Graphene Raman layer checker: monolayer, few-layer or multilayer
Monolayer graphene on Si/SiO₂ has a single-Lorentzian 2D band with a full width at half maximum of 35 cm⁻¹ or less and a 2D peak at least twice as high as the G peak: these are the UK National Physical Laboratory (NPL) criteria. Enter the 2D width, the I(2D)/I(G) ratio and the 2D line shape from your fit, and the checker compares them with the NPL criteria and the published layer trends. It is an indication, not proof.
Last updated
Loading Graphene Raman layer checker…
At a glance
| Monolayer criteria | FWHM(2D) ≤ 35 cm⁻¹ and I(2D)/I(G) ≥ 2 (NPL, Si/SiO₂) |
|---|---|
| Layer trend | 2D stronger than G below five layers (Gupta et al. 2006) |
| Input | 2D FWHM, I(2D)/I(G) or I(G)/I(2D), 2D line shape; optional I(D)/I(G) and I(D)/I(D′) |
| Output | Criteria met or not, and what they suggest; not a layer count |
| Defects | D-band notes; defect type from I(D)/I(D′) (Eckmann et al. 2012) |
| Limits | Doping, strain, defects, substrate, stacking and laser all change the numbers |
| Runs in | Your web browser; nothing is uploaded |
| Price | Free, no account needed |
What it does
How to use the graphene Raman layer checker
- Fit your spectrumFit the G and 2D peaks (and D and D′ if present) with Lorentzians in your Raman software, and note whether one Lorentzian describes the 2D band.
- Set the measurementChoose the laser wavelength and the substrate. The criteria were set for graphene on Si/SiO₂ measured at 532 nm.
- Enter the 2D and G valuesType the 2D FWHM in cm⁻¹ and the peak-height ratio, as I(2D)/I(G) or I(G)/I(2D), and the 2D line shape.
- Add the defect bandsOptionally enter I(D)/I(G) and I(D)/I(D′) to get notes on defects.
- Read the indicationSee which criteria are met and what they suggest, then confirm the layer number with optical contrast, AFM or a Raman map.
How it works
Monolayer (NPL, Si/SiO₂): FWHM(2D) ≤ 35 cm⁻¹ and I(2D)/I(G) ≥ 2
Monolayer criteria. The UK National Physical Laboratory proposed a 2D full width at half maximum of 35 cm⁻¹ or less and an I(2D)/I(G) ratio of at least 2 to identify monolayer graphene on Si/SiO₂, as applied at 532 nm by Muñoz-Vargas et al. (2025). Both numbers are sensitive to other things: I(2D)/I(G) falls with doping and defects, and the 2D width grows with adlayers and with strain variations below the laser spot. A turbostratic (randomly stacked) multilayer can also give a single-Lorentzian 2D narrower than 35 cm⁻¹. Their CVD monolayer averaged 26.2 ± 1.7 cm⁻¹ and 2.4 ± 0.7 over 1024 spectra.
Layer trends. The 2D band changes in shape, width and position with the number of layers (Ferrari et al. 2006). On Si/SiO₂ at 514.5 nm, the 2D band is more intense than G for fewer than five layers (Gupta et al. 2006). At 532 nm, Qiu et al. (2016) measured a single-layer 2D at 2675 cm⁻¹ with a FWHM of about 28 cm⁻¹, and I(G)/I(2D) rising from 0.28 for a single layer to more than 1.4 for multilayers; graphite (HOPG) shows two 2D components near 2680 and 2720 cm⁻¹. The 2D band of AB-stacked bilayer graphene is made of four peaks (Mafra et al. 2012). The checker reports which of these apply to your numbers; it does not claim a layer count.
Defects. The D band needs a defect to appear (Qiu et al. 2016). I(D)/I(G) at a given defect density depends strongly on the laser energy and reaches a maximum at about 3 nm between defects, so one ratio can correspond to two densities; the G width tells them apart (Cançado et al. 2011). The checker therefore does not convert I(D)/I(G) into a defect density. I(D)/I(D′) is about 13 for sp³ defects, 7 for vacancy-like defects and 3.5 for boundaries (Eckmann et al. 2012); the checker names the closest of the three.
Sources: Muñoz-Vargas S et al. Nanoscale Adv 2025;7:7382, doi:10.1039/d5na00668f. Ferrari AC et al. Phys Rev Lett 2006;97:187401, doi:10.1103/PhysRevLett.97.187401. Gupta A et al. Nano Lett 2006;6:2667–2673, doi:10.1021/nl061420a. Qiu C et al. Materials 2016;9:897, doi:10.3390/ma9110897. Mafra DL et al. Nano Lett 2012;12:2883–2887, doi:10.1021/nl300477n. Cançado LG et al. Nano Lett 2011;11:3190–3196, doi:10.1021/nl201432g. Eckmann A et al. Nano Lett 2012;12:3925–3930, doi:10.1021/nl300901a.
Worked examples
A single layer. Qiu et al. (2016) report, at 532 nm, a 2D band at 2675 cm⁻¹ with a FWHM of about 28 cm⁻¹ and I(G)/I(2D) = 0.28 for single-layer graphene. Entered here (I(G)/I(2D) = 0.28, so I(2D)/I(G) = 3.57, one Lorentzian), both NPL criteria are met and the 2D is stronger than G: consistent with monolayer graphene.
A multi-component 2D. With a 2D band that needs several components and I(2D)/I(G) = 1.1, the 2D is still stronger than G but the shape rules out a monolayer: not a monolayer, as for an AB bilayer, whose 2D has four components.
A defect hint. I(D)/I(D′) = 7.4 is closest to the about 7 that Eckmann et al. (2012) found for vacancy-like defects.
Frequently asked questions
How do you tell monolayer graphene from Raman?
Look at the 2D band. For monolayer graphene on Si/SiO₂ it is one Lorentzian, with a full width at half maximum of 35 cm⁻¹ or less, and at least twice as high as the G peak: the UK NPL criteria, FWHM(2D) ≤ 35 cm⁻¹ and I(2D)/I(G) ≥ 2. Confirm with optical contrast, AFM or a map.
What are the NPL criteria for monolayer graphene?
FWHM(2D) ≤ 35 cm⁻¹ and I(2D)/I(G) ≥ 2, proposed by the UK National Physical Laboratory for graphene on Si/SiO₂ and applied at 532 nm by Muñoz-Vargas et al. (2025).
Can multilayer graphene pass the monolayer test?
Yes. A turbostratic (randomly stacked) multilayer can also give a single-Lorentzian 2D band narrower than 35 cm⁻¹, so a spectrum that meets the criteria is consistent with a monolayer, not proof of one.
What does I(2D)/I(G) below 1 mean?
The 2D band is weaker than G. Gupta et al. (2006) found the 2D more intense than G for fewer than five layers on Si/SiO₂ at 514.5 nm, so a weaker 2D suggests five layers or more; strong doping or many defects also lower the ratio.
What does the 2D band of bilayer graphene look like?
Wider than a monolayer’s and made of several components: for AB-stacked bilayer graphene, four peaks (Mafra et al. 2012). Graphite shows two components near 2680 and 2720 cm⁻¹ (Qiu et al. 2016).
Can I get the defect density from I(D)/I(G)?
Not from the ratio alone. Cançado et al. (2011) showed that I(D)/I(G) for a given defect density depends strongly on the laser energy and peaks at about 3 nm between defects, so one ratio can match two densities; the G width and the laser energy are needed too.
What does I(D)/I(D′) tell me?
The kind of defect: about 13 for sp³ defects, about 7 for vacancy-like defects and about 3.5 for boundaries (Eckmann et al. 2012).
Make the whole figure in LabCanvas
Vector figure editor, slides, 3D models, protein structures and Prism-style graphs, in your browser.
Get started