You're reading the documentation for a development version. For the latest released version, please have a look at v0.2.
Transient nutation: FFT
Classes used:
General description
Not all tr-EPR time traces show transient nutations, but if so, it may be of interest to extract the nutation frequency, as this allows to extract the local field strength of the microwave field at the sample location. The easiest way to extract the nutation frequency is by applying a 1D fast Fourier transform (FFT) along the time domain of the data.
The actual analysis step used, trepr.analysis.TransientNutationFFT, works with both, 1D and 2D datasets. The difference of the two scenarios shown below is more in the need for pre- and post-processing of the data.
Transient nutation analysis in 1D
In case of signals showing strong oscillations with the signal passing the zero line multiple times, there is usually no need for any pre-processing of the data, such as subtracting an exponential decay. Experimentally, this situation is typically found in crystalline samples of dilute paramagnetic species, such as a single crystal consisting of pentacene in a p-terphenyl matrix.
Recipe
1format:
2 type: ASpecD recipe
3 version: '0.2'
4
5settings:
6 default_package: trepr
7 autosave_plots: false
8
9directories:
10 datasets_source: ../../tests/testdata/
11
12datasets:
13 - speksim/
14
15tasks:
16 - kind: processing
17 type: PretriggerOffsetCompensation
18 comment: >
19 Compensate for DC offsets of the signal prior to the laser flash
20 - kind: aspecd.processing
21 type: Averaging
22 properties:
23 parameters:
24 range: [370.4, 371.4]
25 unit: axis
26 axis: 0
27 result: 1Ddataset-averaged
28 - kind: singleplot
29 type: SinglePlotter1D
30 properties:
31 parameters:
32 tight: x
33 tight_layout: true
34 apply_to:
35 - 1Ddataset-averaged
36 result: time_trace
37 - kind: singleanalysis
38 type: TransientNutationFFT
39 properties:
40 parameters:
41 padding: 10
42 result: tnfft
43 comment: Perform FFT on dataset
44 apply_to: 1Ddataset-averaged
45 - kind: singleanalysis
46 type: TransientNutationFFT
47 properties:
48 parameters:
49 padding: 10
50 window: kaiser
51 window_parameters: [3]
52 result: tnfft-window
53 comment: Perform FFT on dataset
54 apply_to: 1Ddataset-averaged
55 - kind: multiplot
56 type: MultiPlotter1D
57 properties:
58 parameters:
59 tight_layout: true
60 show_legend: true
61 properties:
62 axes:
63 xlim: [0, 2e6]
64 drawings:
65 - label: no window
66 - label: Kaiser window
67 grid:
68 show: true
69 axis: x
70 result: fft
71 apply_to:
72 - tnfft
73 - tnfft-window
74 - kind: compositeplot
75 type: CompositePlotter
76 properties:
77 properties:
78 figure:
79 dpi: 150
80 size: [8, 4]
81 grid_dimensions: [1, 2]
82 subplot_locations:
83 - [0, 0, 1, 1]
84 - [0, 1, 1, 1]
85 plotter:
86 - time_trace
87 - fft
88 filename: transient-nutation-1d-fft.png
Result
The resulting figure of the above recipe is shown below.
Result of the transient nutation analysis applying a 1D FFT along the time domain. The situation of no window and applying a Kaiser window is compared. For this case, no window would be necessary to extract the nutation frequency.
Transient nutation analysis in 2D
As mentioned above, the trepr.analysis.TransientNutationFFT analysis step does not care whether data are 1D or 2D, and can be applied to 2D data in the exact same way. The other difference here: The data used show still dominant transient nutations, but this time, the nutation is merely a modulation of the exponential decay. Hence, it is usually wise to subtract the exponential decay prior to applying the FFT.
Recipe
1format:
2 type: ASpecD recipe
3 version: '0.2'
4
5settings:
6 default_package: trepr
7 autosave_plots: false
8
9directories:
10 datasets_source: ../../tests/testdata/fsc2/
11
12datasets:
13 #- radical-pair
14 - triplet
15
16tasks:
17 - kind: processing
18 type: PretriggerOffsetCompensation
19 comment: >
20 Compensate for DC offsets of the signal prior to the laser flash
21 - kind: processing
22 type: BackgroundCorrection
23 comment: >
24 Correct for laser-induced background signal.
25 - kind: singleanalysis
26 type: TransientNutationFFT
27 properties:
28 parameters:
29 padding: 10
30 subtract_decay: True
31 result: tnfft
32 comment: Perform FFT on dataset
33 - kind: singleanalysis
34 type: TransientNutationFFT
35 properties:
36 parameters:
37 padding: 10
38 subtract_decay: True
39 window: cosine
40 result: tnfft-window
41 comment: Perform FFT on dataset with cosine window
42 - kind: singleanalysis
43 type: BasicCharacteristics
44 properties:
45 parameters:
46 kind: max
47 output: indices
48 axis: 0
49 result: max_field
50 - kind: processing
51 type: SliceExtraction
52 properties:
53 parameters:
54 position: max_field
55 axis: 0
56 result: time_trace_max
57 comment: Extract transient for global maximum of the dataset
58 - kind: singleanalysis
59 type: BasicCharacteristics
60 properties:
61 parameters:
62 kind: min
63 output: indices
64 axis: 0
65 result: min_field
66 - kind: processing
67 type: SliceExtraction
68 properties:
69 parameters:
70 position: min_field
71 axis: 0
72 result: time_trace_min
73 comment: Extract transient for global minimum of the dataset
74 - kind: processing
75 type: SliceExtraction
76 properties:
77 parameters:
78 position: max_field
79 axis: 0
80 result: frequency_max
81 comment: Extract frequency for global maximum of the dataset
82 apply_to: tnfft
83 - kind: processing
84 type: SliceExtraction
85 properties:
86 parameters:
87 position: min_field
88 axis: 0
89 result: frequency_min
90 comment: Extract frequency for global minimum of the dataset
91 apply_to: tnfft
92 - kind: processing
93 type: SliceExtraction
94 properties:
95 parameters:
96 position: max_field
97 axis: 0
98 result: frequency_max-window
99 comment: Extract frequency for global maximum of the dataset
100 apply_to: tnfft-window
101 - kind: processing
102 type: SliceExtraction
103 properties:
104 parameters:
105 position: min_field
106 axis: 0
107 result: frequency_min-window
108 comment: Extract frequency for global minimum of the dataset
109 apply_to: tnfft-window
110 - kind: singleplot
111 type: SinglePlotter2D
112 properties:
113 parameters:
114 tight_layout: true
115 properties:
116 figure:
117 dpi: 300
118 type: contourf
119 result: original_data
120 - kind: singleplot
121 type: SinglePlotter2D
122 properties:
123 parameters:
124 tight_layout: true
125 properties:
126 figure:
127 dpi: 300
128 axes:
129 ylim: [0, 8e5]
130 type: contourf
131 result: fft_data
132 apply_to: tnfft-window
133 - kind: multiplot
134 type: MultiPlotter1D
135 properties:
136 parameters:
137 tight_layout: true
138 tight: x
139 properties:
140 grid:
141 show: true
142 axis: x
143 result: time_traces
144 apply_to:
145 - time_trace_max
146 - time_trace_min
147 - kind: multiplot
148 type: MultiPlotter1D
149 properties:
150 parameters:
151 tight_layout: true
152 show_legend: true
153 properties:
154 axes:
155 xlim: [0, 5e5]
156 drawings:
157 - label: max
158 - label: min
159 - label: max cosine
160 - label: min cosine
161 grid:
162 show: true
163 axis: x
164 result: frequency_traces_compare_windows
165 apply_to:
166 - frequency_max
167 - frequency_min
168 - frequency_max-window
169 - frequency_min-window
170 - kind: multiplot
171 type: MultiPlotter1D
172 properties:
173 parameters:
174 tight_layout: true
175 properties:
176 axes:
177 xlim: [0, 5e5]
178 grid:
179 show: true
180 axis: x
181 result: frequency_traces
182 apply_to:
183 - frequency_max-window
184 - frequency_min-window
185 - kind: compositeplot
186 type: CompositePlotter
187 properties:
188 properties:
189 figure:
190 dpi: 150
191 size: [8, 7]
192 grid_dimensions: [2, 2]
193 subplot_locations:
194 - [0, 0, 1, 1]
195 - [0, 1, 1, 1]
196 - [1, 0, 1, 1]
197 - [1, 1, 1, 1]
198 plotter:
199 - original_data
200 - fft_data
201 - time_traces
202 - frequency_traces
203 filename: transient-nutation-2d-fft.png
Result
The resulting figure of the above recipe is shown below.
Result of the transient nutation analysis of a full 2D dataset by applying a 1D FFT along the time domain. The transient nutation frequencies for the absorptive and emissive signal maxima are shown.