【讨论】再问：碳膜铜网和微栅网的区别

碳膜铜网和微栅铜网的制作是不完全一样的，微栅并不是就在已制好的碳膜铜网的基础上，在其上的碳膜上做一些微孔。

我的意思是，排除制作方法，如果只看最后形貌，是否可以这样认为

最后形貌也完全不一样。碳膜膜厚度为7-10nm，而微栅膜厚度为15-20nm。

微栅上的膜是碳膜吗

确切地说，微栅上的是碳骨架。

转一个SPI网站上关于微栅的介绍。SPI算得上世界顶级电镜配件供应商，他们的说法还是很可信的。这里讲到了所有方面的问题。
http://www.2spi.com/catalog/grids/cusctgrd.php

Most samples for transmission electron microscopy must be "supported" on some kind of a thin electron transparent film, to hold the specimen in place while in the objective lens of the TEM. Only samples that are "self-supporting" do not need some additional support film.

Should an electron microscope user use a support film at all if one is not absolutely needed? Probably not, because even the very best quality of support films do result in more material being placed in the beam's path, resulting even if ever so slightly, a reduction of the contrast in the final image. But for many users, they don't have the luxury of having to make that decision: More often than not, even what appear to be self-supporting samples, still do need the assistance of a support film to cut down on sample drift.

The selection of the "right" support film has never been an easy exercise, even the most experienced of TEM experts, and the only way to really determine which support film is the best for a particular application it to just "try it!" Therefore, we at SPI Supplies would like to contribute in a useful way to the needed information base to help make the optimum selections, but the giving of such "guidance" itself is an imprecise science.

We have therefore broken down the different kinds of support films and have tried to present the "trade offs" between one support film system vs. another. We can offer no guarantees, since the range of samples studied is so wide and the environmental range so broad (from cryo temperatures to temperatures exceeding 1000°C) to which the grids and their coatings might be subjected. But the best decisions are usually made by just trying several possibilities and if one works better than the other, the choice for future samples becomes obvious.


Just remember:

There are actually two decisions to be made, one being the physical form of the support film (e. g. continuous, holey, lacey) and the other being the chemical composition of the support film (e.g. carbon, polymeric, silicon monoxide/dioxide, silicon nitride, etc. These decisions are of course always in addition to the one involving the kind of grid being used (e.g. copper, nickel, gold, molybdenum, etc.). Actually there is another decision and that is whether you should be using "conventional" TEM mesh grids or the newest technology in grid fabrication, silicon nitride membrane window grids. We can now also offer silicon oxide membrane window grids, where the membrane is silicon oxide instead of silicon nitride.

Polymer Support Films, both Carbon Coated and Uncoated

There is a choice of different polymer support films and the polymer film support most often asked for is Formvar®. We are able to make continuous films of Formvar as well as holey Formvar, even in specific hole sizes and also lacey Formvar. The final films are of an excellent quality and very stable under the electron beam in most instances. The specific grids and prices listed on our website are all for 3.05 mm diameter grids, however we can also coat without difficulty 2.3 mm diameter grids as well (however we might need additional lead time).

There are a number of other polymeric support films some of which are essentially the same or similar polyvinyl formal polymer as Formvar where as others are clearly different polymers and with different properties. One should also be aware that most of these resins are commercial grades that have been carefully selected, sometimes purified beyond commercial purities and in some instances, we are blending two or more grades of commercial materials. That is why we present the resins by name only without reference to a specific grade or molecular weight. What is used is that specific formulation that we have determined to be the very optimum for the filming of grids for this kind of application. In any case, we present the following information to help reduce some of the mystery related to the different coating resins:

* Formvar® and Vinylec® are all polyvinyl formal and while there might be some chemical differences, it is our opinion that from a grid making standpoint, they would be indistinguishable. Formvar is our standard resin, however upon special request, we would make grids from any specifically requested resin, but possibly at a slightly higher price.
* Parlodion® and Collodion, sometimes also known as "gun cotton" or nitrocellulose are also used as support films. Our standard product is based on the use of Parlodion because of its well known consistency and virtually zero batch to batch variation in properties.
* Pioloform® and Butvar®, which generically are polyvinyl butyral are also sometimes used, however even though they are not standard support film resins for SPI, we will coat with Pioloform or Butvar upon special request

Both Parlodion and Collodion in amyl acetate will float on water, while Formvar and Vinylec preparations sink. As a result, a Formvar or Vinylec film will have the structure of the glass on which it was formed, while a Parlodion or Collodion film can be made "structureless" and "featureless" by forming the films on water. This is an essential step in making the structureless carbon films on which DNA is studied.

So far we have addressed only the issue of the polymer and not the issue of whether it should be carbon coated. The general rule is this: If possible, carbon coat, because carbon coated grids will be more stable than uncoated support films. However, the addition of the carbon coating has the disadvantage that it can also impart some discernible structure to the support film system, and for those working with either nanoparticles or other fine structures for example, in sections, the structure from the carbon film could be misinterpreted as being structure coming from the sample. We find today the mix is about 50/50 between those who want their support films carbon coated vs. those who want them uncoated. To give an example of how important that carbon film could be in terms of film stability, for the filming of slot grids, the cost to produce Formvar only is about 50% higher than the cost to produce carbon coated Formvar. Why? Very simple: The yield is greater for the carbon coated Formvar grids. Remember, only SPI Supplies is batch checking by TEM our entire production and when a batch just won't stand up in the beam uncoated, we can not proceed and make such a shipment to our customers.

The SPI produced carbon only and carbon coated Formvar grids are known for their high beam stability and absence of structure. Indeed, it is generally recognized that our ability to make stable support films of the highest quality generally exceeds that of the typical user making them on an occasional basis and also exceeds the quality consistent from other commercial providers who do not have their own in-house TEM for quality checking and leave the Quality Control function up to their customers. As a result, SPI experience almost never the request by a customer to return unstable or defective grids. However, carbon films show a significant phase grain, presumably from C60 microcrystals.

Finally there is the question of using "holey" or "no holes" carbon or carbon coated polymeric support films and this depends on the kinds of samples being studied and the need to have completely unobstructed regions of sample with the effect of any substrate effects. In those instances, data is taken only from those portions of the sample essentially suspended over "holes", or in the case of lacey films, over the open areas of the lacey network. This kind of coated grid is preferred for example by those doing EELS (electron energy loss spectroscopy).

Note: If one is contemplating the application the sample from solvents (that would dissolve the support films) or examination (or treatment prior to examination) at elevated temperatures, one just can not use polymeric support films. So read on!

Carbon support films, continuous (no holes), holey carbon, and lacey carbon of the very highest quality are produced in an SPI Vacu Prep™ Table Top Vacuum Evaporator. The grain present is about the smallest possible in a vacuum evaporator and is very stable in the electron beam. When holey carbon films are needed, a holey Formvar filmed grid is then carbon coated and the polymer then dissolved away, leaving only the holey carbon film. Similarly, when lacey carbon films are needed, a similar procedure is followed.

One note of interest:

Most of the world knows these different kinds of holey support films as "holey" and "lacey". But in some parts of the world, most notably, Japan, these kinds of films are known as "microgrids". We make that point because we are sometimes asked for "microgrids" by researchers wanting to duplicate results described in a Japanese published paper and vice versa, but not realizing they are the same thing.


Thinking about making your own coated grids:

We have some advice for those contemplating doing that. And if you are contemplating the making of lacey carbon or lacey Formvar, since the process is quite different, we will also share our many years of experience for the making of lacey films as well.

Carbon support films are very popular because there is very little that can really destroy a carbon film. And holey carbon films are also popular because for those doing EELS work, or other types of work where no support can be tolerated, then the data can be taken from those portions of the sample suspended over holes.


What about the thickness?

We are often asked about the thickness of our carbon coated grids. This is a tough question to answer. Our goal at all times is to make the thickness the minimum necessary to result in a film that will be dimensionally stable in the electron beam. Anything thicker than that works against our customer's best interests. And anything thinner than that results in unstable grids. Our grid coating experts have a pretty good track record when it comes to making the optimum thickness and when they miss it, our own quality control checking (since we check each batch by TEM ourselves) rejects that batch and it never gets sent to a customer. But we are still constantly asked about the thickness, just the same. Measuring the thickness is far more easily said than done, because of the propensity of most embedding resins to swell in some indeterminable amount, a carbon film. But our best estimates, based on the appearance of the films floating on water, would put the thickness between 10-20 nm. Don't let anyone fool you into thinking they really know what is the thickness of their coated grids, measuring that thickness is an almost impossible objective.


Silicon dioxide support films:

Silicon dioxide/monoxide support films are made by the evaporation of silicon monoxide chunks from a tungsten basket. The final film is in all likelihood a mixture of SiO2 and SiO the exact ratio being an almost impossible to do kind of analysis. However, we don't see any real need even to know that number so long as the films are stable and are inert in the presence of the oxygen plasma.

Actually there are a number of not-too-well-understood reasons why a TEM user might want to be using silicon dioxide/monoxide filmed grids. For example, when organics are to be etched away with an oxygen plasma, the support film will remain. Another reason is when the grids are going to be heat treated in air to temperatures where the oxygen present would cause a reaction with the support film if it was carbon. And finally, but perhaps the most important reason, is when one is planning to do an EDS analysis for carbon and it then becomes desirable to not have any carbon present in the support film.

The links below for the different coated grids are meant to be more as "examples" than a statement of grids onto which coatings can be applied. Remember that we can coat just about anything one might want, so we can even prepare silicon dioxide coated grids that are actually a silicon dioxide coated Formvar, for example.

Allred™ substrates from titania and precious metals:

The Allred™ Substrates are produced for SPI Supplies by Dan Allred & Company. These represent a class of grid based products that, while they are in essence, "support films", they are really more than support films, since they are designed for one to do certain types of "processing", resulting in a final "composite" that is sufficiently electron transparent so that it can be characterized by TEM without further sample preparation. The films are also sufficiently rigid so that one can also do FESEM and under some circumstances, SPM as well. These grids are available in 400 mesh only unless specifically requested otherwise. In addition, these coatings are extremely thin, in fact thinnner than one can normally make, ranging from ~ 2 nm for the precious metal films up to ~ 5 nm for the titanium dioxide films. In general, these substrates at times can be a lower cost alternative to the SPI Supplies Brand Silicon Nitride Membrane Window Grids.


Available coated grid products:

* Platinum
* Gold
* Palladium
* Titania (TiO2)


Silicon nitride membrane window grids:

When high temperature studies are involved or when nanoparticles are being studied and a completely amorphous yet highly robust support is required, the system of choice could be our silicon nitride membrane window grids.


Hydrophilic nature of the SPI Supplies produced filmed grids:

Some users of these grids require a support film surface to be hydrophilic, perhaps so that a sample applied will spread out on a grid and not stay agglomerated in one small droplet. Although Formvar only as well as SiOx coated grids are stable and do not deteriorate with time in terms of their hydrophilic nature, that is not true for carbon coated grids. Indeed, and it is a recognized fact, as carbon coated grids age, they very definitely tend to become less hydrophilic. We are unaware that the aging of grids can be slowed or prevented although some have reported that refrigerated storage does result in a slowing of the process.

For some number of years, our policy was to make carbon coated grids "fresh", as ordered, so that there would never be any shelf life issues and the grids would arrive in the customer's hands with the longest possible shelf-life, that is, the time when they would remain hydrophilc. But this caused hardships for our customers who just did not want to wait and for those customers who did not require highly hydrophilic grids, this approach just was not optimum for some of our customers. We have now changed our long time policy for carbon coated grids: When you view the carbon coated grid products, the original product numbers will still reference grids that are freshly made. But for those customers who don't need the highest hydrophilic nature, we will be keeping some of these grids "in stock" and available for immediate shipment. We are introducing a new family of SPI numbers, they will have the same format as the original numbers but will have an "S"inserted to denote "From stock". Hence for those needing the highest possible hydrophilic nature to their carbon coated grids, they should order them '"freshly made". But for those who are not as concerned about freshness, and who want to have faster delivery of their grids, use the new numbers with the "S" inserted to denote that you are ordering stock items from stock. The period of time between when the grids were manufactured vs. when the goods are shipping will always be a variable, but we would estimate that the typical shipping time would be roughly four weeks.

For those who want their grids to be even more hydrophilic, SPI can offer as a special service, glow discharge exposure to greatly increase the hydrophilicity of the grids. Actually this is done in an SPI plasma cleaner. We have, as have been some of our customers, been quite pleasantly surprised at the impact of exposure in the SPI Plasma Cleaner. The exposure is at very low power and therefore the treatment is quite unlike the higher power treatments given to the grid in other types of glow discharge equipment. At 10 watts, there is not enough energy to actually "etch" the carbon which in fact is what does happen in some units that operate at higher power. So in the case of SPI treated grids, to increase the hydrophilicity, one does not have to give up on the wettability in order to keep the grid from being attacked.

For further information, consult the following reference: Aebi U. and Pollard T.D., A glow discharge unit to render electron microscope grids and other surfaces hydrophilic. J. Electron Microsc.Technique, 7:29-33 (1987).