NIR Cuvettes with JGS3 Quartz: Near-Infrared Spectroscopy 700–2500 nm Guide
NIR Cuvettes with JGS3 Quartz: Near-Infrared Spectroscopy 700–2500 nm Guide
On this page
Material Guide · NIR · JGS3 Quartz
NIR Cuvettes with JGS3 Quartz
Near-infrared spectroscopy needs cuvettes that transmit cleanly through the 1300, 1900, and 2200 nm water-overtone bands. JGS3 quartz does — JGS1 does not. Here is how to pick the right grade for 700–2500 nm work.
700 – 2500 nm
Low-OH JGS3
FT-NIR · process · food · pharma
An NIR cuvette is a quartz sample cell designed for the 700–2500 nm near-infrared region, where the dominant analytical features are water and C-H/N-H/O-H overtone absorption bands at roughly 970, 1190, 1380, 1450, 1900, and 2200 nm. Standard UV-Vis cuvettes made from JGS1 quartz contain residual hydroxyl (OH) groups that themselves absorb in the 1300–2200 nm region — overlapping the analyte bands and biasing every NIR measurement. JGS3 quartz is the low-OH grade specifically processed to transmit cleanly through that window. For any NIR measurement above ~1100 nm, the cuvette must be JGS3 or a comparable low-OH fused silica; mixing a JGS1 reference with a JGS3 sample cell produces strong artifacts in the same bands that carry the analytical information.
TL;DR · If your method scans above 1100 nm, specify JGS3 quartz. Standard JGS1 cuvettes have an OH absorption band centered around 2200 nm that climbs steeply from 1400 nm — invisible in the UV-Vis but dominant in the NIR. For dilute aqueous samples in NIR, use 5–10 mm path; for moisture analysis in solids/films, use 1–2 mm path or transmission accessories; for FT-NIR process work, use matched pairs from a single fabrication batch (Sintered 80/83 minimum).
1. What NIR cuvettes are for — and why the cell matters more than in UV-Vis
💡 Takeaway — In UV-Vis, the cuvette is a transparent container. In NIR, the cuvette itself absorbs at the same wavelengths your analyte does — so the cuvette is part of the chemistry, not just packaging.
Near-infrared spectroscopy covers the 700–2500 nm range (sometimes extended to 3500 nm with specialty cells). The information it carries is fundamentally different from UV-Vis: instead of electronic transitions of chromophores, NIR detects vibrational overtones and combinations of C–H, N–H, O–H, and S–H bonds. Those overtones cluster in narrow bands at 970, 1190, 1380, 1450, 1900, and 2200 nm — and one of them (around 1900–2200 nm) is exactly where standard quartz has its own OH-related absorption.
Three consequences for cell selection:
- Material matters more. In UV-Vis the cuvette is essentially transparent across the analytical window; the optical path geometry is what controls accuracy. In NIR the material itself absorbs in the analytical window, so the choice between JGS1 and JGS3 quartz directly changes what your spectrum looks like.
- Path length is shorter on average. Water has strong fundamental and overtone absorption in NIR. Aqueous samples saturate the detector at 10 mm path lengths in many regions. Typical NIR cuvettes use 1, 2, or 5 mm paths.
- Matched-pair quality is critical. NIR signals are inherently small (overtones are weaker than fundamentals by ~100×). A 0.005 A baseline drift that is invisible in UV-Vis quantification can become a sign-flipping artifact in NIR.
2. JGS1 vs JGS3 — the OH-band problem
💡 Takeaway — JGS1 is the deep-UV grade (185 nm cutoff, retains ~1000 ppm OH). JGS3 is the IR-optimized grade (260 nm cutoff, <10 ppm OH). For any NIR work the difference between them is the difference between a clean spectrum and one corrupted by a 1.5 µm hydroxyl band.
The Chinese GB/T 6071 standard defines four fused-silica grades; for cuvettes only JGS1 and JGS3 are commonly stocked (JGS2 exists as a designation but is not part of MachinedQuartz’s product line). The numerical difference between them is the hydroxyl (OH) content of the fused silica — and that single difference creates two very different transmission profiles in the NIR.
| Property | JGS1 (synthetic, deep-UV) | JGS3 (low-OH, NIR-optimized) |
|---|---|---|
| UV cutoff | 185 nm | 260 nm |
| Useful range | 185 – 2200 nm | 260 – 3500 nm |
| OH content | ~ 1,000 ppm | < 10 ppm |
| OH band at 2200 nm | Strong absorption | Negligible |
| OH band at 1380 nm | Visible dip | Negligible |
| Visible appearance | Indistinguishable from JGS3 by eye | Indistinguishable from JGS1 by eye |
| Best for | UV-Vis 200–800 nm, deep-UV work | NIR 1100–2500 nm, FT-NIR, IR transmission |
Because the two grades look identical visually, the etched part number on the cuvette is the only field-reliable identifier. Mixing a JGS1 reference cell with a JGS3 sample cell during NIR measurement is the single most common artifact source in NIR cuvette work — it produces a strong negative absorbance band centered around 1380 nm and a smaller one near 2200 nm, both of which fall on the same wavelengths labs typically measure water and aromatic C–H combinations.
3. JGS3 transmission curve — what you can and cannot measure
💡 Takeaway — JGS3 transmits cleanly through 260–2500 nm and most of 2500–3500 nm. The two regions to be careful of are 1900 nm (water overtone, even in low-OH quartz there is a residual dip) and above 3500 nm (Si–O fundamental absorption — that is the mid-IR boundary where any quartz fails).
4. NIR application matrix — which method needs which cuvette
💡 Takeaway — Every major NIR application maps to a path length and a matched-pair grade. Use the table to specify the cell at the start of method development, not after the first artifact appears.
| NIR application | Wavelength window | Path length | Cuvette grade | Why this combination |
|---|---|---|---|---|
| Moisture in solvents / pharma intermediates | 1380 nm, 1900 nm | 1–2 mm | JGS3, Sintered 80/83 matched pair | Strong water bands — path must be short to keep A in linear range; OH-free quartz so cuvette doesn’t add baseline |
| Protein concentration / structure (overtones) | 1450, 2050, 2180 nm | 2–5 mm | JGS3, Molded 83 sealed | Sealed cell prevents evaporation during long FT-NIR scans |
| Food (oils, fats, sugars) | 1100–2500 nm broad | 5–10 mm | JGS3, Standard 80 (cost-effective) | Routine production analytics; broad sweep doesn’t demand tight matched-pair |
| Process NIR (online monitoring) | 900–1700 nm | 1–5 mm flow cell | JGS3, Molded 83 (sealed, fused) | No cap to leak under process pressure; tight tolerance for long-term calibration stability |
| Agricultural products (grain, feed) | 1000–2500 nm broad | 10 mm | JGS3, Standard 80 | Diffuse reflectance more common than transmission for solids; transmission cells used for slurries |
| Petrochemicals (hydrocarbons) | 1100–1850 nm | 2–10 mm | JGS3, Sintered 80/83 | Need solvent compatibility — Sintered joint-free body resists aromatics |
| Carbon nanotubes / graphene (research) | 800–1600 nm | 2–10 mm | JGS3, Molded 83 | Research-grade tight tolerance for absolute spectra reproducibility |
| Polymer film analysis | 1100–2500 nm | 1 mm or transmission accessory | JGS3 windows + plates | Films typically measured in transmission accessory, not a sealed cuvette — but window grade still needs JGS3 |
5. Compatible NIR instruments — Z-dimensions and stocked formats
💡 Takeaway — NIR instruments use the same 12.5 × 12.5 mm exterior footprint as benchtop UV-Vis. Z-dimensions vary; matched-pair packs need to fit your specific instrument before procurement.
| Instrument family | Type | Z-dim (mm) | Typical cell holder | JGS3 cuvette suits |
|---|---|---|---|---|
| Bruker MPA / MPA II / MATRIX | FT-NIR | 15 | 12.5 mm standard | 1, 2, 5, 10 mm path |
| Thermo Antaris II / Nicolet iS50 (NIR module) | FT-NIR | 15 | Liquid transmission | 2, 5, 10 mm path |
| PerkinElmer Lambda 950 / 1050 / Spectrum 400 | UV-Vis-NIR | 15 | Standard cuvette holder | 10 mm + short-path for moisture |
| Agilent Cary 5000 / 6000i / 7000 | UV-Vis-NIR | 15 | Standard universal holder | 1–100 mm path range |
| Shimadzu UV-3600 / 3700 / SolidSpec | UV-Vis-NIR | 15 | Universal cell holder | 1–50 mm path range |
| Jasco V-770 / V-780 NIR | UV-Vis-NIR | 15 | Standard 12.5 mm | 1–10 mm path |
| FOSS XDS / NIRS DS2500 | Process / dispersive NIR | Custom (often 12) | Specific holder | 1, 2, 5 mm sealed (Molded 83) |
| ABB FTLA2000 / FTIR-NIR | Process FT | Variable | Custom flow cell | Sealed flow cell, JGS3 |
For full Z-dimension data on each instrument and tips on Z-dim adapters when a short cell needs to fit a 15 mm holder, see the cuvette-spectrophotometer compatibility matrix.
6. Path length for NIR — why shorter is usually better
💡 Takeaway — Water has strong NIR overtones that saturate 10 mm path in many regions. Default path for NIR aqueous work is 1 or 2 mm; for transparent non-aqueous samples, 5 or 10 mm; for dilute analytes, longer path is sometimes possible but the water spectrum has to be subtractable.
| Sample / region | Recommended path | Notes |
|---|---|---|
| Aqueous biological / pharma at 1380 nm | 0.5 – 1 mm | Water overtone is intense; longer paths saturate detector |
| Aqueous protein at 2050 nm | 1 – 2 mm | Need to subtract water reference accurately |
| Organic solvent (acetonitrile, methanol) | 5 – 10 mm | Lower absorptivity in NIR than water |
| Oils (vegetable, motor) | 1 – 2 mm transmission | Or use ATR / diffuse reflectance accessory instead |
| Polymer dissolved in chloroform | 2 – 5 mm | Chloroform NIR window is reasonable |
| Dilute NIR-active dye / nanoparticle | 10 mm | Standard path; ensure JGS3 reference |
| Petroleum products (octane, aromatics) | 2 – 5 mm | Hydrocarbon C-H combinations 1100–1800 nm |
For the formal Beer–Lambert calculation that picks a path length from expected concentration and absorptivity, see the path length calculator and the measure-absorbance SOP.
7. Five common NIR cuvette errors
💡 Takeaway — NIR signals are 10–100× weaker than UV-Vis signals at the same concentration. Errors that are tolerable in UV-Vis become fatal in NIR.
- Using JGS1 cuvettes from the UV-Vis bench. The visual identical look means analysts grab “the quartz cuvettes” without checking the etched part number. Result: strong negative bands at 1380 and 2200 nm corrupting the spectrum. Fix — segregate JGS3 cells in a labeled box; check the part number every time.
- Filling with cold sample into a warm cell. NIR is highly sensitive to refractive index changes during thermal equilibration; the baseline drifts for the first 60–120 seconds. Always equilibrate sample and cell to the same temperature before reading; for thermostatted measurements, wait ± 0.5 °C of target.
- Using a 10 mm cell for water-band measurements. Water overtones saturate the detector at 1450 and 1900 nm with 10 mm path. Switch to 1 or 2 mm for any moisture analysis.
- Reference cell empty for NIR. NIR baselines do not always behave like UV-Vis baselines with an empty reference. Match the reference solvent to the sample matrix as closely as possible — for biological samples, use buffer not water.
- Mismatched matched pair. NIR signals are small. A 0.005 A baseline drift that is invisible in UV-Vis is large compared to a typical NIR overtone signal (often 0.05–0.20 A). Specify Molded 83 matched pairs (ΔA ≤ 0.002 A) for quantitative NIR work.
For broader cuvette-side troubleshooting see the negative absorbance guide and the UV-Vis troubleshooting guide.
Why we wrote this guide
NIR cuvette pages in the wider web are surprisingly thin — most either lump NIR cells together with UV-Vis cells or quote a generic “190–2500 nm” range that does not address the OH-band problem. We wrote this guide so the JGS1 versus JGS3 distinction, the overtone-band map, and the application matrix are all in one place and tied to specific MachinedQuartz fabrication grades. Lab buyers should be able to specify a NIR cuvette the same way they specify a quartz window for a Cary 5000 — with confidence in the material and the path.
Authority references
8. Frequently asked questions
Up to about 1100 nm, yes — standard JGS1 quartz transmits well in that region. Above 1100 nm, JGS1 has hydroxyl absorption bands at 1380, 1900, and 2200 nm that overlap with the analytical bands used for water, hydroxyl, and C-H combinations. For any NIR work above 1100 nm, specify JGS3 quartz instead. Visually JGS1 and JGS3 cuvettes look identical; the etched part number is the only field-reliable identifier.
JGS1 is the synthetic deep-UV grade with about 1000 ppm hydroxyl content and a UV cutoff at 185 nm. JGS3 is the low-OH grade with less than 10 ppm hydroxyl and a UV cutoff at 260 nm. The high OH content in JGS1 absorbs at 1380, 1900, and 2200 nm in the NIR region, while JGS3 transmits cleanly through those bands. JGS1 is the right choice for UV-Vis work needing deep-UV transmission; JGS3 is the right choice for NIR work above 1100 nm.
For aqueous samples at the strong water overtone bands (1380, 1900 nm), use 0.5 to 1 mm path length. Standard 10 mm cells saturate the detector at these wavelengths and produce a flat-topped spectrum with no usable information. For weaker overtones at 970 nm or for organic solvents like methanol or acetonitrile, 5 to 10 mm is appropriate. Use the path length calculator for the formal Beer-Lambert sizing.
No. Plastic cuvettes (PMMA, polystyrene) have strong C-H overtone absorption across the entire NIR region. The plastic itself absorbs at the same wavelengths used for analytical detection, producing severe baseline interference. NIR cuvettes must be made from low-OH fused silica (JGS3 or equivalent) for any wavelength above 700 nm.
JGS3 quartz transmits cleanly to about 3000 nm and with reduced transmission to 3500 nm. Above 3500 nm the silicon-oxygen bond fundamental absorption dominates and all quartz materials become opaque. For mid-IR work above 3500 nm, switch to calcium fluoride (CaF2, useful to 8000 nm) or potassium bromide (KBr, useful to 25000 nm) windows.
For visible-range UV-Vis (260 to 800 nm), yes — JGS3 transmits the same as JGS1 in that region. For deep-UV work below 260 nm, JGS3 cannot be used because its cutoff is at 260 nm. Many labs stock a single JGS1 set for UV-Vis and a separate JGS3 set for NIR; mixing the grades during measurement is the single biggest error source in NIR cuvette work, so segregate them clearly.
For quantitative NIR work, yes. NIR overtone signals are 10 to 100 times smaller than UV-Vis signals at the same concentration, so a 0.005 A baseline drift between cells in a pair that is tolerable in UV-Vis becomes a major fraction of the analytical signal in NIR. Specify Molded 83 matched pairs (ΔA ≤ 0.002 A at 280 nm) for FT-NIR and quantitative NIR. For qualitative or process screening work, Sintered 80/83 matched pairs (ΔA ≤ 0.005 A) are sufficient.
Two main possibilities. First, if your cell is JGS1 instead of JGS3, the 1380 nm dip is the hydroxyl absorption band of the quartz itself — switch to JGS3. Second, if the cell is confirmed JGS3, the dip likely comes from residual water absorbed onto the cell walls, especially if the cell is new or has been stored uncapped. Dry the cell in a 60 to 80 °C oven for 30 minutes before use, or run a baseline at the same wavelength to subtract the residual water signature.
Need JGS3 NIR cuvettes for FT-NIR or process work?
MachinedQuartz fabricates JGS3 quartz cuvettes in path lengths from 1 mm to 100 mm, with optional Molded 83 sealed construction for high-temperature or matched-pair-critical work. MOQ 2–4 pieces, lead time 1–4 weeks for custom builds, 5–8 business days for stock 10 mm JGS3.
Recent Comments