of Single Living Cells in Aqueous Environment by Synchrotron IR Microscopy Using Microfabricated Fluidic Devices." Optics Express, 2010, 18(11): 11061–11070. DOI:10.1364/OE.18.011061

[11] Mazzeo R, Joseph E, Prati S, et al. "Attenuated Total Reflectance—FTIR Imaging for the Analysis of Pigments in Cross-Sections of Paint Layers." Analytical and Bioanalytical Chemistry, 2007, 387(8): 2833–2842. DOI:10.1007/s00216-006-1053-5

[12] Belu A M, Graham D J, Castner D G. "Time-of-Flight Secondary Ion Mass Spectrometry: Techniques and Applications for the Characterization of Biomaterial Surfaces." Biomaterials, 2003, 24(21): 3635–3651. DOI:10.1016/S0142-9612(03)00159-5

[13] Tien C, Lai Y, Lin C. "Forensic Identification of Fibers by FTIR Microscopy." Forensic Science Journal, 2014, 13(1): 19–34.

[14] VaporFit: Phys. Chem. Chem. Phys. 2025, 27. DOI:10.1039/D5CP01007A. GitHub: https://github.com/piobruzd/V…

[15] OpenSpecy: Cowger W et al. Analytical Chemistry 2021, 93(32): 11184–11193. DOI:10.1021/acs.analchem.1c01349. GitHub: https://github.com/wincowgerD…

[16] Marchetti A et al. "Novel Optical Photothermal Infrared (O-PTIR) Spectroscopy for the Noninvasive Characterization of Heritage Glass–Metal Objects." Science Advances, 2022, 8(9): eabl6769. DOI:10.1126/sciadv.abl6769

[17] ftir.fun Alkyl C-H Functional Group Page. https://ftir.fun/ir/group/alk…

[18] ftir.fun Ester Functional Group Page. https://ftir.fun/ir/group/est…

[19] ftir.fun Amide Functional Group Page. https://ftir.fun/ir/group/ami…

[20] ftir.fun Carbonyl Functional Group Page. https://ftir.fun/ir/group/car…

[21] ftir.fun Nitrile Functional Group Page. https://ftir.fun/ir/group/nit…

[22] ftir.fun Carboxyl Functional Group Page. https://ftir.fun/ir/group/car…

[23] ftir.fun Methacrylate Functional Group Page. https://ftir.fun/ir/group/met…

[24] ftir.fun Aromatic Ring Functional Group Page. https://ftir.fun/ir/group/aro…

[25] ftir.fun Sulfate Functional Group Page. https://ftir.fun/ir/group/sul…

[26] ftir.fun Urethane Functional Group Page. https://ftir.fun/ir/group/ure…

Next Episode Preview: Ep 37 — FPA Focal Plane Array Imaging: High-Throughput Chemical Imaging
In this episode, we learned about the three modes of μ-FTIR, diffraction limits, and detector selection. The next episode will focus on FPA focal plane arrays—the technology that makes "a spectrum per pixel" a reality. We will dive into the principles of FPA detectors, data cube structure, speed comparison with single-point MCT, and demonstrate how FPA turns chemical imaging from a "luxury" into an "everyday tool" through two cases: automated analysis of microplastics on full filter membranes and tissue section imaging. We will also discuss the big data challenges brought by FPA and chemometric solutions.

This article is licensed under CC BY-NC-SA 4.0. All images are from public domain or web resources with attributed sources; copyrights belong to the original owners.