MALDI-MSI vs DESI-MSI vs SIMS: A Platform Comparison for Spatial Metabolomics

Selecting an imaging mass spectrometry platform for spatial metabolomics depends on whether the priority is detecting the widest range of molecules, achieving the highest spatial resolution, minimizing sample preparation, or operating under ambient conditions. Three platforms — MALDI, DESI, and SIMS — dominate the field, but they differ at every level: ionization mechanism, spatial resolution, molecular coverage, sample preparation requirements, and the type of biological question each was designed to answer.

This guide provides a direct, data-backed comparison of MALDI-MSI, DESI-MSI, and SIMS for tissue metabolomics and lipidomics. It covers what each technology can and cannot detect, what sample preparation each requires, and what the resulting data actually looks like — enabling an informed platform decision before committing samples and budget.

For projects that combine spatial imaging with complementary LC-MS-based profiling, our untargeted metabolomics service provides comprehensive molecular identification from tissue extracts.

How MALDI, DESI, and SIMS Differ at the Ionization Level

The three platforms diverge at the very first step — how molecules are transferred from the tissue surface into the gas phase as ions. This distinction determines everything downstream: which molecule classes are detectable, how much spatial detail is preserved, and how the sample must be prepared.

MALDI-MSI: Laser desorption with a chemical matrix

The tissue section is coated with a small organic acid matrix — typically DHB, CHCA, or DAN — that absorbs ultraviolet laser energy at 337 or 355 nm. When the laser fires at each pixel, the matrix absorbs the energy, rapidly heats, and transfers protons to co-desorbed analyte molecules, producing singly charged ions detected by the mass analyser (Norris & Caprioli, 2013).

MALDI provides the broadest molecular coverage of the three platforms — lipids, metabolites, peptides, and small proteins — across an m/z range of approximately 80–1,000. The operation under vacuum limits sample types to fresh-frozen or FFPE tissue sections.

The MALDI-2 technique adds a secondary laser pulse to post-ionize neutral molecules in the gas phase, increasing sensitivity by up to 100-fold for certain lipid classes and pushing detection limits into the low femtomole range (Soltwisch et al., 2015).

DESI-MSI: Ambient ionization with minimal preparation

A pneumatically assisted electrospray generates charged solvent microdroplets directed at the tissue surface at an angle. The droplets dissolve analytes from a micro-localized area, and secondary droplets containing the dissolved molecules are splashed into the mass spectrometer inlet — all at atmospheric pressure (Takáts et al., 2004).

DESI primarily detects lipids and small metabolites in the m/z 200–1,000 range and is particularly effective for mapping phospholipid distributions across large tissue areas. Most experiments operate at 50–200 µm spatial resolution, though nano-DESI systems using pulled glass capillaries achieve 7–10 µm (Laskin & Lanekoff, 2016).

Because DESI requires no matrix, no vacuum, and minimal preparation, it offers the fastest sample-to-data turnaround of the three platforms. Published studies have demonstrated lipid-based tissue classification in under 30 minutes using DESI-MSI (Kriegsmann et al., 2022).

SIMS: Ion beam sputtering at the nanometre scale

A focused beam of high-energy primary ions — typically Bi₃⁺, Au₃⁺, or C₆₀⁺ for ToF-SIMS, or Cs⁺ or O⁻ for nanoSIMS — bombards the tissue surface, sputtering secondary ions from the top 1–2 nm of the sample. The primary ion beam can be focused to a spot diameter of 50 nm or less (Passarelli & Ewing, 2013).

SIMS achieves spatial resolution orders of magnitude beyond MALDI or DESI — routinely 100–500 nm for ToF-SIMS and down to 50 nm for nanoSIMS. The trade-off is severe: high-energy impact fragments most molecules, limiting intact detection to atomic species, small fragments, and a subset of lipids below m/z 1,000.

SIMS requires ultra-high vacuum (10⁻⁸ to 10⁻¹⁰ mbar), thoroughly dehydrated samples, and electrically conductive substrates such as silicon wafers or ITO-coated slides. These operational requirements make SIMS the most sample-preparation-intensive of the three platforms.

Spatial Resolution, Molecular Coverage, and Sensitivity: A Technical Comparison

Comparison of MALDI, DESI, and SIMS ionization mechanisms: MALDI uses matrix-assisted laser desorption under vacuum, DESI uses charged solvent spray at ambient pressure, and SIMS uses a focused primary ion beam under ultra-high vacuum.

When MALDI-MSI Provides the Broadest Molecular Coverage

MALDI-MSI is appropriate when the research question demands detecting the widest possible range of molecules from a single tissue section — particularly when the experiment must capture multiple molecular classes (lipids, metabolites, peptides) in one imaging session.

• Tumour microenvironment and oncometabolite mapping. MALDI-MSI at 10–50 µm resolution distinguishes metabolic signatures across tumour sub-regions. In glioblastoma, it maps 2-hydroxyglutarate (2-HG) distribution to IDH1-mutant tumour regions — a disease-relevant oncometabolite that bulk LC-MS would average across the entire tissue (Alexandrov, 2020). Published studies using MALDI-derived lipid signatures have reported 96% accuracy in distinguishing tumour from stroma and 95% in classifying adenocarcinoma versus squamous cell carcinoma subtypes (Kriegsmann et al., 2022).
• Drug distribution and pharmacokinetics. MALDI-MSI directly images both the parent drug and its phase I and II metabolites across tissue compartments, providing spatially resolved pharmacokinetic data that neither bulk LC-MS nor the other imaging platforms can match for combined coverage of drug, metabolite, and endogenous lipid signals.
• Neuroanatomical lipid mapping. Brain regions have characteristic lipid profiles — phosphatidylcholines dominate grey matter; sulfatides and galactocerebrosides mark white matter. MALDI-MSI maps these lipid classes to specific neuroanatomical structures at resolutions sufficient to distinguish cortical layers. For integrated spatial multi-omics studies, our spatial metabolomics service supports combined MALDI imaging with complementary transcriptomic data integration.

When DESI-MSI Offers Practical Advantages

DESI-MSI is appropriate when the experimental bottleneck is sample preparation time, when vacuum compatibility is a constraint, or when the study requires rapid screening of large tissue cohorts.

• High-throughput tissue screening. Because DESI requires no matrix application — the most time-consuming step in MALDI workflows — sample turnover is substantially faster. A single DESI experiment can proceed from tissue section to data in minutes, making it suitable for screening dozens of sections in a single day. Published studies have reported 98.2% classification accuracy in breast cancer tissue analysis and 100% accuracy in distinguishing tumour from normal lung tissue using DESI-derived lipid signatures (Kriegsmann et al., 2022).
• Rapid-turnaround tissue classification. DESI's ambient operation and minimal sample handling make it suitable for research workflows requiring fast sample-to-data turnaround. Tissue sections are analysed directly from standard glass slides with no vacuum cycle or matrix deposition. Published proof-of-concept studies have demonstrated lipid-based tissue classification within a 20–30 minute analysis window.
• Large-area tissue mapping. The combination of ambient operation and moderate resolution makes DESI ideal for mapping entire tissue sections — whole mouse brain cross-sections, large tumour resections, or tissue microarray serial sections. For targeted lipid analysis across tissue cohorts, our targeted lipidomics service provides complementary quantitative data from tissue extracts.

When SIMS Is Required for Subcellular or Isotopic Analysis

SIMS addresses a set of analytical questions that MALDI and DESI cannot: imaging at the organelle level and quantifying isotopically labelled compounds at nanometre-scale resolution. Its molecular coverage and sample preparation demands, however, make it appropriate only for specific, targeted applications.

• Subcellular lipid and drug localization. ToF-SIMS at 100–500 nm resolution images the distribution of specific lipids — phosphatidylcholine, cholesterol, sphingomyelin — across individual cell membranes, distinguishing plasma membrane from organelle membrane composition. NanoSIMS at 50 nm resolution can visualize isotopically labelled compounds within individual organelles (Passarelli & Ewing, 2013). These capabilities are inaccessible to MALDI and DESI at micrometre-scale resolution.
• Stable isotope tracing and absolute quantification. NanoSIMS is the only MSI platform capable of measuring isotope ratios (¹³C/¹²C, ¹⁵N/¹⁴N, ¹⁸O/¹⁶O) at subcellular resolution. This enables experiments tracking ¹³C-glucose incorporation into specific lipid pools within the endoplasmic reticulum versus the plasma membrane. Absolute quantification of pro-drug concentrations in individual organelles has been demonstrated at attomole sensitivity (Thomen et al., 2020).
• Limitations to consider. The fragmentation inherent to SIMS ionization limits intact molecular identification compared to MALDI or DESI. Most SIMS imaging is restricted to fragment ions, elemental species, and a subset of intact lipids. SIMS is not suitable for untargeted metabolite discovery — it is a targeted technique for mapping known molecular species at the highest possible spatial resolution.

Sample Preparation Requirements per Platform

The practical difference between these three platforms is most apparent at the sample preparation stage.

MALDI-MSI

• Tissue sections at 10–12 µm on ITO-coated conductive slides
• Matrix application by automated sprayer or sublimation (15–45 min per slide)
• Optional ammonium formate wash to remove ion-suppressing salts
• Desiccation under vacuum before loading into the instrument
• Adjacent serial section for H&E staining required for histological annotation

DESI-MSI

• Tissue sections at 10–20 µm on standard glass microscope slides — no conductive coating required
• No matrix application — the section is analysed as-mounted
• No desiccation or vacuum preparation needed
• Solvent composition (typically methanol:water 95:5) is the only variable to optimize; solvent choice affects which lipid classes are preferentially detected

SIMS

• Requires electrically conductive substrate — silicon wafers or ITO-coated slides
• Tissue sections must be thoroughly dehydrated; residual water outgasses under ultra-high vacuum
• Surface flatness at the nanometre scale is critical — uneven sections create topographic artefacts that distort secondary ion yields
• Sample must be stable under prolonged ultra-high vacuum and high-energy ion bombardment
• Chemical fixation is generally avoided — introduced exogenous elements dominate the SIMS spectrum

Sample preparation workflows for MALDI-MSI, DESI-MSI, and SIMS: DESI requires the least preparation with no matrix and no vacuum, while SIMS demands conductive substrates and ultra-high vacuum conditioning.

What to Expect in Your Results: Data Output and Interpretation

Each platform produces fundamentally different output data. Understanding these differences before starting a project prevents mismatched expectations about what the final dataset will contain and how it can be used.

For projects requiring definitive molecular identification of imaging-detected features, our spatial metabolomics service integrates MALDI-MSI with laser capture microdissection and LC-MS/MS, transitioning from spatial mapping to molecular identification within the same tissue sample.

References

1. Norris, Jeremy L., and Richard M. Caprioli. (2013). Analysis of tissue specimens by matrix-assisted laser desorption/ionization imaging mass spectrometry in biological and clinical research. Chemical Reviews, 113(4), 2309–2342.
2. Kriegsmann, Mark, et al. (2022). Comparing DESI-MSI and MALDI-MSI mediated spatial metabolomics and their applications in cancer studies. Frontiers in Oncology, 12, 891018.
3. Passarelli, Melissa K., and Andrew G. Ewing. (2013). Single-cell imaging mass spectrometry. Current Opinion in Chemical Biology, 17(5), 854–859.
4. Takáts, Zoltán, et al. (2004). Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science, 306(5695), 471–473.
5. Soltwisch, Jens, et al. (2015). Mass spectrometry imaging with laser-induced postionization. Science, 348(6231), 211–215.