Dear Gert,

I think that an internal plasma cleaner is always a great benefit to an SEM, especially in long-term investigations. We have internal cleaners on almost every FIB and SEM in our labs and use them frequently. Once you have an SEM without a load-lock an in-situ cleaner becomes essential, because at every door opening you contaminate the chamber with hydrocarbons. SEMs with load locks are not so much affected, but after some time they are also heavily contaminated from dirty sample holders, pump oil, dirty samples, etc. With an internal cleaner you can always clean both chamber and sample, and I think this advantage weights much heavier than the disadvantage of possible performance losses.

I forgot an important part about the differences of external and internal cleaners: The in-situ cleaners are typically working with a so-called "gentle cleaning" which means that the plasma which reaches the sample and the chamber is only containing radicals at thermal energies and no high-energy ions. This leads to a slower cleaning, but it also doesn't damage sensitive samples.

In an external cleaner the samples are typically right inside the burning plasma which means they are bombarded by high-energy ions and even sputtering occurs. Where an in-situ cleaner needs 10 minutes, an external cleaner needs only a few seconds to remove hydrocarbons, but it also damages the surface, oxidizes metals and sputters the top layer slowly.

The main difference is that in the in-situ cleaners the plasma is burning in a small chamber attached to the SEM. From there the plasma is lead into the SEM chamber and has enough time and distance to uncharge and to lose its energy in collisions with the walls and residual gas.

How does it affect the vacuum?

When the plasma is running the pressure increases, of course. The two large competitors (evactron and ibss) have different models that work at different pressure ranges. In our labs we use the ibss cleaners which are able to function at full turbo pump speed (at least in our Tescan microscopes). During the cleaning the pressure increases about one order of magnitude. After cleaning the pressure instantly drops below the initial value, because less hydrocarbons evaporate from the walls.

Other cleaners require a partial shutdown of the turbo pump or going into the VP mode to reach their cleaning pressure. Depends on the model of the SEM and the cleaner.

Fun fact: In our FIB the cleaning pressure is so low that I am able to open the gun valve during plasma cleaning. So I can scan the beam and look with the BSE detector what is going on. By the way: The cleaining pressure can be adjusted with a needle valve on the cleaner.

Hi, plasmaceaning is a ruthless method for the sample. We have in our TEM lab a cleaner from binder, but try it as infrequently as possible to take. Important samples we are place already in the evening prior the day of the investigations into the TEM and these can outgas there. For cleaning we often use "beam showering" as the Japanese call it

http://www.binder-labortechnik.de/

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3064681/

As I understand in-situ cleaner should be used to remove contamination developed in a specimen chamber, during work. Have not used it. 

From older times (when working with EPMA) I remember wonders of cold trap which effectively prevented in-situ contamination. 

In-Situ plasma cleaner is a more complete solution. Usually there are two sources of hydrocarbon contamination, from sample itself or from the contaminated SEM/FIB sample chamber. In-situ plasma cleaner, such as EM-KLEEN from PIE Scientific LLC, can clean sample and chamber at the same time. EM-KLEEN plasma cleaner can remove hydrocarbon contamination at sample chamber pressure below 0.25mTorr (0.03pa). It has been demonstrated on JOEL JSM 7800F Extreme-resolution Analytical FESEM. It can effectively improve image contrast and resolution as shown in the image below with a cleaning time less than 10 minutes. Operation pressure of 0.03pa is already much lower than the operation pressure of many VP SEM. Some SEM users even use EM-KLEEN plasma cleaner to regulate the pressure inside the VP SEM for charge control purpose. It has integrated an automatic gas flow controller. It also have LCD touchscreen controller that supports 60 recipes. User can program multiple sample cleaning recipe, chamber cleaning recipe and charge control recipe without any manual adjustment.  Simply load the recipe and click the run recipe button. Charge control recipe only regulates the gas flow rate into the SEM sample chamber without igniting plasma. 

For external plasma cleaner, it is also possible to achieve ion-free gentle cleaning. Tergeo series plasma cleaner form PIE SCientific integrates two plasma sources in one machine so that it can perform direction/immersion mode plasma cleaning and remote/downstream mode plasma cleaning in one machine. 

http://www.piescientific.com/Product_Remote_Plasma_Cleaner

http://www.piescientific.com/Tergeo-EM_Plasma_Cleaner

We have a Zeiss Merlin FEG-SEM and a XEI plasma cleaner is attached on one the EDS/WDS ports. During the acceptance tests we have noticed no drop in performance (resolution), and the longer EBSD mappings run very smooth, no sign of drift or instability. 

I personally recommend an in-situ plasma cleaner, you can clean your samples and the chamber of the SEM. The plasma cleaner is integrated into the control software of the SEM (SmartSEM), which is very practical. 

During cleaning, you will partially loose vacuum. After cleaning the microscope will pump to get to its normal operating vacuum level. All of it is just clicking buttons in the software. 

The only drawback is the window of the EDS detector. Most, if not all, EDS detector windows are polished off during plasma cleaning, therefore you should retract the detector before cleaning. Some vendors nowadays offet Si3N4 windows, that can be plasma cleaned. 

In-short the cleaner is practical, and we noticed no reduction of SEM performance.

Happy researching !

Intersting information regarding the EDS detector. You only retract the detector but it doesn't mean that you don't "clean" it as well. You only reduce the effect, don't you? Would it help to use a shutter? Does plasma cleaning affect the EBSD screen as well?

As a matter of fact YES the cleaner will affect everything inside the chamber. By retracting the EDS detector you reduce the effect. However; for both EDS and EBSD detectors the "cleaning" effect seems to be negligible when retracted; unless you apply plasma cleaning every night ! For EBSD screen I think you will change it (i.e. a newby student crashes the specimen into it) way before it gets considerably thinned by plasma clean.  

We apply plasma cleaning for every 50% of all specimens, and planning to clean the whole chamber (an overnight, 12 hr cleaning) every 6 months. With this frequency, we are expecting to use the EDS-EBSD system over 8 years without any problem that can be caused by "cleaning" via plasma (according to other users experience). 

Last year during M&M conference in Portland, OR, USA; the Si3N4 window was plasma cleaned for over 72 hours, continuously to demonstrate the durability.  

A shutter can be useful; however I think it is much more cost-effective to just retract the detector itself.  

Oh, I was more thinking about an additional shutter, e.g. if you retract it a shield closes automatically and protects EDS and EBSD detector. Like a saloon door :-). 

Both the instrument and the sample must be clean = free of contamination precursor molecules.

There is plasma etching that can be very useful to etch away some of the materials on the sample or a part of the sample itself. This, unless proper parameters are used, can easily ruin nanometer-scale features of a sample. The etch rates for various materials vary.

There is downstream plasma, which is gentler, no (strong) etching occurs, but contamination precursor molecules can be removed from surfaces. A few companies sell these units for instrument and for sample cleaning. With these one can reach ultra-high cleanliness (UHC) when no carbonaceous contamination appears of the sample surface, even after hours of continuous irradiation/measurement/imaging. Depending on the vacuum level and temperature, some molecules will stay on the surfaces, for example, in high vacuum at room temperature, water molecules.

Another possibility is to use low energy (at around 100 eV) electron flood instead of downstream plasma. This works in HV or UHV and is effective to reach and maintain UHC.

It is important to be able to decide whether the sample or the instrument needs cleaning (or both). For this NIST has a sample available for free as part of a collaboration.

Postek et. al gave electron beam induced contamination a pretty thorough treatment in part 2 of "Does Your SEM Really Tell the Truth?" series. DOI 10.1002/sca.21124. Personally, I have had more success aggressively cleaning the sample ex situ because it is much easier to tailor a cleaning process to each sample. An in situ plasma cleaner can then be used to remove any light contamination that may occur in the short time it takes me to load the sample into the SEM.