One-dimensional (1D) NMR is the first stop for almost every structure question in a synthetic or analytical laboratory. A clean 1H spectrum tells you how many chemically distinct protons a molecule contains, and a 13C spectrum adds the carbon count and type information that a hydrogen trace cannot provide. For simple, well-resolved molecules this is usually enough. But the moment a structure becomes crowded, asymmetric, or stereochemically rich, the one-dimensional spectrum stops answering the questions that matter: which proton sits on which carbon, which fragments are joined together, and how far apart two groups really are in space. This is where two-dimensional (2D) NMR enters. Instead of reporting chemical shifts along a single axis, a 2D experiment spreads the signal across two frequency axes and maps the relationships between nuclei, through-bond connectivity, direct and long-range couplings, and spatial proximity. The result is a connectivity map of the molecule that is far more informative than any stack of 1D spectra. This article explains in practical terms when 2D NMR is genuinely needed for structure elucidation, which samples benefit most, and which experiment to choose for each structural question.
Understanding the value of 2D NMR begins with recognizing what it actually measures. A 1D experiment reports chemical shifts and integrals; a 2D experiment adds the critical element of relationship. Each experiment type correlates two sets of nuclei and encodes a specific kind of information, so choosing the right experiment is the central decision in any elucidation project. The table below summarizes the core experiments and the structural information each one provides, which frames the rest of this discussion.
Table.1 Core 2D NMR Experiments and the Structural Information They Provide.
| Experiment | Correlation Detected | Structural Question Answered |
| COSY | 1H–1H through-bond, typically 2–3 bonds | Which protons are neighbors in the spin system, revealing the proton backbone. |
| TOCSY | 1H–1H through an extended coupled network | Which protons belong to the same continuous spin system, even without direct coupling. |
| HSQC | 1H–13C one-bond, direct attachment | Which hydrogen is attached to which carbon; resolves overlapping 1H signals. |
| HMBC | 1H–13C long-range, 2–4 bonds | How fragments connect, including quaternary carbons and heteroatom centers. |
| NOESY / ROESY | 1H–1H through space (typically within 5 Å) | Which protons are spatially close, informing relative configuration and conformation. |
Signal overlap is the most common reason a structure stops being interpretable. In a molecule with many similar protons, such as a long alkyl chain, a steroid core, or a molecule with multiple equivalent-looking methylene groups, 1H signals collapse into broad, unresolved envelopes. Integration becomes unreliable, coupling patterns are obscured, and it is impossible to say how many distinct protons each region contains. A 2D experiment solves this by spreading the overlapping information onto a second dimension. An HSQC, for example, separates protons by the chemical shift of the carbon they are attached to, so two protons that coincide in the 1H spectrum become distinct when viewed against their different 13C shifts. The same principle makes COSY and TOCSY far easier to read in crowded spectra, because connectivity is displayed as discrete cross-peaks rather than buried in overlapping multiplets.
A 1D 1H spectrum can tell you that a proton exists, but it cannot always tell you which carbon it belongs to. That linkage is exactly what HSQC provides: each cross-peak pairs a 1H chemical shift with the shift of its directly attached 13C. This direct assignment is essential whenever two candidates for the same proton have similar shifts, or when a carbon count in the 13C spectrum needs to be reconciled with the proton count. HSQC also exposes subtle issues, such as diastereotopic protons on a single carbon that give two distinct 1H–13C cross-peaks sharing the same carbon, a pattern that is invisible in 1D data and that frequently changes how a spectrum should be interpreted.
The hardest part of many elucidation problems is not reading individual fragments but joining them together. COSY and TOCSY establish the proton skeleton within a continuous spin system, but where two fragments meet through a quaternary carbon, a carbonyl, or a heteroatom, there is no proton–proton coupling to follow. This is the situation that makes long-range heteronuclear correlation essential. HMBC maps protons to carbons two, three, or four bonds away, so a proton on one fragment can be connected to a carbon on the neighboring fragment even when no proton bridge exists between them. Without HMBC, the assembly of fragments into a full skeleton rests on guesswork, and an incorrect connection is a real risk.
Quaternary carbons, carbonyl groups, and heteroatom-substituted centers carry no attached protons, which makes them silent in 1H and COSY data and difficult to place in a structure. DEPT editing can count them, but it cannot say where they sit. HMBC is the experiment that repairs this gap. Every proton within two to four bonds of a quaternary center will show a cross-peak to that carbon, so a set of surrounding protons effectively anchors the position of the otherwise invisible atom. This is indispensable for molecules built around carbonyl, quaternary, or nitrogen-substituted cores, where the entire connectivity map may otherwise be impossible to assemble.
Structural isomers that differ only in the position of a substituent or a ring junction often produce nearly identical 1D spectra. A single 1H or 13C trace may not reveal whether a methyl group sits at one position or another, because the chemical-shift differences are small and both candidates fit the data. In these cases the connectivity information in HMBC, and the spatial information in NOESY, decides the issue. A long-range correlation from a diagnostic proton to a specific ring carbon, or a through-space contact that is only possible in one isomer, cleanly separates the candidates. The same logic applies to distinguishing regioisomers in substituted aromatics and heterocycles, where a misplaced substituent changes the correlation pattern in a way that 1D data cannot capture.
Connectivity experiments establish which atoms are bonded to which, but they say nothing about how the molecule is folded in space. Relative stereochemistry and conformation depend on through-space proximity, which is measured by the nuclear Overhauser effect. NOESY, and ROESY for medium-sized molecules, detect protons that lie close together in three dimensions, typically within about 5 Å, regardless of how far apart they are along the bond network. If two groups show a cross-peak that is only geometrically possible on one side of a ring or one face of a double bond, the configuration is assigned. For conformational questions, the relative intensity of NOE contacts reveals which groups are preferentially positioned near each other, which is central to understanding how a flexible molecule behaves. This makes 2D NMR the method of choice whenever stereochemistry matters and a crystal structure is not available.
Table.2 Common Structure Elucidation Problems and the 2D Experiment That Resolves Them.
| Structure Elucidation Problem | 1D Data Limitation | Recommended 2D Experiment |
| Overlapping 1H signals | Proton count and integrals unreliable | HSQC to separate protons by attached carbon shift |
| Ambiguous proton–carbon attachment | Cannot assign 1H to specific 13C | HSQC |
| Fragments cannot be joined | No proton bridge across the connection | HMBC long-range correlations |
| Quaternary or proton-deficient centers | No 1H signal to locate the atom | HMBC to anchor via surrounding protons |
| Positional or constitutional isomerism | 1D shifts too similar to distinguish | HMBC connectivity plus NOESY spatial contacts |
| Relative stereochemistry and conformation | 1D data contain no spatial information | NOESY / ROESY through-space contacts |
Some sample classes are far more likely than others to demand the extra dimension that 2D NMR provides. Recognizing these patterns early helps a team decide, before spending time on 1D interpretation, whether a 2D suite is justified. The categories below are the ones that recur most often in elucidation projects, each with a distinct reason that 1D data are insufficient.
Synthetic intermediates, especially those with multiple stereocenters, fused rings, or heavily functionalized backbones, accumulate exactly the kind of proton congestion that defeats 1D analysis. As a route proceeds, the products grow in complexity, and an intermediate that was interpretable at one step can become impossible to assign at the next. HSQC and HMBC restore clarity by placing every proton and carbon unambiguously, and COSY confirms the connectivity of the carbon skeleton. This is particularly valuable during route optimization, where confirming that a new bond formed at the intended position, and not a regioisomer, is essential before committing to the next step. Integrated structure characterization using a 2D suite gives synthetic chemists confidence that the molecule in hand is the molecule they intended to make.
Impurity and byproduct identification is one of the highest-value applications of 2D NMR, because the analyte is by definition unknown and often present in small amounts. A degradant may be a rearrangement product, a ring-opened form, or a structurally unexpected species that no amount of 1D data will reveal. 2D experiments supply the connectivity and spatial information needed to propose a full structure, and the logic applies equally to impurity isolation and identification programs. When combined with mass spectrometric data for the molecular formula, a compact 2D suite, typically COSY, HSQC, and HMBC, frequently allows an unknown to be assigned with confidence. For stereochemical unknowns, a NOESY or ROESY experiment adds the final piece, distinguishing diastereomers that share the same connectivity.
Natural products are the classic test of any structure elucidation platform. These molecules are built from many rings, oxygenated or nitrogenated centers, and stereocenters, and they present overlapping signals in every 1D spectrum. A complete natural-product assignment typically requires the full battery of experiments, and the workflow is so standard that it has become a template: COSY defines the proton spin systems, HSQC attaches every proton to its carbon, HMBC assembles the fragments into a skeleton and locates every quaternary and heteroatom center, and NOESY fixes the relative configuration. The information density of a well-executed 2D analysis rivals a crystal structure for the connectivity and configuration questions, with the advantage that the sample remains fully recoverable. For structurally dense small molecules that are not natural products, such as highly substituted drug-like scaffolds, the same logic applies.
Biopolymers and their fragments present a special challenge because they contain many chemically similar building blocks. In an oligosaccharide, each sugar residue generates a set of closely spaced anomeric and ring signals; in a peptide, many backbone and side-chain protons occupy a narrow chemical-shift window. Distinguishing individual residues and connecting them in the correct sequence is impossible from 1D data alone. TOCSY is especially valuable here because it traces all the protons within a single residue, even when direct couplings are hidden, while HSQC and HMBC identify and connect the individual sugar or amino-acid units. The result is a residue-by-residue assignment that underlies sequence determination, linkage analysis, and detailed conformational study.
Compounds that exist as close structural isomers, or as mixtures of diastereomers, benefit disproportionately from the second dimension. NOESY and ROESY are the decisive tools for relative configuration, detecting contacts that are only geometrically possible in one arrangement. The same spatial readout supports conformational analysis of flexible molecules, where the preferred folded or extended forms are reflected in the pattern and intensity of NOE contacts. Whenever a project requires knowing not just what is bonded to what, but how the whole assembly is oriented in space, a 2D spatial experiment is the difference between an assumption and an assignment.
Table.3 Sample Types and the 2D NMR Experiments Most Often Applied to Them.
| Sample Type | Main Structural Challenge | Most Useful 2D Experiments |
| Complex synthetic molecules and intermediates | Signal congestion and connectivity confirmation | COSY, HSQC, HMBC |
| Unknown impurities, byproducts, degradation products | Full structure of an unknown, often at low level | COSY, HSQC, HMBC, NOESY |
| Natural products and dense small molecules | Multi-ring skeleton, quaternary centers, stereochemistry | COSY, HSQC, HMBC, NOESY |
| Oligosaccharides, peptides, signal-rich molecules | Many similar residues, sequence assignment | TOCSY, HSQC, HMBC |
| Isomeric and stereochemically complex compounds | Relative configuration and conformation | NOESY / ROESY |
Tell us about your compound and its elucidation question, and our scientists will recommend the right 2D experiment suite.
The single most important skill in 2D NMR is not running every experiment on every sample, but choosing the smallest suite that answers the question. Each experiment is designed to detect one kind of correlation, and running an unnecessary experiment wastes instrument time and adds little. The guidance below maps each core experiment to the specific structural question it answers, so a project can be planned as a short, efficient sequence rather than a broad sweep.
COSY, or correlation spectroscopy, detects through-bond coupling between protons that are separated by two or three bonds. Each cross-peak connects two coupled protons, and by following the cross-peaks it is possible to trace a continuous chain of protons along the carbon skeleton. COSY is the natural starting point after 1D 1H data, because it converts the overlapping multiplet patterns of the 1D spectrum into a clean map of neighboring relationships. It is most informative for molecules with well-resolved proton spin systems, and it immediately reveals connectivity that 1D coupling analysis can only guess at.
TOCSY, total correlation spectroscopy, extends the logic of COSY across an entire continuous spin system. Where COSY shows only directly coupled neighbors, TOCSY reveals correlations between all protons within a coupled network, even those separated by several bonds with no direct coupling between them. This makes TOCSY invaluable for identifying the boundaries of a spin system, such as distinguishing the protons of one sugar ring or one amino-acid residue from those of the next. A TOCSY trace effectively marks out whole building blocks, which is why it is the workhorse for oligosaccharides, peptides, and any molecule composed of repeated, similar units.
HSQC, heteronuclear single-quantum coherence, correlates each proton with the carbon to which it is directly attached through a one-bond coupling. The experiment places every proton in the context of its carbon, resolving 1H overlaps and giving a single, unambiguous assignment of each proton to its carbon. HSQC is also an efficient way to obtain a 13C-type spectrum at much higher sensitivity than a direct 13C experiment, because detection runs through the abundant protons. Because the number of cross-peaks equals the number of directly attached proton–carbon pairs, HSQC naturally reports the carbon multiplicity, and it is the essential foundation for any HMBC interpretation, since HMBC shifts must be read against a known HSQC assignment.
HMBC, heteronuclear multiple-bond correlation, detects long-range proton–carbon couplings, typically two to four bonds. This is the experiment that connects fragments across quaternary carbons, carbonyls, and heteroatom centers where no proton bridge exists, and it is indispensable for locating the proton-deficient regions that COSY and HSQC cannot reach. Because every proton within several bonds of a quaternary carbon shows a correlation, a small set of surrounding protons anchors the position of the otherwise silent atom. HMBC is therefore the decisive experiment for skeleton assembly, and nearly every structure elucidation project depends on it at the assembly stage.
NOESY and ROESY measure through-space proximity rather than through-bond connectivity. A cross-peak appears when two protons lie close in three-dimensional space, typically within about 5 Å, whether or not they are bonded anywhere near each other. This spatial readout is the basis for determining relative configuration and conformation. NOESY is used for small molecules where the positive NOE regime applies, while ROESY is preferred for medium-sized molecules where NOESY signals are weak or can change sign. A cross-peak that is only geometrically possible on one face of a ring, or between groups that can only approach each other in one conformation, decides the stereochemical and conformational question.
Table.4 Choosing the Right 2D NMR Experiment for Each Structural Question.
| Structural Question | Experiment | Correlation Type |
| Which protons are coupled neighbors? | COSY | 1H–1H, 2–3 bonds |
| Which protons belong to one spin system? | TOCSY | 1H–1H, whole coupled network |
| Which carbon holds each proton? | HSQC | 1H–13C, one bond |
| How do fragments connect, and where are quaternary carbons? | HMBC | 1H–13C, 2–4 bonds |
| What is the relative configuration and conformation? | NOESY / ROESY | 1H–1H, through space |
Our NMR specialists match each structural question to the smallest, most informative experiment suite for your sample.
2D NMR is inherently less sensitive than 1D NMR, because the signal is distributed across a second dimension and sampled over a longer experiment. Good sample preparation and appropriate instrument choices therefore matter more here than they do in routine 1D work. The considerations below are the ones that most directly determine whether a 2D dataset is useful or marginal.
The amount of material, its concentration, and the choice of solvent set the practical ceiling on what a 2D experiment can achieve. As a rule, more sample gives a cleaner result, and higher concentration shortens acquisition time by improving signal-to-noise. A few concrete benchmarks keep a project realistic:
Sample amount: for routine 1D 1H and 13C work, roughly 1–5 mg is usually sufficient, whereas a complete 2D suite typically calls for 10–20 mg or more of a small molecule in a standard 5 mm tube. Trace unknowns can be tackled with far less, but at the cost of many more scans and a sensitive probe.
Sample concentration: higher concentration improves signal-to-noise and shortens acquisition, so dissolving fully at the highest workable concentration, commonly a few to tens of millimolar for a 5 mm tube, is good practice. Un-dissolved particles degrade shimming and broaden every signal, so complete solubility matters more than the absolute amount.
Solvent selection: the deuterated solvent should separate the signals of interest from solvent and residual-water peaks, and exchanging or labile protons that would otherwise confuse the spectrum can be suppressed or controlled. Sample purity is equally important, since a 2D experiment of a mixture folds the contaminants into the map and complicates every assignment.
Higher magnetic field strength increases both resolution and sensitivity, which translates directly into better-resolved and faster 2D data. Dispersion along the 13C axis improves with field, so signals that overlap on a lower-field instrument separate cleanly at higher field. Sensitivity gains also matter for the heteronuclear experiments, HSQC and HMBC, which transfer magnetization from protons to the less abundant carbon-13 nuclei and benefit substantially from a stronger magnet and a cryogenic probe.
Field strength: modern high-field systems, commonly in the 500–600 MHz class for routine work and higher for the most demanding cases, resolve crowded signals far more cleanly than older low-field instruments. The 13C-dispersion benefit is directly proportional to field strength, so pushing resolution is often the cheapest route to a cleaner HSQC or HMBC.
Probe configuration: a cryogenically cooled probe can recover much of the sensitivity lost when only a small amount of sample is available, and is often the deciding factor for trace impurity and natural-product work. The choice of probe design, whether a dedicated 1H/13C probe or a broadband configuration, also governs which nuclei and experiments run efficiently. The practical message is that the instrument should be matched to the problem, and where sample is limited or complexity high, a higher-field system with a sensitive probe is the investment that pays off.
Table.5 Practical Considerations for Successful 2D NMR Data.
| Factor | Why It Matters | Recommended Practice |
| Sample amount | 2D is less sensitive; more sample gives cleaner, faster data | Use several milligrams where available; expect longer acquisition for trace samples |
| Concentration | Higher concentration improves signal-to-noise | Dissolve fully at the highest workable concentration in a standard 5 mm tube |
| Solvent | Affects shimming and peak separation | Choose a deuterated solvent that separates target signals from solvent peaks |
| Sample purity | Contaminants fold into the 2D map | Purify before analysis; interpret against a clean sample |
| Field strength | Higher field improves resolution and sensitivity | Prefer higher field for crowded or low-level samples |
| Probe configuration | Determines efficiency of heteronuclear experiments | Use a cryogenic probe for trace and heteronuclear work |
From crowded synthetic intermediates to unknown degradation products, our NMR team designs the right experiment suite for your elucidation question.
BOC Sciences provides a dedicated NMR capability within its broader analytical services platform, supporting pharmaceutical, biotechnology, and fine-chemical clients from early research through mature development. Our NMR laboratory combines high-field instruments, sensitive probe configurations, and experienced elucidation scientists who plan every dataset around the structural question rather than a generic acquisition routine. Whether the need is a routine connectivity confirmation, a complete unknown structure, or a stereochemical assignment, projects are designed to return the smallest suite of experiments that delivers a defensible answer.
Our NMR testing service covers the full range of experiments described in this article, from routine 1D 1H and 13C spectra to complete 2D suites including COSY, TOCSY, HSQC, HMBC, and NOESY or ROESY. Samples are accepted as synthetic intermediates, APIs, impurities, natural products, formulations, and reference materials, and each project is scoped with the analyte and the question in mind. For stereochemical and conformational questions, dedicated NOESY and ROESY experiments are run and interpreted with the spatial relationships made explicit, and where connectivity must be proven, a full COSY–HSQC–HMBC assembly is executed and reported with clear assignments. Every result is delivered as processed spectra with detailed peak assignments, so that the structural conclusions are transparent and reproducible.
When a standard acquisition is not enough, our scientists develop fit-for-purpose NMR methods from the experimental question outward. Method development selects the experiment type, acquisition parameters, and solvent system needed to resolve the specific connectivity, impurity, or stereochemistry problem, balancing resolution against instrument time and sample availability. Once a candidate approach is in place, method validation demonstrates that it performs reproducibly and delivers the intended information, with the setup documented for transfer to client laboratories where repeat testing is required. This is particularly valuable for impurity identification workflows and for reference-standard characterization, where the NMR signature must be stable and comparable across sites.
Many structure elucidation problems cannot be solved by NMR alone, and our laboratory is structured to combine NMR with the rest of the analytical platform. Molecular-formula data from mass spectrometry narrow the candidate list, and MS testing supplies the elemental composition that NMR connectivity must match. Complementary spectroscopy and hyphenated techniques add functional-group and positional evidence, and an integrated structure characterization package assembles all of these into a single, coherent structural conclusion. Projects are managed with a single point of coordination, so clients receive consolidated data packages rather than disconnected reports, and difficult or low-level samples are handled with the dedicated attention they require.
Table.6 NMR and Structure Elucidation Services at BOC Sciences.
| Service Name | Description | Inquiry |
| NMR Testing | Routine 1D and complete 2D NMR suites, including COSY, TOCSY, HSQC, HMBC, and NOESY or ROESY, with detailed peak assignments. | Inquiry |
| Structure Characterization | Integrated connectivity, configuration, and conformation analysis combining NMR with complementary techniques into a single structural conclusion. | Inquiry |
| Method Development | Fit-for-purpose NMR method development, selecting experiment type and parameters for the specific elucidation question. | Inquiry |
| Impurity Isolation and Identification | Full structure determination of unknown impurities and degradation products using NMR connectivity and spatial data. | Inquiry |
Connect with our NMR and structure elucidation specialists to discuss your connectivity, impurity, or stereochemistry project and receive a tailored experimental plan.
If you have any questions or encounter issues on this page, please don't hesitate to reach out. Our support team is ready to assist you.
