Copper color transition temperature 2

[rphill12]

Does anyone know the temperature at which copper will begin to change colors (i.e. blue, violet, and even gray.. I think gray is when the oxide formation happens, but I'm not positive).

I'm really interested in knowing when the gray forms since it seems like that would be the highest temperature.

I am trying to do a failure report on a material and I can't seem to find this information anywhere.

Thank you!!

 

[TVP]

There is some information in a previous thread here at Eng-Tips:

thread330-38947

I recall that there was an article in Industrial Heating magazine by Daniel Herring (The Heat Treat Dr.) that discussed the color change of copper during heating, but I don't have the specifics. Perhaps someone else will be able to provide the details. Calling kenvlach...

 

[kenvlach]

rphill12,
It isn't simply a matter of temperature. The variables include time, copper purity/alloying and initial surface condition (texture & coatings), method of heating, atmosphere during heating, etc. For example, I found a study which studied the oxide formed on electropolished 99.999% Cu single crystal at 200 °C in 1 atm pure oxygen. Like yours?? You can correlate color to the oxide thickness (see below), but there will be an entire curve of time vs. temperature yielding a given thickness. As described by TVP in earlier thread, the oxidation weight gain (oxide thickness is proportional) is given by
W² = kt,
with t = time,
and k = rate constant for a given T & P_O2.
More generally, k varies with temperature and oxygen concentration, so
k = k_o (P_O2)^1/n e^-Ea/RT

On bare metal, the initial oxidation rate is rapid and linear, then slows to the diffusion limiting case. At low temperatures, the oxide may be amorphous, with very slow kinetics. With heating, amorphous oxides crystallize and the rate becomes more rapid. Although the equation has the same form, different terms apply above the transition temperature. The oxygen pressure exponent depends upon the mechanism of the reaction; (often ½) and is experimentally determined.

The copper hasn't changed color, but its surface has oxidized. If metallic copper is still visible through the oxide film, you see the so-called 'interference colors' which are a function of film thickness. When light hits the oxide, a portion reflects off its top surface and the remainder passes through to the metal and reflects. This portion of the light will be varying amounts out of phase with and thus interfering with light from the top surface. The observed light is the combination. The amount of interference varies primarily with oxide thickness and wavelength of the light (there is also some refraction). For simplicity, white light is always used, resulting in a rainbow-like variation of color vs. thickness.

Some figures with color vs. thickness:

http://www.anvilfire.com/FAQs/temper_colors_hardness.htm

[oxide on steel]

http://www.nasascale.org/howtos/termpering.htm

[oxide on steel]

http://www.ee.byu.edu/cleanroom/color_chart.parts/sin_chart.jpg

[Si₃N₄ on Si]

Some photos of Titanium color anodizing (oxide formed electrolytically rather than by heat), from Russamer Lab (

http://www.russamer.com):

The colors formed on copper as a function of thickness, from Constable, Proc. Royal Soc. A, vol. 117, pages 376+ (1928) and Miley, Nature, vol. 139, p. 283+ (1937):[tt]
color thickness
Dark-brown 370-380 Å
Red brown 410-420 Å
Dark purple 450-560 Å
Dark violet 480-485 Å
Dark blue 500-520 Å
Pale blue-green 830 Å
Pale silver-green 880 Å
Yellowish-green 940 Å
Yellow 980 Å
'old gold' 1100 Å
Orange 1170-1200 Å
Red 1240-1260 Å [/tt]

I don't recall ever seeing gray copper oxide. It may be the start of opaqueness, with a thin black CuO outer oxide atop a pinkish-red Cu₂O layer. It takes a pretty high temperature (e.g., 400 °C) to form a visible black layer.

Is the oxide translucent or opaque? Please give any information on the copper, its initial condition (any anti-tarnish treatment or varnish), shape and finish, and the suspected temperature, time and atmosphere of heating.

We will try to figure something out, but likely an accurate answer requires experimentation. Put several identical samples in a lab oven, heat to a given temperature, and remove samples at different times. If resulting oxides seem too thin or too thick, repeat with higher or lower temperatures.

For further study. Searching the Journal of the Electrochemical Society (1955-date) for 'copper + oxidation + kinetics' yields 68 articles. The following 5 seem most relevant. Links are to the abstracts.

Oxidation of Copper between 250° and 450°C and the Growth of CuO “Whiskers”
E. A. Gulbransen, T. P. Copan, and K. F. Andrew
J. Electrochem. Soc. 108, 119 (1961)

http://tinyurl.com/3d7tzw

Influence of Oxygen Pressure on the Oxidation Kinetics of Copper in Dry Air at Room Temperature
P. K. Krishnamoorthy and S. C. Sircar
J. Electrochem. Soc. 116, 734 (1969)

http://tinyurl.com/2ptqb9

In Situ Rapid Thermal Oxidation and Reduction of Copper Thin Films and Their Applications in Ultralarge Scale Integration
Yao Zhi Hu, Rahul Sharangpani, and Sing-Pin Tay
J. Electrochem. Soc. 148, G669 (2001)

http://tinyurl.com/35cxuk

In Situ Studies of the Initial Atmospheric Corrosion of Copper Influence of Humidity, Sulfur Dioxide, Ozone, and Nitrogen Dioxide
T. Aastrup, M. Wadsak, C. Leygraf, and M. Schreiner
J. Electrochem. Soc. 147, 2543 (2000)

http://tinyurl.com/2k7stg

In Situ Observation of Dissociation, Microchannels, and Duplex Layer Formation in Cu₂O Scales Grown at 300°C
Bülent Önay
J. Electrochem. Soc. 136, 1578 (1989) [no abstract]

Oxidation of Oxygen-Free High Conductivity Copper to Cu₂O
J. P. Baur, D. W. Bridges, and W. M. Fassell, Jr.
J. Electrochem. Soc. 103, 273 (1956)

http://tinyurl.com/2hmwk9

 

[rphill12]

Thank you very much for taking the time to both write that up and to research additional information.

The copper in question is a laminated layer on a printed circuit board so the environment is not controlled to any type of lab standards. Next time I will be specific, but the information you provided is still very helpful!