How the electron source shapes resolution, running cost and daily uptime, and how to write an SEM tender that matches the research programme you actually have rather than the one on the brochure.
A procurement committee in Dhaka usually meets the SEM decision as a single number. One bid quotes a smaller nanometre resolution figure than the other, it scores higher on the technical evaluation, and eighteen months later the instrument spends most of its working week imaging jute fibres, cement clinker and fracture surfaces at magnifications where either machine would have been indistinguishable. That headline figure is measured on a gold on carbon standard, at the highest accelerating voltage, at a short working distance, in a room built for the purpose. Very little of your actual work happens under those conditions.
The electron source is the decision underneath the number. It sets what the column can do at low accelerating voltage, how much current you can put into a small probe for X-ray analysis, how often a technician opens the gun, and how badly the instrument reacts to a bad week on the grid. Getting it right is mostly a matter of being honest about the samples.
Before any bid document is opened, write down the twenty samples the instrument will look at in its first year. Not the samples in the grant proposal: the ones already sitting on the bench. In most Bangladeshi university laboratories that list falls into five groups.
If fewer than a quarter of that list falls into the last two groups, you are probably not buying a field emission instrument this cycle. That is not a counsel of modesty. It is a recognition that an over specified column in an unprepared room will underperform a modest column in a good one, and it will cost more every year it runs.
All three source types produce a focused electron probe. What separates them is brightness, meaning current per unit area per unit solid angle, and the physical size of the emitting region. Brightness decides how much current you can push into a small probe. That decides whether you get a usable signal at low accelerating voltage, where surface detail lives, and at short pixel dwell times, where drift stops ruining long acquisitions.
A bent tungsten wire heated until it emits. It is the least expensive source, the most forgiving of an imperfect vacuum, and the only one your own technician can change without a service visit. Filaments are a consumable and they fail, sometimes in the middle of a booked session, so a laboratory running a tungsten column keeps a small stock on the shelf rather than ordering one at a time. For micron and sub-micron work at moderate to high accelerating voltage, a well aligned tungsten instrument produces images no reviewer will question. The Hitachi SU3500 that Vvon supplies sits in this class, and its variable pressure chamber removes much of the coating burden that otherwise slows a shared facility down.
A LaB6 single crystal tip runs brighter and with a smaller emitting area than tungsten, so a probe of a given size carries more current. It lasts considerably longer than a tungsten filament, but it demands a better gun vacuum and it punishes careless venting. Where it pays back is analytical work: X-ray counting and elemental mapping benefit directly from beam current, and long maps benefit from stable emission. Where buyers are disappointed is the assumption that LaB6 transforms low voltage imaging. It improves it. It does not turn a thermionic column into a field emission one.
A cold or Schottky field emitter has an emitting region orders of magnitude smaller and a brightness no thermionic source approaches. That is what delivers sub-nanometre and low kilovolt performance, and it is the right choice when the research programme genuinely depends on imaging nanoparticles, thin film cross sections or lightly coated soft materials. The costs are structural rather than incidental: a much harder vacuum, a gun pump that should never be switched off, real sensitivity to vibration and stray magnetic fields, and an emitter replacement that is a service event rather than a technician task. Hitachi field emission columns come through the same channel as the SU3500, and the choice between them belongs to the sample list.
| Source | Best suited to | What you give up | Practical note for a lab in Bangladesh |
|---|---|---|---|
| Tungsten filament | Teaching, failure analysis, fibres, minerals, cement, general morphology | Fine detail at low accelerating voltage | Filaments are user replaceable; hold spares in country, not on order |
| LaB6 | Analytical work, elemental mapping, longer unattended acquisitions | Vacuum discipline and a higher source price | Worth it only if the vacuum is genuinely maintained between sessions |
| Field emission | Nanoparticles, thin film cross sections, low voltage surface imaging | Room quality, uninterrupted gun power, dependence on a service engineer | Do not order until the site survey on the actual room has passed |
An instrument that runs beats one that is better on paper. On a Bangladeshi campus, uptime is decided by three things the datasheet does not mention: the quality of the electrical supply, the presence of spare parts inside the country, and how quickly a trained engineer can reach the room. A field emission gun that loses its vacuum during an extended outage becomes a service call. A tungsten column shrugs, pumps down and carries on. That difference should be priced into the decision, not discovered afterwards.
Put two questions to every bidder in writing and score the answers. Where is the nearest engineer factory trained on this exact column, and what was the actual response time on a comparable installation in Bangladesh or the region? Then ask which parts are held locally. An instrument with a four week wait for a detector component is not a fast instrument, whatever the specification claims.
Variable pressure, sometimes called low vacuum, admits a controlled amount of gas into the chamber. Ions generated in that gas neutralise the charge building on a non-conductive surface, which means ceramics, minerals, polymers, paper, textiles and dry biological material can be imaged without a conductive coat. For a shared facility that turns around many samples from many groups, this is the single feature that most reduces queueing, because sample preparation, not imaging, is what fills the day.
Be clear about the trade. In variable pressure mode you are working mainly with backscattered electrons, resolution is reduced, and the signal is degraded by scattering in the gas. It is not a substitute for good secondary electron imaging on a properly prepared sample. It is a way of getting a useful answer quickly, and a way of imaging things you are not allowed to coat, such as forensic or heritage material.
Capital cost is settled once. The rest arrives every year, and in a university budget it is the part that gets forgotten at approval and fought over later. Build the five year figure into the evaluation and ask for it in writing at bid stage.
The most expensive microscope in the country is the one that was over specified for the room it went into and now runs at half the performance it was bought for.: Vvon Technologies installation team