🚀 go-pugleaf

Status #8 Cell 4A3

To the non-experts out there, please don't copy this data down as if it were 
the gospel.  To the experts, please look at this and tell me if there is a 
reason to do it right i.e. with about 200 grams of Pd that has been properly 
outgassed.  

The experiment is presently in the "watching paint fade" stage.  At about 400 
hours into the run, there is nothing to see except small changes (one sigma or 
less) due to outside temperature variations.  Since I would like to present 
something about the experiment every few days, I have spent a little time 
looking at the charging data.  (Oh! for a student.)  The 1 cm x 1 cm x 1 mm 
cathode was charged over a period of 21 hours by a current ramp which 
increased from 1 ma to 46.9 ma over that time.  The hourly data is presented 
below: 

JSUM        Gas         D/Pd        G*26.34     Diff/G      K 
    0       0           0             0          0
    +.6     0.02        0.000        -0.53    -28.2
  -16.0     0.59        0.009       -15.54      0.4         37
  -33.4     1.18        0.018       -31.08      1.0         92
  -56.2     1.89        0.030       -49.78      1.7        156
  -45.1     3.05        0.048       -80.34      5.8        533
  -52.5     4.41        0.070      -116.16      7.2        662
  -71.1     6.27        0.099      -165.15      7.5        689
  -85.1     8.61        0.136      -226.79      8.2        754
 -130.1    11.31        0.178      -297.91      7.4        680
 -197.4    14.66        0.231      -386.14      6.4        588
 -283.2    18.41        0.290      -484.92      5.5        505
 -375.2    22.41        0.353      -590.28      4.8        441
 -479.8    26.73        0.422      -704.07      4.2        386
 -597.9    31.43        0.496      -827.87      3.7        340
 -714.6    36.41        0.574      -959.04      3.4        312
 -859.5    41.84        0.660     -1102.07      2.9        266
-1003.1    46.53        0.734     -1225.60      2.4        220
-1129.9    49.41        0.779     -1301.46      1.7        156
-1216.1    50.66        0.799     -1334.38      1.2
-1279.7    50.37        0.794     -1326.75      0.5
-1332.2    50.40        0.795     -1327.54     -0.0

The first two columns are raw data.  JSUM is the integrated balance in 
joules.  The minus sign means that extra heat had to be added to keep the 
calorimeter in balance.  Gas is the cc of gas accumulated in the syringe.  The 
rest of the columns are my computations and subject to horrible errors.  The 
measured one sigma calorimeter drift at the start of this run is of order 5 
mw.  Actual measured peak to peak drift in the 10 hours before the run started 
was 9 mw.  (Dieter, I know you have noted my 35 mw one sigma, but this is for 
making a large temperature change coming back to an old balance point - and 
likely turned out to be optimistic.)  This means that the error of the JSUM 
measurement is time dependent.  This would give a one sigma error after one 
hour of 18 joules, and for the last entry at hour 21 of 378 joules.  The gas 
measurements are somewhat better, the servo resolution is 1/70 cc, and the 
balance is always good to a few steps.  There should be a temperature 
correction proportional to the gas in the syringe (the gas in the calorimeter 
is at constant temperature) and the ambient temperature.  It has not been 
made.  There are also unknown barometric pressure changes, but there was no 
violent weather during the test period.  The peak to peak ambient temperature 
variation over this run was 0.9 C.  There is also a syringe non-linearity that 
has not been thoroughly investigated but which I have checked at a few points 
to 1% with a 1% gas burette.  All in all, I would put a limit on the error of 
the gas measurement at 2% of the reading. 

What we would like to get from this data, is a measure of heat of absorption 
of Palladium/Deuterium as a function of the gas loading.  There are problems.  
First there is a calorimeter servo time lag.  Even though we tried to charge 
very slowly, the calorimeter thinks a while before it reports the balance.  
This makes the last few points particularly suspect where the conditions are 
switching from the Palladium absorbing nearly all of the Deuterium to none of 
it.  Since this results in a real change of operating power level, the servo 
must compensate, and so lags behind.  Next, there is the measurement problem 
due to the known errors mentioned above.  We note that these errors could be 
reduced by a factor of up to 500 if we just run a larger Palladium cathode.  
We could also reduce the servo errors by charging somewhat more slowly.  But 
this is an "anomalous heat" run and was not optimized for measurement of the 
heat of absorption. 

The major problem with this run is that it has an unknown D/Pd starting point.  
Since the first try at starting this run was aborted by a gas leak, it is 
quite possible that the starting point was not 0.0 D/Pd but more like 0.1.  We 
attempted to remove the gas from the aborted start by running the cell weakly 
backwards, but with an unknown result.  There is also the problem that our 
Palladium "coin" bar has an unknown gas absorption history.  

The G*26.34 column is the joules lost per cc because the measured oxygen did 
not recombine to form water.  This is derived from the 1.53*I energy for dis-
association, 96,484 coulombs per mole, two atoms per D2, and the fact that we 
are measuring O2 not D2 giving another factor of two.  Someone please check as 
I always worry that I will mis-count charges or something. 

The Diff/G column is the difference between the joules per cc of un recombined 
oxygen and the calorimeter balance, divided by the absorbed D (2*observed 
oxygen) to get the average joules per cc of D absorbed integrated up to the 
indicated D/Pd loading.  The plus sign for most of this data indicates that 
the absorption of D by Palladium releases heat.  

The K column converts the Diff/G column to kilocalories per mole D2 absorbed 
at the indicated D/Pd loading.  I have not computed all the entries in this 
column where the error limits make the measurement meaningless.   

It seems to me that there is a hint in this data that much of the heat of 
absorption comes from the first 0.13 or so of D/Pd (or 0.23 if this sample 
started at the suspected 0.1).  After that, less and less heat is released per 
unit absorbed.  Is there a thermodynamic reason for this?  I have F. A. Lewis, 
"The Palladium Hydrogen System", can someone point me to the right curve for 
comparison?
                            
Tom Droege