23.05.2014 Views

Nuclear Spectroscopy

Nuclear Spectroscopy

Nuclear Spectroscopy

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

SUPPLIES<br />

• NaI(Tl) detector with MCA<br />

• Radioactive sources: 22 Na, 54 Mn, 60 Co, 65 Zn, 88 Y,<br />

133<br />

Ba, 134 Cs, 137 Cs, 207 Bi<br />

SUGGESTED EXPERIMENTAL PROCEDURE<br />

1. Start the MCA and check the calibration using<br />

the 137 Cs source. The observed photopeak should<br />

be centered near 661.6 keV. If it is not within 10<br />

keV of the accepted value, you may need to<br />

recalibrate your system.<br />

2. Place another source on the fourth shelf under<br />

the detector and acquire a spectrum with the full<br />

vertical scale at 1,000 counts.<br />

3. Select a Region Of Interest (ROI) around your<br />

photopeak. Details for doing this are in Appendix<br />

A. Record the centroid of the photopeak and<br />

its FWHM. This is available to you by displaying<br />

the Peak Summary, or for the ROI where the<br />

cursor is located, by selecting the Display Region<br />

button to the right of the displayed spectrum.<br />

4. Repeat step 3 for all the sources available to you.<br />

You may use all the data from Experiment #2 to<br />

avoid retaking spectra.<br />

DATA ANALYSIS<br />

From your data of photopeak energies and corresponding<br />

FWHM calculate the fractional resolution<br />

and graph its square as a function of the inverse of the<br />

corresponding photopeak energy. Use a linear least<br />

squares program to get the equation for the best<br />

straight line fit to your data. Graph this equation on<br />

the graph with your data. Is your fit to the data a good<br />

fit?<br />

Another way to test the theoretical explanations for<br />

the fractional resolution is to graph the natural logarithm<br />

of the fractional resolution as a function of the<br />

natural logarithm of the corresponding photopeak<br />

energy. Obtain the slope and intercept from a linear<br />

least squares fit to the data. From equation 3.1,we get<br />

the following<br />

Gamma Intensity<br />

40<br />

30<br />

20<br />

10<br />

0XXXX<br />

5%<br />

7.5%<br />

10%<br />

400 450 500 550 600<br />

Gamma Energy (keV)<br />

Figure 3.2 The overlapping spectra above are of two equalarea<br />

photopeaks centered at 477 keV and 511 keV. The<br />

individual photopeaks are shown at 10% resolution underneath<br />

the other spectra. The combination of the two photopeaks<br />

is shown for resolutions of 5%, 7.5%, and 10%.<br />

which predicts the slope to be -0.5. How well does<br />

your experimentally determined slope compare to<br />

the expected value of -0.5?<br />

NE ( ) =<br />

N<br />

0<br />

2πσ<br />

2 2<br />

−( e E − E m ) / 2σ<br />

2<br />

original peaks<br />

This is the equation for the normal (Gaussian)<br />

distribution with a peak area of N 0<br />

. The average<br />

energy is E m<br />

and σ is the standard deviation<br />

of that average. The FWHM = 2.3548 σ. See<br />

Experiment #9 for more information.<br />

3.3<br />

⎛ ∆E⎞<br />

E<br />

ln ln ln E . lnE<br />

⎝ E ⎠ ∝ ⎛ ⎝ ⎜ ⎞<br />

E<br />

⎟<br />

⎠<br />

= ⎛ − 1 2 ⎞<br />

⎝ ⎠<br />

=−05<br />

13

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!