Research Update : Fringe Intensity vs. Water content in Nafion

1
Research Update Fringe Intensity vs. Water
content in Nafion

• Jungik Kim

2
Performance and Degradation of a Fuel Cell

• Performance depends on Temperature, Water
  content, and pressure
• Degradation can be regarded as unrecoverable
  performance loss
• T, RH, and P should be monitored while operating
  a fuel cell to maximize the performance and
  understand the degradation mechanism

3
Approach

• T and RH measurement by optical method
• Refractive index of Nafion is dependent on its
  density which varies with T and RH changes
• Changes in refractive index can be measured by
  interferometry
• Localized and in-situ measurement of T and RH is
  possible

4
Experimental Procedure

• Construct a data base on the refractive index of
  Nafion as a function of T and RH
• Measure changes in refractive index during fuel
  cell operation
• Measured refractive index changes can be used to
  estimate the changes in T and RH during fuel cell
  operation

5
Interferometry with T Variation
Sample
w
w
Glass ng
t
Dried Nafion115 np

• Sample cooled from an elevated temperature by
  natural convection at room temperature
• During the cooling process, fringe pattern
  recorded

6
Interferometry with T Variation

• Exponential decrease of T
• dn/dT -10-4 / C
• T profile is extracted from measured fringe
  intensity

7
Interferometry with RH Variation

• A wet membrane was dried by blowing dry oxygen
• During the drying process, fringe pattern
  recorded

8
Expectation
RH and n
Intensity
Time
Time

• Assume Refractive index is proportional to RH
• The slope would correspond to the frequency of
  the intensity oscillation

9
Reality

• Intensity seems to vary randomly during the
  measurement of 1 minute
• Cant fine sinusoidal repetition
• The interval between each measurement was limited
  to 5 sec, which is slower than the time scale of
  intensity variation

10
Effect of Low Sampling Frequency
Intensity
Intensity
Time

• Though the original signal is sinusoidal, low
  sampling freq. makes the measurement look random

11
Statistical Analysis

• RH of the sample will reach steady state
  eventually and then the intensity wont change
  any more

• Standard deviation may be meaningful

12
Statistical Analysis

• Standard deviation taken over time period of 60
  sec(12 data points) averaged over positions(1300
  pixel)
• Averaged standard deviation decreases with time

13
Simulation on Statistical Analysis
Averaged standard deviation
Mean frequency of intensity oscillation (Hz)

• Artificially added noise in frequency prevents
  ASD from reaching zero
• Until 0.01 Hz, ASD is more or less constant
• When the oscillation is synchronized with the
  sampling frequency (1, 0.8, 0.6, ... Hz), small
  ASD can be measured.

14
Consideration

• The effect of RH in the sample on the
  interference intensity variation was demonstrated
  by calculating averaged standard deviation
• However, it is difficult to establish
  quantitative relationship between ASD and RH
  variation
• High speed camera is necessary

15
Testing MEA with Electrospun Electrode

• 1 cm2 MEA made
• Highest OCV observed was 0.79 V