What's new

Scanning Electron Microscope Blade Edge Images

So what if we accept your thesis, but say that the SEM is picking up the area of the blade as being solid platinum? Is it able to detect that the coating is only one atom thick or is it considering a larger volume than is actually on the apex? This seems possible, but I don't feel like going and reading how the SEM works again. The atom thick layer would be on either side of the apex, encapsulating it so it seems possible because these aren't the images from another site where he cut the blades in half before scanning.
Not clear what depth these SEM analysis went to as the depth is dependent on the energy level used for the scan. One published report I've seen talks to a depth of 2 - 5 micrometers (microns) for scans of metals. A micrometer is one thousandth of a millimeter so we are talking about a thin subsurface blade layer. This is a small fraction of the width of a human hair that can run at an average of around 70 microns. For comparison the blades are made from metal that is 0.09 mm or 90 microns thick.

2-5 microns is still a layer that is many atoms thick. An iron atom is only .126 nanometers wide. A nanometer is one thousandth of a micrometer so a 2 micron analysis would over 15,000 atoms thick. Again based on the actual power of the above SEM analysis the depth scanned could be less than a tenth of 2 microns but that would still be 1,500 atoms thick.

One Gillette razor blade patent application I've seen that discusses coatings measures thickness in angstroms. 10,000 angstroms equal a micron/micrometer. Thickness ranged from 100-200 angstroms for an inner chromium/platinum coating to 1,500 - 4,000 for the outer teflon layer. At 10,000 angstroms/micron we are looking at a coating layer that could roughly range up to 0.5 of a micron in this case. The 2014 SEM analysis clearly went deeper than this on coated blades since iron was always the most common element found.

 
Last edited:
One Gillette razor blade patent application I've seen that discusses coatings measures thickness in angstroms. 10,000 angstroms equal a micron/micrometer. Thickness ranged from 100-200 angstroms for an inner chromium/platinum coating to 1,500 - 4,000 for the outer teflon layer. At 10,000 angstroms/micron we are looking at a coating layer that could roughly range up to 0.5 of a micron in this case. The 2014 SEM analysis clearly went deeper than this on coated blades since iron was always the most common element found.
If this is fun for you, it used to be for me, you might be able to estimate the volume of Platinum that would be sputtered on the edge. I am pretty sure you can see how wide the strip of coating on the apex is with the naked eye, maybe estimate at 1mm? Then, you already have the thickness from the patent so you just need to look up the density and do some math. Maybe you could produce a range? I like to do modelling to test ideas and see if I can get into a range that seems believable, then see how that relates to these SEM analyses. It seems like the percentage is not very relevant for figuring out the total quantity of any one material, only it's presence as you have said. Maybe it's the size of the test area being analysed is unknown? I guess this works like a core sample, but one that you don't have it's total weight so percentages are of questionable value. I am just throwing things at the wall.
 
Hi @APBinNCA, Great suggestion, combining a physical with the economic analysis will give us a better answer.

The calculation you suggest is doable with some assumptions and use of the density of platinum that is 21.45 grams per cubic centimeter of metal (cm^3) or .02145 grams per millimeter cubed mm^3. Assumptions used confirm expectation that only a small amount of platinum is likely in our coated blades. 0.018% by weight or about 29 cents worth per 100 pack based on a coating of the entire blade.

Calculation Data/Assumptions:
  • DE blade surface area is 43mm x 22mm x .9 (holes and edge shaping) or 851.4mm^2
  • Entire blade surface is coated (worst case, may be a lot less, patent unclear as to how much of the blade is coated)
  • Estimated platinum coating at 50 angstroms (1/3 to 1/2 of chromium/platinum layer) that is the same as 0.000005 mm
  • 851.4 x .000005 = a platinum volume of 0.004257 mm^3
  • Multiply estimated volume by density per mm^3 or: .004257 x .02145 = .0000913 grams/side
  • Times two sides = .0001826 grams/blade
  • Times 100 for platinum in a 100 blade box gives us 0.01826 grams of platinum
  • 0.01826 grams (roughly one hundredth of a gram) of platinum represents 0.036% of the 50 grams of metal in a 100 blade pack. As expected this is a fraction of the .125% used in the economic analysis above. Well below the 2.31% SEM analysis result.
  • At $32 per gram this represents 58 cents of platinum in a 100 blade pack.
  • If only a portion of the blade is coated then there is even less platinum. A thicker coating will add more. If just the edges are coated (roughly 2.5mm) then only 5 x 43mm or 215mm^2 is coated per side. This would only require 1/4 of the platinum or around 15 cents worth. Roughly 0.009% of the blades by weight.
Feel this is much closer to reality than the $1-$2 assumption used in the economic analysis. Rolled strip steel for razor blades is listed for roughly $2 -$4 per kilogram on alibaba. Assuming double that for Swedish steel, $8 per kg., there is an estimated 40 cents worth of steel in a 50 gram 100 blade package. 58 cents for a platinum coating is starting to get into a reasonable ballpark. Even more so if only part of the blade is coated. Many 100 packs of blades need to be profitable at a selling price of under $10. $8 for Astra SP blades.

Back to the question of validating the 2014 SEM analysis. Since this methodology is only looking a sample of the blade surface area to a likely depth of 2 - 5 microns we should only use it to confirm that an element is present. As the methodology relies on a homogeneous consistency, that is not the case with coated razor blades, we cannot rely on the weight percentages as those will overstate the presence of elements in the coating that are near the surface.

Interesting exercise. Feel free to change any assumptions and recalculate if you would like or have better information on the assumption areas.
 
Last edited:
DE blade surface area is 43mm x 22mm x .9 (holes and edge shaping) or 851.4mm^2
Did you forget to multiply by 2 since there are two sides? I went and grabbed a blade since I am being too lazy for the amount of thought you are puting into this. On this, I would say 0.5mm X 37mm X 4 or 74mm^2 is more likely. The ground height of the bevel is this much so that is the maximum area that is sputtered with platinum. Remember, platinum is designed to strengthen the edge to prevent acidic degradation. This goes back to the 1960's patents. Logic also says that the whole blade isn't coated because you don't harden the entire blade otherwise it would become brittle.

So we are dealing with roughly 8% of what you got which seems even more likely, say 3 pennies per 100. On the economic front, modern manufacturing makes up around 20% cost of an item(or less). So $2.50/100 blades, we might be getting closer as that puts platinum at 1% of the cost. The thing is in India, the direct currency conversion puts most higher quality tier blades at roughly half of this amount. That's how they are able to buy them and re-sale them on ebay to us! What I am getting at is that a large multi national corporation is paying in whatever the cheapest currency is where they are manufacturing, I highly doubt that the cost is double to manufacture in RU vs IN. Meaning they are probably closer to $2/100 or even less and the other costs involved are probably higher(distribution). The wholesale price should be roughly $4-$5/100 assuming retail of $12/100, that's a pretty standard retail markup. The parent company needs to be making roughly $1 in profit on a $4 sale so now we have $3. The part that I don't know is what distribution costs, if it's another $0.50 then there we are at $2.50, but's that's the total manufacturing cost not just materials(and is still too high to fit within the standard markup for the above mentioned Indian blades when purchased in IN and excludes any profit). If we were to say that blades cost roughly $1.50-$2.50 to manufacture, depending on the country, I think we would be getting closer to reality. So I think I am pretty close guessing $1 or less per 100 for materials!
 
Last edited:
Did you forget to multiply by 2 since there are two sides? I went and grabbed a blade since I am being too lazy for the amount of thought you are puting into this. On this, I would say 0.5mm X 37mm X 4 or 74mm^2 is more likely. The ground height of the bevel is this much so that is the maximum area that is sputtered with platinum. Remember, platinum is designed to strengthen the edge to prevent acidic degradation. This goes back to the 1960's patents. Logic also says that the whole blade isn't coated because you don't harden the entire blade otherwise it would become brittle.

So we are dealing with roughly 8% of what you got which seems even more likely, say 3 pennies per 100. On the economic front, modern manufacturing makes up around 20% cost of an item(or less). So $2.50/100 blades, we might be getting closer as that puts platinum at 1% of the cost. The thing is in India, the direct currency conversion puts most higher quality tier blades at roughly half of this amount. That's how they are able to buy them and re-sale them on ebay to us! What I am getting at is that a large multi national corporation is paying in whatever the cheapest currency is where they are manufacturing, I highly doubt that the cost is double to manufacture in RU vs IN. Meaning they are probably closer to $2/100 or even less and the other costs involved are probably higher(distribution). The wholesale price should be roughly $4-$5/100 assuming retail of $12/100, that's a pretty standard retail markup. The parent company needs to be making roughly $1 in profit on a $4 sale so now we have $3. The part that I don't know is what distribution costs, if it's another $0.50 then there we are at $2.50, but's that's the total manufacturing cost not just materials(and is still too high to fit within the standard markup for the above mentioned Indian blades when purchased in IN and excludes any profit). If we were to say that blades cost roughly $1.50-$2.50 to manufacture, depending on the country, I think we would be getting closer to reality. So I think I am pretty close guessing $1 or less per 100 for materials!
Did multiply by two sides - see bullet # 6. Excellent that you are taking the analysis forward.

My estimates based on coating the entire blade are worst case while the 2.5mm is a conservative assumption on partial coatings nearer the edge. Issue is that I haven't been able to find a definition of the actual edge area sputtered. Suspect reality is somewhere in between our two sets of assumptions. Did find an quote from someone who visited a Gillette plant in the U.K. years ago that a kilogram of platinum was enough to coat millions of blades that supports something in the range between our sets of assumptions. Based on what I saw for the cost of blade steel suspect you are right that total bill of materials should be under $1 for 100 blades. If you can find any specifications on exactly how much of a blade is coated we can go with it.

Margins could be slimmer than typical retail as blades are consumer packaged goods (CPG) so margins may only be in the 30-40% range. Found what appears to be an Astra razor blades corporate wholesale site that will sell a pallet of 30 cases of 100 boxes (100 blades each) for $16,500 or $5.50 each. Gives a margin of just over 30% on the current Amazon price of $8. Single cases of 100 boxes are $650 implying a margin of under 20%.

Another wrinkle is that since the 100 packs we buy are really wholesale inner packs from cases of as many as 5,000 to 10,000 blades I think there is a lot of old inventory out there that vendors are heavily discounting to move. Also for most manufactures DE blades are highly commoditized and not heavily marketed with most sales and marketing efforts going into much higher margin cartridge shaving systems. Just look at the Derby and Dorco catalogs and of course Gillette. All want to move consumers away from low margin DE blades to the shaving systems.
 
Did multiply by two sides - see bullet # 6. Excellent that you are taking the analysis forward.

My estimates based on coating the entire blade are worst case while the 2.5mm is a conservative assumption on partial coatings nearer the edge. Issue is that I haven't been able to find a definition of the actual edge area sputtered. Suspect reality is somewhere in between our two sets of assumptions. Did find an quote from someone who visited a Gillette plant in the U.K. years ago that a kilogram of platinum was enough to coat millions of blades that supports something in the range between our sets of assumptions. Based on what I saw for the cost of blade steel suspect you are right that total bill of materials should be under $1 for 100 blades. If you can find any specifications on exactly how much of a blade is coated we can go with it.

Margins could be slimmer than typical retail as blades are consumer packaged goods (CPG) so margins may only be in the 30-40% range. Found what appears to be an Astra razor blades corporate wholesale site that will sell a pallet of 30 cases of 100 boxes (100 blades each) for $16,500 or $5.50 each. Gives a margin of just over 30% on the current Amazon price of $8. Single cases of 100 boxes are $650 implying a margin of under 20%.

Another wrinkle is that since the 100 packs we buy are really wholesale inner packs from cases of as many as 5,000 to 10,000 blades I think there is a lot of old inventory out there that vendors are heavily discounting to move. Also for most manufactures DE blades are highly commoditized and not heavily marketed with most sales and marketing efforts going into much higher margin cartridge shaving systems. Just look at the Derby and Dorco catalogs and of course Gillette. All want to move consumers away from low margin DE blades to the shaving systems.
Sounds like we are on the same page, getting narrowed in. I actually though my area estimate was high based on measuring a blade that I have, but understand we need a patent reference.

I am prone to getting hyper focussed and missing something from the very beginning, you are completely right about the margins. The only thing I have some concern with is using Amazon prices as a basis, they are obviously undercutting the market. I mean some of these blades are $24/100 at reputable retailers! I just paid about $30/300 for ASPs, but that was directly from a lower economic tier country and I am taking advantage of their cheaper shipping. A more reasonable price is probably $12-$15, same with Nacets. Even then, I am not so sure. I would think 50% would be a more believable margin and that allows wiggle room for discounts.

The one interesting thing about India that I don't know much about, but all the blades have a listed price, implying that those prices are fixed and there is some price controls. I think it's the most transparent thing to look at that's why I mention it. I took another look and they are about $4/100 for high quality blades, not meant for export. India is still one of the lowest cost manufacturing locations and the fact that in many cases the money is staying in country, I could see the wholesale being $2/100. That's why I didn't think RU made could be more than $2.50-$3, accounting for additional costs. India must be even cheaper than I thought if Astras are $5.50.

I suppose we should leave this part of the discussion for now since you have the wholesale price. I think my original thought was that I have Indian blades that specifically say they have platinum sputtered edges so I can't image that is adding more than a few pennies considering their lower cost. I have read a few times that ordering Sandvik roles from Sweden definitely ups the price and puts the blades into a premium tier, I guess relative to the unknown source for the standard blades. I even have some blades that claim to use Japanese steel, I guess it's possible.
 
Sounds like we are on the same page, getting narrowed in. I actually though my area estimate was high based on measuring a blade that I have, but understand we need a patent reference.

I am prone to getting hyper focused and missing something from the very beginning, you are completely right about the margins. The only thing I have some concern with is using Amazon prices as a basis, they are obviously undercutting the market. I mean some of these blades are $24/100 at reputable retailers! I just paid about $30/300 for ASPs, but that was directly from a lower economic tier country and I am taking advantage of their cheaper shipping. A more reasonable price is probably $12-$15, same with Nacets. Even then, I am not so sure. I would think 50% would be a more believable margin and that allows wiggle room for discounts.

The one interesting thing about India that I don't know much about, but all the blades have a listed price, implying that those prices are fixed and there is some price controls. I think it's the most transparent thing to look at that's why I mention it. I took another look and they are about $4/100 for high quality blades, not meant for export. India is still one of the lowest cost manufacturing locations and the fact that in many cases the money is staying in country, I could see the wholesale being $2/100. That's why I didn't think RU made could be more than $2.50-$3, accounting for additional costs. India must be even cheaper than I thought if Astras are $5.50.

I suppose we should leave this part of the discussion for now since you have the wholesale price. I think my original thought was that I have Indian blades that specifically say they have platinum sputtered edges so I can't image that is adding more than a few pennies considering their lower cost. I have read a few times that ordering Sandvik roles from Sweden definitely ups the price and puts the blades into a premium tier, I guess relative to the unknown source for the standard blades. I even have some blades that claim to use Japanese steel, I guess it's possible.
It's possible that in India some manufacturers, like Dorco, are using DE blades as loss leaders to further penetrate the market with the goal of then shifting consumers to the higher margin shaving systems that they are really focused on. For example 60 Klass blades (rebranded Dorco Titan STL300s) have sold for 120 rupees or $1.46 U.S. at Joie Mart on-line. Equivalent to $2.43 for 100 blades that are made in the same Vietnam plant as the Dorcos we buy in the U.S. Marketing message is blades based on superior Korean Technology. Bet that is below cost and an effort to build market share.

Agree we are on the same page on the physical analysis. Did some more research and it's likely that just the edges of the blades are coated so somewhere between our 2.5 mm to 1.0mm estimate is correct. Probably closer to the 1.0mm figure. It looks like the sputtering process has the blades stacked in bayonets so likely it's primarily the edges that get exposed to the coating materials though I still couldn't find anything that confirms the exact part coated as part of a specification.

Using a 37mm actual cutting edge length and a 1.0 to 2.5 mm edge coating width as a range our calculations evolve to the following:

  • 37mm x 1 to 2.5mm x 4 edge sides = 148 - 370 mm^2
  • Times a 50 angstrom or 0.000005 mm coating depth: 148-370 x .000005 = a platinum volume of 0.00074-0.00185 mm^3
  • Multiply above by estimated volume by density per mm^3 or: x .02145 = .0000159 - .0000397 grams/blade
  • Times 100 for platinum in a 100 blade box gives us 0.00159 - 0.00397 grams of platinum/100 blades
  • 0.00159 - 0.00397 grams of platinum represents 0.0032% - 0.0079% of the 50 grams of metal in a 100 blade pack.
  • At $32 per gram this represents 5 - 13 cents of platinum in a 100 blade pack.
Even more compatible with our latest blade economics analysis. At current blade prices manufacturers could not afford much more of the metal. Per these estimates a kilogram of platinum would cover between 25 and 63 million razor blades.

Further affirms that SEM analysis can identify the presence of an element and will overstate the coating material percentages that are prevelant near the surface of a razor blade.

How It's Made Video on Razor Blades


Sputtering Machine Info Link

 
Top Bottom