Peak #10 · NMR Training δH 9.05 · δN 118.2

NMR Training: the ¹H–¹⁵N HSQC

The HSQC is the first spectrum almost every protein NMR project records, and the one you will record most often. This guide walks through it with real data from a Bruker Avance NEO 700 MHz spectrometer with a cryoprobe: what every peak and artifact means, what every pulse in hsqcfpf3gpphwg does, how to set up the experiment command by command in TopSpin 4, how to tell that the instrument is working, and how to process the result.

Start the training Analyse a titration

Recording a ¹H–¹⁵N HSQC: a hands-on guide

Everything on this page comes from one real experiment: the spectrum, the pulse program, every parameter and the audit trail. Click the figures to explore them, follow the setup step by step, and test yourself at the end.

Magnet Bruker Ascend 700Console Avance NEOProbe 5 mm QCI cryoprobe (H/F–C/N–D, z-gradient)Software TopSpin 4.2Pulse program hsqcfpf3gpphwgSample 100 µM ¹⁵N-labelled protein, pH 7, 298 KRun time 38 min

Training

1Anatomy of an HSQC spectrum

Each peak is one N–H group: its ¹H shift on the horizontal axis and the shift of the ¹⁵N it is bonded to on the vertical axis. A folded protein gives roughly one peak per residue (except prolines and usually the N-terminal residue, whose NH₃⁺ exchanges too fast), plus side-chain NH and NH₂ groups. Click a numbered region to learn what it tells you, or click any peak to see its 1D slices.

2The pulse sequence: hsqcfpf3gpphwg

An HSQC moves magnetization from ¹H to the attached ¹⁵N, lets the ¹⁵N shift evolve during t₁, brings the signal back to ¹H and records it. Flip-back pulses keep the water magnetization along z, and WATERGATE dephases any residual transverse water. Click any pulse, gradient or delay, or a line of the pulse program, or step through the sequence in time order.

Narrow bars = 90° · wide bars = 180° · bells = shaped/selective pulses · grey lobes = gradients

Lines in white can be clicked. Comment lines (;) explain parameters; ased in TopSpin shows exactly the parameters this program uses.

3Before you go to the magnet: the sample

Most failed HSQCs fail in the tube, not in the spectrometer. Tick through this list before you book time.

¹⁵N labelled protein, 50–500 µMThis run used 100 µM and needed 16 scans (38 min) on a cryoprobe. Halving the concentration needs four times the scans for the same signal-to-noise. 5–10% D₂O addedThe deuterium signal is what the lock and TopShim use. Without D₂O you cannot lock. Buffer pH 6–7, low saltAmide protons exchange with water faster at higher pH and temperature, which weakens or erases peaks. Salt lengthens the 90° pulse on a cryoprobe and costs sensitivity; keep it as low as the protein allows. Avoid protonated buffer components at high concentrationStrong non-¹⁵N signals (HEPES, Tris, glycerol, MES) leave t₁-noise stripes, as at 3.5–3.7 ppm in this spectrum. Phosphate or deuterated buffers are cleaner. Correct volume and a clean, unscratched tubeAbout 500–550 µL in a 5 mm tube (or a matched Shigemi/3 mm tube for small or salty samples). Set the depth with the facility gauge. No bubbles, no precipitateSpin the tube briefly in a hand centrifuge. Bubbles and particles ruin the shim and the water suppression. Reference compound (optional)A trace of DSS gives a 0.00 ppm ¹H reference; ¹⁵N is then referenced indirectly (section 8). Temperature decidedRecord at the temperature you will use for all related spectra (298 K here). Chemical shifts move with temperature.

4Setting up the experiment, command by command

The order matters: each step depends on the one before. Open a step to see what the command does, why it is needed, what a good result looked like on this spectrometer, and what to do if yours looks different. Commands are typed in the TopSpin command line.

Safety first

The magnet is always on. Keep tools, gas cylinders, steel chairs and other iron or steel objects away from it; watches, phones and bank cards can be damaged close to it.
People with pacemakers, defibrillators, neurostimulators or other active implants must stay outside the line marked by the facility (usually the 5-gauss line).
Cryogens (liquid helium and nitrogen) are handled only by trained staff. A loud hiss and a white cloud around the magnet means a quench: the escaping helium can displace oxygen, so leave the room at once, alert the facility and do not go back in.
Never force a tube into the magnet, and use only clean, undamaged tubes and spinners.
Your facility’s induction and rules always take priority over this page.

What really happened: the audit trail

TopSpin logs every change in audita.txt inside the dataset. This is the log of the run on this page (13 August 2025).

Times are local (EDT); the acquisition entries in the audit trail are written in UTC (+0000) and have been converted. The first acquisition was stopped after 6 minutes and restarted; zg on the same EXPNO overwrites the data (“raw data deleted by zg”). The final run took 38 min 4 s, matching the time that expt and the planner in section 5 predict.

5The parameters that decide time, resolution and sensitivity

Change the numbers to see the consequences before you spend magnet time. The defaults are the values from this run; the calculator predicts its duration to within a minute.

Acquisition parameters of this run

Experiment planner

¹H frequency

MHz

SW ¹H (F2)

ppm

TD F2

SW ¹⁵N (F1)

ppm

TD F1

NS DS

¹⁵N T₂ (assumed)

6Is the spectrometer OK? Checks before you press `zg`

Compare your numbers with this good run on the same instrument. A value far outside the reference does not always mean a fault, but it always deserves a second look. These are reference values from one session; your facility’s current values take priority.

7What can go wrong, and how it looks in your data

Each problem below is shown on the same spectrum region of this run (δH 7.4–8.95, δN 107.8–118.2 ppm). The left panel is the real spectrum; the others were simulated from it so you can compare like with like. Contour levels are identical in every panel unless stated. The slice below each set cuts through the same peak (δH 8.46, δN 113.7 ppm).

TopSpin procedures referred to below are described in the Bruker TopSpin 4 User Manual (sections given in each panel). Figures on this page are original illustrations made from this dataset or drawn as schematics.

8Processing the spectrum

The processed spectrum above was made from 1500 × 128 raw points with these settings (from procs, proc2s and the processing audit trail).

Steps

Settings used

The processing log shows the forward linear prediction being tuned step by step (NCOEF 2 → 4 → 10 → 20 → 22 → 28 → 32) before zero-filling was raised to 4K × 1K. Linear prediction extends the short ¹⁵N time signal to sharpen peaks; too many coefficients on noisy data can create false peaks, so compare with a spectrum processed without it.

9Processing after data collection

How a non-uniformly sampled HSQC goes from raw data to a finished spectrum with NMRPipe, step by step. Everything below is transcribed from two screen recordings of a real processing session: the commands, the scripts and what the programs printed. Folder names have been changed.

Dataset¹H–¹⁵N HSQC, 800 MHza different experiment from sections 1–8

SamplingNon-uniform (NUS)60 complex ¹⁵N increments recorded, NS 32, 298 K

SoftwareNMRPipe, nmrDraw, ISTrun on NMRbox

CompareSection 8the same job in TopSpin, for uniformly sampled data

The pipeline

Press ▶ Replay to watch the commands of a step being typed. Long program output (the IST progress) is shortened.

10Troubleshooting

Symptom first, then the usual causes, then what to do.

11Test yourself

Nine questions on this page’s material. Pick an answer to see the explanation.

★Glossary

The ideas behind the page first, then every parameter used by hsqcfpf3gpphwg with its value in this run. The search box filters both lists.

Concepts

Parameters

Spectrum and parameters: ¹H–¹⁵N HSQC recorded on a Bruker Avance NEO 700 MHz spectrometer with a QCI cryoprobe, TopSpin 4.2. Facility procedures and safety rules always take priority over this guide.