Electrical, magnetic, mechanical and thermal scanning probe modes each need a particular cantilever, a signal path and a holder. What each one genuinely measures, what it does not, and which options to specify while the instrument is still a quotation.
Most atomic force microscopes in university laboratories run one mode. They produce height images, the images go into papers as evidence that a film is smooth or that particles are a certain size, and the rest of the instrument's capability sits unused because nobody specified the accessories at purchase and retrofitting them means a new quotation, a new import and another eight months. The topography image is the least interesting thing a scanning probe can do.
Every additional mode reuses the same tip and the same feedback loop to record a second channel alongside height. To do that it needs three things, and all three have to be present.
That is why a mode is a purchase rather than a menu item, and why the single most valuable thing to specify at quotation is the signal access or input and output board that lets the controller talk to accessories. It is inexpensive relative to the instrument, it does not oblige you to buy any particular mode now, and without it a future mode may not be possible at all.
A conductive tip in contact with a biased sample carries a current, and mapping that current alongside topography shows where the sample conducts. This is how you find pinholes in a dielectric, conductive filaments in a resistive switching layer, grain boundary conduction in a polycrystalline film, or dead regions in a photovoltaic absorber. It needs a current amplifier with a wide dynamic range, because currents of interest can span many decades across one image, and it wears tips quickly since the contact is both mechanical and electrical.
Scanning spreading resistance microscopy is the higher force, logarithmic amplifier version used to profile dopant concentration across a polished semiconductor cross section. It is a specialist capability with an equally specialist sample preparation requirement, and it is only worth specifying if the group works on device cross sections.
Kelvin probe mapping measures the contact potential difference between tip and sample, which gives surface potential and, with a calibrated tip, work function. It is the technique for mapping band alignment across a heterojunction, charge trapped in a dielectric, potential drop across a working device, or the effect of a surface treatment on work function.
It is also the mode most sensitive to the laboratory environment. Adsorbed water changes surface potential, so a humid room gives drifting, irreproducible potential maps. In Dhaka this is a real constraint rather than a theoretical one: plan on an enclosure with a dry gas purge, or a glovebox, if the group intends to publish work function numbers.
Scanning microwave microscopy couples a microwave resonator to the tip and maps permittivity and carrier density below the surface. It needs its own holder, a source and receiver, and calibration standards, and it answers questions about buried structure that no contact mode reaches.
Electrochemical scanning probe work puts the tip in an electrolyte cell under potential control, so that surface changes can be imaged while an electrochemical reaction proceeds. That requires a liquid cell, a compatible holder and a potentiostat, and it is the natural pairing for any group already doing battery, corrosion or catalysis work with a three electrode setup.
A magnetically coated tip scanned at a small lift height above the surface responds to the gradient of the stray field, producing a map of magnetic contrast: domains, domain walls, recorded bits, particle stray fields. Two honest limitations should be stated whenever this data appears in a paper. The contrast is a field gradient, not a magnetisation, so extracting a quantitative moment from it requires modelling and assumptions. And the tip carries its own field, which can move domains in a soft magnetic sample, so what you image may be what you created. Low moment tips exist for exactly that reason.
Instead of scanning, the tip is driven into the surface and retracted while force is recorded against distance. The approach curve gives stiffness and, through a contact mechanics model, an elastic modulus. The retract curve gives adhesion and, on a polymer or a biological sample, unbinding events. Modern instruments do this at every pixel and produce modulus and adhesion maps alongside topography.
The number that comes out is only as good as two calibrations and one choice. The deflection sensitivity has to be measured against a hard surface, the cantilever spring constant has to be determined, usually by a thermal noise method, and the contact model has to be chosen. Hertz, Derjaguin-Muller-Toporov and Johnson-Kendall-Roberts assume different things about adhesion and tip shape and give different moduli from the same data. State which model was used, or the modulus is decorative.
A scanning thermal probe carries a small resistive element at the tip that acts as both heater and thermometer. Scanned in contact, it maps thermal contrast: where heat flows away quickly and where it does not. That is genuinely useful for finding buried voids, delaminations, filler distribution in a composite and hot spots on a powered device.
Turning contrast into a thermal conductivity is much harder, because the heat flow depends on the contact area, the tip condition and any water meniscus, all of which change as the tip wears. Treat scanning thermal microscopy as a comparative technique unless you have calibration samples spanning the conductivity range of interest and you re-calibrate within the same session. Used that way it answers real questions. Used as an absolute measurement it produces numbers that will not survive review.
Work backwards from the group's research questions rather than from the mode list. A battery group needs electrochemical and mechanical capability. A photovoltaics group needs conductive and Kelvin probe. A magnetics group needs magnetic force microscopy and probably variable temperature. A polymer group needs nanomechanical mapping and heating stages.
Then check five practical things before signing.