No announcement yet.

Getting a shiny surface after etching

  • Filter
  • Time
  • Show
Clear All
new posts

  • Getting a shiny surface after etching

    Hi Folks,
    I’m an artist who’s working on manufacturing a piece I’ve developed that uses fine lines etched into metal to reflect light in specific directions. Here’s a demonstration:

    The issue I’m running into is that I’ve produced my original prototypes on a CNC, where it takes 10-30 hours, making it prohibitively expensive. I commissioned a photochemical machining shop to make me some samples using chemical etching (ferric trichloride), which sort of works, but they’re around 5 to 10 times less reflective than the CNC etched samples. I think this is because of the microscopic texture of the metal post etching - matte rather than shiny. I’ve tried stainless steel and brass, and tried gold plating as well. The etch lines are 0.1mm wide and around 0.7mm deep.

    My questions are:
    1. I have some flexibility on what metal I use. Is there some combination of metals and etching process that produces a shiny surface?
    2. Is there some form of post processing that can make the surface shiny? Say electroplating. Obviously the challenge here is not destroying the etched texture.
    3. If not, is there some other manufacturing method that might work?


  • #2
    *That should be 0.07mm deep.


    • #3
      Hello Rahms. I saw your video. It's beautiful, did you make this?

      The reflectiveness of metals mostly comes from a smooth surface on a microscopic level. when you etch the metal, you change the microstructure of the surface, because the grain boundaries of the metal is usually more corroded than other regions. so your answers:

      1) probably no. however pure metals are better for your application than alloys. because in pure metals there are no different phases with different corrosion resistance and the etched surface is going to be more even.

      2) you can etch the copper with the same method you used, but you need to brighten the surface afterwards. you have few options for this:

      - bright dip the copper. you can buy commercial ready to use solutions to brighten the copper or brass substrate. or you can make it yourself.
      - plate some other metal on the substrate after etching, with leveling and brightening additives (eg bright nickel), but this will change your dimensions for about 20-100um.

      3) can you draw a schematic of the cross section you want to make?


      • #4
        I did - thank you.

        1) Thanks for the tip on pure metals. I was looking to understand the physical reason for the matte texture so this is helpful. I'm willing to make many samples to find something that works. Off the top of my head some metals that might be worth trying are copper, nickel, aluminum, zinc, tin, titanium(?). Are these all etchable through conventional methods?

        1.a) If I do use copper (it is attractive when shiny) what kind of seal do you suggest to keep it bright? I've covered some samples in epoxy and that seems to work, but it might be overkill and I imagine there's a better way.

        a) Brightening treatments: Bright dip is an interesting option. This might be a silly question - but this seems like it removes imperfections via an acid solution. Why does the acid in this process make the metal brighter while the one in etching makes it more matte? Are there similar options for the metals I mentioned other than copper?

        b) Plating: I tried gold plating two of the pieces, and surprisingly they might have been the least reflective of the lot. I'm not sure if it's a fundamental issue with gold plating or just the specific treatment that was used. I didn't know about leveling and brightening additives. Are there any specific products I can refer the shop to? A big part of the challenge right now is that I'm working with a shop in China that doesn't speak english that well, so I'll need to be very specific.

        Electropolishing - do you know if the result is usually more matte or more shiny?

        3) For cross section, do you mean the shape of the ideal etch trench? To be honest I'm not sure: my microscope wasn't able to resolve many details about the diamond etch lines. (See image below). I'm considering paying for 3D profilometry on my samples so I can see what's going on.

        Click image for larger version

Name:	Screen Shot 2021-01-11 at 12.06.55 PM.png
Views:	20
Size:	763.8 KB
ID:	144745


        • #5
          Hi Rahms

          1) yes all the metals are etchable. but it is better to use a metal which is studied enough, so you can find data on the processing. (and the prices might be important as well)

          aluminum and titanium form a protective film on the surface in corrosive environments. so its a little harder to etch them. but its not impossible. (you can dissolve aluminum and probably titanium in strong alkaline solutions. but it release toxic gases, be careful)

          (a) you can protect the shiny surface from corrosion by different methos:

          - you can use inhibitors such as benzotriazole dissolved in solution. then dip your parts in that solution and let it dry. (or you can spray it)
          - you can do a conversion coating such as chromating. it gives a color a little bit more reddish than copper itself, but it's beautiful.
          - you can use clear paints, like epoxy as you said, or lacquer.
          - you can plate a thin layer of a corrosion resistant metal, such as gold, palladium or rhodium on the surface. this plating will be as bright as the substrate, not more not less.


          (a) it's not silly. you are sharp man

          when you etch a metal you may have different goals.

          - sometimes you just want to remove the metal, as fast, cheap and convenient as possible, and the surface finish is not important. like etching of copper for PCB manufacturing.

          - sometimes you actually need to change the metal surface, so you can study it. you etch some phases, and keep some other, so different phases can bee distinguished under the microscope. this method is used in metallography techniques.

          - and in your case, you just need a shiny surface, so contrary to PCB etching, you don't want to dissolve the metal, you just want to brighten it.

          different formulations are developed based on each demand. so it's no surprise that they have different effect on the finish.

          (b) about plating

          gold is not a good choice for this. because gold platings are almost always very thin (from less than a micron to few macrons). so it can not cover up the imperfections.

          Bright nickel plating is a better option as it can easily get as thick as over 100 um without cracking.

          the additives are formulated for each plating bath, so you should first know your electroplating bath. its better to tell your supplier for a bright nickel plating, and not the additives itself.

          please note that the plating is going to change the geometry of your surface. you may plate too thick, and loose all the precious notches you made!

          (c) electropolishing:

          the concept is similar to a bright dip bath, except you use a power supply instead of oxidizing agents (generally speaking).
          electropolishing and chemical polishing can give you a very shiny surface. some of them may produce a matt layer that needs to be removed later (called desmuting process). but the surface will be flat in microscopic level. (you ever seen the gigantic mirrors in telescopes? they are aluminum sheets, electropolished and anodized, most reflective thing made by humans, ever)

          However, there is a bug but involved:

          BUT you can not safely use chemical polishing as it may dissolve the notches as well as imperfections. keep in mind that your beloved notches are a kind of imperfection in the eye of the polishing process

          3) I actually want to know the exact shape you want. the microscope and profilometry will only show what you currently have. if it is the exact thing you want, then yes it is going to help.


          I have another idea. you can paint the shape of notches you want on the surface of a bright metal sheet, and then electroplate the other regions to come up and form the notches you want. the notches will be as deep as the thickness of your paint, and do not need any post processing.

          a pic similar to what i mean from Wikipedia
          Attached Files
          Last edited by yarim; 01-12-2021, 06:23 AM.


          • #6
            A good idea came to my mind

            grab a piece of aluminium sheet, and engrave the (negative) pattern you want with CNC like you did before.

            then use this aluminium part as a mandrel for electroforming.

            you plate a thick layer of metal (copper and nickel are most used metals in electroforming) on the aluminium part, and then you detach the aluminum from freshly formed part. you can plate as thick as you want.

            (the same concept is used in copying vinyl disks if you are familiar with)

            the only thing different than your previous CNC is that you should engrave the negative pattern. the notches will be walls, the walls will be notches.

            I attached a pic from a quick google search. note that you don't need the last step

            the good point about this method is that you get a part with exactly same finish as your mandrel. and you can use your mandrel as many times as you want


            • #7
              Very helpful, thanks.

              Below is the ideal surface I'm looking for - a series of V shaped grooves that are a bit less than 90 degrees. I'm not sure how much the diamond etched samples actually resemble this - I'm hoping the profilometry can help answer that. I've made them on my home machine and as you can see from the microscope picture it's certainly a little messy. If I'm going to invest in one master copy that then gets replicated might there be a better way to make it? (At a minimum I'd have it professionally made on a high end mill.)
              Click image for larger version

Name:	grooves.png
Views:	10
Size:	21.9 KB
ID:	144759

              Electroforming is an interesting option. Can it help make the negative too? (As in electroform a negative then electroform the positive on that.) Since I'm using a diamond engrave bit rather than an any kind of end mill I'm not sure I can use the same CNC process to create the negative.

              Attached Files


              • #8
                Hi Rahms

                Your CNC sample might look something like what you want, but the etched sample is definitely nothing similar. the etching solution do not create v shaped groove. it dissolves wherever it can reach. So even your printed lines will have a "under cut" in some degree (depending on your dimensions). So I think you should forget about etching.

                You can use the machined master model to produce electroformed replicas. I suggested Aluminum or Titanium, because they have a natural oxide layer on the surface that helps you detach the replica easily and without rupture.

                So if you use the machined mandrel to electroform the master model, then you may have adhesion issue, as the same metals usually sticks together.

                you can do this, but after making the master model you should plate it with a hard to stick layer. I have no experience but a thin layer of chromium may do the trick as it has a natural chromium oxide on the surface (that is why stainless steel is corrosion resistance). But you can not plate chromium directly on copper. so nickel electroform may be a better choice.

                Then you have these two options:

                1) machine a negative master model out of aluminum or titanium (or even, electroform copper on it, detach the part.

                2) machine the model from Al or Ti, electroform Nickel onto model, detach the nickel master model, plate a layer of chromium on the master model, use it to electroform copper parts.

                your schematics does not show the dimensions, but note that a layer of chromium is going to be at least 1um. see how much is it going to alter the geometry.