Inverse Temperature (1000/K)

[Michael]

Hi All,

I am reading ahead for a microfab class I am taking in January and I've
noticed alot of the literature publishes graphs that characterize a given
phenomenon as a dependency on temperature and thus the x-axis is scaled in
inverse temperature like 1000/K (K^-1). In some cases the top of the graph
is scaled in ascending Celsius to assist in the graphs interpretation.
However my primary question why do the authors use inverse temperature
rather simply scaling the x-axis using Celsius or Kelvin directly?

I read some brief blurb on the Internet that inverse or reciprocal
temperature is more meaningful than its counterpart. But why?

Any info is always appreciated.

--Michael

 

[redbelly]

Hi All,

I am reading ahead for a microfab class I am taking in January and I've
noticed alot of the literature publishes graphs that characterize a given
phenomenon as a dependency on temperature and thus the x-axis is scaled in
inverse temperature like 1000/K (K^-1). In some cases the top of the graph
is scaled in ascending Celsius to assist in the graphs interpretation.
However my primary question why do the authors use inverse temperature
rather simply scaling the x-axis using Celsius or Kelvin directly?

I read some brief blurb on the Internet that inverse or reciprocal
temperature is more meaningful than its counterpart. But why?

Any info is always appreciated.

--Michael

Some phenomena are inversely proportional to absolute temperature.
Making a plot vs. 1/T is just a way to get a straight line, so that:
1. It's easier to read off the values on a graph, and
2. The straightness of the graph demonstrates how well the
inversely-proportional behavior is followed.
2a. Any slight variation from the inversely-proportional behavior is
easily apparent by the graph not being perfectly straight.

Mark

 

[Mike]

Hi There,

Thanks for the response. Makes sense now! I am curious though, why 1000/T
versus 1/T? Alot of these plots, like diffusion coefficients for various
elements in Si use 1000/T. I suspect it follows along the same lines ...
making it easier to make relative temperature comparisons using larger
decimals values (i.e. 1000 deg C is 1273 K so 1/K is 7.8E-4 as compared
with the more friendly 1000/K which would be 0.78 in this case).

--Michael

 

[redbelly]

Hi There,

Thanks for the response. Makes sense now! I am curious though, why 1000/T
versus 1/T? Alot of these plots, like diffusion coefficients for various
elements in Si use 1000/T. I suspect it follows along the same lines ...
making it easier to make relative temperature comparisons using larger
decimals values (i.e. 1000 deg C is 1273 K so 1/K is 7.8E-4 as compared
with the more friendly 1000/K which would be 0.78 in this case).

You're exactly right. Using 1000/T, the axis values are numbers close
to 1. Just like a plot of electrical current might use mA or uA
instead of Amps, when the current being plotted is in the mA or uA
range. (There, at least we're somewhat on-topic now.)

Mark