PRINTED! circuit boards

Laser induced copper deposition from a liquid precursor on different substrates, including 3D printed parts.

Similar projects worth following
A modern PCB prototyping process for a more more civilized time.
Copper(II) formate (the copper salt of formic acid) is know to decompose into CO2, water and solid copper crystals when heated since long. Laser and flash lights (photonic curing) as heat source is used since at least 30 years and till today under active research e.g. for conductive inks, maskless copper interconnects etc.
The required chemicals are safer then whats typically used for etching and the process by far less disturbing than milling.
This project should help to make this available for DIYers, Hackers and Makerspaces as the proposed tools and materials are widely available.

The goal is to make hardware prototyping easier by providing a middle ground between dealing with either developer and etching chemicals, loud mills or long shipping times from china. Ideally a medium dense single layer 4x4" PCB for SMD components can be manufactured in less than an hour with little to no attention.

Status Quo:

"Printing" a single layer of copper on a PET-G works and conductivity 2-3 orders of magnitude better than conductive plastic, but 1-2 orders of magnitude below that of solid copper, is achievable.

Component mounting still needs to be solved, low temp solder seems to promising, alternatively conductive epoxy.


Copper(II) format preparation

I mixed 12g Copper carbonate with 126ml of DI water and 10ml of 75% Formic acid in a glass jar. The reaction is slow and takes a few hours at room temperature. CO2 is released during this process, do NOT close the jar tightly. Use the usual safety gear (googles, gloves) and keep the room ventilated.


I'm using a 900mW 405nm laser diode, probably extracted from a Pioneer BDR-209 BR recorder. The driver limites the current to 200mA  (about 300mW power) and PWM is used to limit the power further while printing.

450nm seems to work as well, but requires significantly more power, it might be difficult to use on heat sensitive substrates. I stooped using it once I had a 405nm laser available.


  • Glass (microscope slides): very easy to work with, can't be destroyed by the laser. Adhesion is limited.
  • Polyimide (on glass): somewhat sensitive to power. Better adhesion than glass.
  • Fiber reinforced epoxy (FR4): sensitive to power, easy to burn. Difficult to wet, sanding the surfaces helps but exposes porosetiy.
  • white PET-G: very sensitive to power. Easy to wet with window cleaner. Quality traces and conductivity achievable with multiple layers (3 typical)
  • black PLA: didn't work, maybe because black
  • TPU (flexible): didn't work

Sample preparation:

I'm using a syringe to apply small quantities of saturated Copper(II) format solution about (0.1ml - 0.5ml per square inch). Usually a wetting agent is needed to overcome the surface tension. Dishwasher detergent works quite well, alternatively window cleaner as it contains ethanol, particularly good with PET-G. The samples are usually dried with the heated bed set to 75°C.

Power Control:

I'm using Repetier Firmware, but not in Laser mode as it only allows a fixed PWM uncoordinated to the actual gantry. Instead I'm using the E parameter of the G1 command to control the energy that is applied to each segment. The principal is fairly simple, the E axis is set to 15625 steps per mm with a maximum speed of 1mm/s. A Arduino Nano is  used to broaden the step pulse to 64µs (1/15625). This creates a PWM signal coordinated to the actual gantry speed.


  • Component mounting is unsolved. PET-G might actually work with Tin Bismuth solder (138°C)
  • System (pump/needle) to apply the liquid precursor selectively only where needed
  • Better integration with 3D printing, e.g. multiple layers, add features to help with component placement

  • Closeups

    rolmie03/24/2018 at 20:46 0 comments

    The dimensions are about 2.5cm x 1.25cm (1"x0.5"). The features and footprints are well 'hidden' in somewhat oversized traces, but this is an easy way to reduce resistance. 

    Printing strategy is to print as many "shells" as possible, as I suspect that conductivity along the printed lines is better than across. Additionally each layer is cured twice, interleaving the lines. The laser spot is about 0.1mm in diameter, the feedrate is 33mm/s. Curing a single layer at this size takes about 4:30 minutes.  

    A well hidden SOT-23 footprint.

    A 0805 foot print for a resistor:

    A "defect", reason unknown

  • Explaining the samples

    rolmie03/24/2018 at 18:33 0 comments

    substrate: white PET-G
    precursor: saturated Copper(II) formate solution, with about 5% (by weight) glycerol
    process: window cleaner as wetting agent, 3 layers with 50% power increase per layer

    The resistance measured from bottom-left corner to top-right is below 1 Ohm. The dimension of the sample are about 2.5cm x 1.25cm (1"x0.5").

    substrate: glass (microscope slide)
    precursor: saturated Copper(II) formate solution
    process: dishwasher detergent as wetting agent, 2 layers

    Glass seems to be the easiest of all substrates, it is impossible to irradiate with too much power.
    Tin bismuth solder (melts at 138°C) seems to work (limited experience). The footprints are SOT-23 and 0805 mostly, traces are about 10mil.

    substrate: polyimide (Kapton) on glass (microscope slide)
    precursor: saturated Copper(II) formate solution
    process: dishwasher detergent as wetting agent, single layer
    The distance between the traces in the center is about 0.5mm (20mil), based on that the trace width is estimated as 0.15mm (6mil). Lead Tin flux dissolves the copper.

    substrate: fiber reinforced epoxy (FR4)
    precursor: saturated Copper(II) formate solution
    process: dishwasher detergent as wetting agent, single layer

    Tin Lead solder doesn't work.

View all 2 project logs

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates