Organic Chemistry · Structure Determination: Nuclear Magnetic Resonance Spectroscopy

13C NMR Spectroscopy: Signal Averaging and FT–NMR

7 min read
Constants verified against current reference sources (2026-08): ¹³C natural abundance 1.1%; γ(¹³C) = 6.728 × 10⁷ rad·s⁻¹·T⁻¹; γ(¹H) = 2.675 × 10⁸ rad·s⁻¹·T⁻¹; ¹³C receptivity ≈ 10⁻⁴ that of ¹H.
Want it in plain words first? Jump to Eli explains — the same idea, no jargon.
On this page 9 sections
  1. In 30 seconds
  2. Why this matters
  3. The college version
  4. Eli explains
  5. Worked example
  6. Key takeaway
  7. Check yourself
  8. Study tools
  9. Sources & references

In 30 seconds

Carbon-13 NMR faces two obstacles that proton NMR does not. First, the abundant isotope ¹²C has nuclear spin zero — invisible to NMR; only ¹³C, at 1.1% , has the spin I = ½ needed to absorb. Second, ¹³C has a much smaller gyromagnetic ratio (about one quarter of the proton's), further weakening its signals. Together these factors make an individual ¹³C signal roughly four orders of magnitude weaker than a ¹H signal from the same number of nuclei.

Modern spectrometers defeat this handicap with two techniques: (co-adding many scans so real signals add while random noise partially cancels) and Fourier-transform (FT) NMR (a short pulse excites every carbon at once, making thousands of scans fast enough to be practical). With broadband proton decoupling (next topic), these advances turned ¹³C NMR into a routine experiment.

Why this matters

¹³C NMR is the ideal complement to ¹H NMR: its spectra are simpler (one signal per unique carbon, usually singlets), its shifts span a far wider range, and it sees carbons with no attached hydrogen (carbonyls, quaternary carbons). Understanding why the experiment is insensitive (abundance, gyromagnetic ratio) and how the instrument compensates (pulsing, averaging, the FT) explains practical choices: why ¹³C spectra take many scans, why they are not integrated, and why modern NMR — including MRI — uses pulse methods.

The college version

Core Concepts

Why ¹³C is hard: spin, abundance, and gyromagnetic ratio

NMR signal strength depends on a nucleus's abundance and on the cube of its gyromagnetic ratio γ (γ controls both the spin-state population difference and the voltage induced in the detector coil). For ¹³C relative to ¹H:

S(13C)S(1H) = (0.011) × (γ(13C)γ(1H))3 = 0.011 × (0.2514)3 ≈ 1.7 × 10-4

So a ¹³C signal is about 1/5700 as strong as a ¹H signal from the same number of nuclei — roughly four orders of magnitude weaker. Constants: ¹³C abundance 1.1%; γ(¹³C) = 6.728 × 10⁷ rad·s⁻¹·T⁻¹; γ(¹H) = 2.675 × 10⁸ rad·s⁻¹·T⁻¹.

The resonance frequency: Larmor equation

A nucleus resonates at the Larmor frequency, proportional to the applied field B₀:

ν= γB02π

Since γ(¹³C) ≈ 0.2514 × γ(¹H), ¹³C resonates at roughly one quarter of the proton frequency at the same field (at 9.4 T: ¹H ≈ 400 MHz, ¹³C ≈ 100 MHz). Instrument names ("a 400 MHz spectrometer") refer to the proton frequency.

Signal averaging: the √N law

Random noise fluctuates unpredictably, but a genuine signal is identical in every scan. Adding N scans adds signal linearly while noise, being random, adds only as √N:

(SN)N = (SN)1 N

Doubling the number of scans improves S/N by only √2 ≈ 1.41 — which is why a good ¹³C spectrum requires thousands of scans, and why each scan must be fast.

FT–NMR: pulse, FID, transform

Older continuous-wave (CW) instruments swept the frequency slowly, detecting one resonance at a time — hopeless for ¹³C. In , a short radiofrequency pulse excites all nuclei at once; each responds by precessing and inducing a decaying oscillating voltage called the free induction decay (), a time-domain signal containing every resonance frequency. A converts the FID into the familiar frequency-domain spectrum. One pulse = one complete spectrum in a fraction of a second, so thousands of FIDs can be averaged in minutes — the multiplex (Felgett) advantage.

Relaxation and the recycle delay

After each pulse, nuclei must return to equilibrium before the next; the time constant is the spin–lattice relaxation time T1. If pulses arrive faster than nuclei can relax, signals saturate (weaken or vanish). Carbon T1 values range from milliseconds to many seconds, so the experimenter sets a (typically a few seconds) between pulses — a compromise between speed and signal.

Common Confusions

Do not confuseWithDifference
"¹³C NMR is just ¹H NMR for carbons"Same sensitivity and rules¹³C is ~5700× less sensitive, resonates at ~25% of the frequency, and is not routinely integrated
Signal averagingAveraging different samplesScans repeat on the same sample; signals reinforce, noise averages away
The Fourier transformCreating new dataIt only reformats the FID from time to frequency domain — no information is added
Sensitivity loss from abundanceSensitivity loss from γ aloneBoth matter: 1.1% abundance and γ³ ≈ 0.016 combine to give ~1/5700
T1 relaxationT2 relaxationT1 (spin–lattice) governs how fast you can repeat pulses; T2 (spin–spin) broadens lines
¹³C frequency¹H frequencyAt any field ¹³C resonates at about 25% of the ¹H frequency (e.g., 100 vs 400 MHz at 9.4 T)
Eli, the EliExplains learning guide

Eli explains

The same idea, in plain words

Explain it like I’m 10

Only about one carbon atom in a hundred can "sing" in an NMR machine, and its voice is thousands of times weaker than a hydrogen's. So scientists act like photographers in a dark room: take the same picture many times and stack the copies — the real picture gets brighter while random fuzz cancels out. One clever pulse takes the whole picture in a flash, so stacking thousands of copies is fast.

Worked example

Example 1: Calculating the ¹³C resonance frequency (dimensional analysis)

Find the resonance frequency of ¹³C in a 9.4 T magnet, given γ(¹³C) = 6.728 × 10⁷ rad·s⁻¹·T⁻¹.

Write the Larmor equation first:

ν= γB02π

Substitute the values:

ν(13C) = (6.728 × 107 rad s-1 T-1)(9.4 T)2π = 1.006 × 108 s-1 ≈ 100.6 MHz

Unit check: radians are dimensionless, so rad·s⁻¹·T⁻¹ × T leaves s⁻¹ (hertz) — the answer is a frequency. The same calculation for ¹H (γ = 2.675 × 10⁸ rad·s⁻¹·T⁻¹) gives 400 MHz, confirming the "one quarter" rule: 100.6/400 = 0.2514.

Example 2: How many scans do you need?

A single scan of a dilute sample gives S/N = 5, and the experimenter wants S/N = 50.

Start from the averaging law:

SN = (SN)1 N

Rearrange to solve for N before substituting:

N = (S/N(S/N)1)2 = (505)2 = 100

So 100 scans are needed. To double the S/N again (to 100) would require 400 scans — four times as long; signal averaging is a time budget, not a free lunch.

Example 3: Why one pulse measures everything at once

A CW instrument spends minutes sweeping the ~100 MHz ¹³C range; an FT instrument excites the whole range with a ~10-microsecond pulse and records the FID in about a second. If 2000 scans are needed and each takes 2 s, the FT experiment needs about 4000 s (~1 hour); the CW approach would take weeks. The multiplex advantage makes routine ¹³C NMR possible.

Key takeaways

  • ¹²C (spin 0) is NMR-invisible; only ¹³C (spin ½, 1.1% abundance) gives signals.
  • Relative receptivity of ¹³C ≈ (0.011)(0.2514)³ ≈ 1.7 × 10⁻⁴ → about 1/5700 that of ¹H (≈ 4 orders of magnitude weaker).
  • Larmor equation ν= γB0/2π: ¹³C resonates at ~25% of the ¹H frequency (100 MHz vs 400 MHz at 9.4 T).
  • Signal averaging: S/N grows as √N — quadrupling scans doubles S/N.
  • FT–NMR: one pulse excites all nuclei; the FID is Fourier-transformed to a spectrum; many FIDs are averaged quickly (multiplex advantage).
  • T1 relaxation limits pulse repetition; recycle delays prevent saturation.
  • ¹³C–¹³C coupling is negligible (≈1.2 × 10⁻⁴ probability), so carbon signals are singlets before H-coupling.

Check yourself

5 review questions from the chapter. Try each one, then open the answer.

  1. Why is ¹²C invisible in NMR, and what fraction of carbon atoms are ¹³C?

    Show answer

    ¹²C has nuclear spin I = 0, so it has no magnetic moment to interact with the field. ¹³C (spin ½) is 1.1% of natural carbon.

  2. Compute the relative of ¹³C versus ¹H and express it as "1 in N."

    Show answer

    (0.011) × (0.2514)³ = 1.75 × 10⁻⁴, i.e., about 1/5700 — roughly four orders of magnitude less sensitive than ¹H.

  3. If 16 scans give S/N = 8, what S/N will 64 scans give?

    Show answer

    S/N ∝ √N: √64/√16 = 8/4 = 2, so S/N = 8 × 2 = 16. (Quadrupling scans doubles S/N.)

  4. What is the FID, and what does the Fourier transform do to it?

    Show answer

    The FID is the decaying oscillating voltage induced by precessing nuclei after a pulse — a time-domain signal containing all resonance frequencies; the Fourier transform converts it to the frequency-domain spectrum.

  5. Why is ¹³C–¹³C coupling never observed in ordinary spectra?

    Show answer

    Observing C–C coupling requires two adjacent ¹³C nuclei; the probability is (0.011)² ≈ 1.2 × 10⁻⁴, so essentially no molecules contain such a pair.

Keep learning

Ready to build on this? Continue to the next lesson.

Study tools & related lessonsKey vocabulary · Related

Key vocabulary

natural abundance
Fraction of an element's atoms that are a given isotope
gyromagnetic ratio (γ)
Constant linking a nucleus's magnetic moment to its spin
receptivity
Overall NMR sensitivity of a nucleus (abundance × γ³)
signal averaging
Co-adding many scans to improve signal-to-noise
FID
Free induction decay: the time-domain signal after a pulse
Fourier transform
Mathematical conversion of time-domain data to a frequency spectrum
FT–NMR
NMR using pulses plus Fourier transformation
spin–lattice relaxation (T1)
Time for nuclei to return to equilibrium after a pulse
recycle delay
Pause between pulses to allow relaxation

Sources & references

  1. openstax.org — Organic Chemistry

This lesson was adapted from the open educational references above; their licenses and attributions are preserved. See Copyright & Licensing.

Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.