This site is supported by donations to The OEIS Foundation.

User:Jianing Song

From OeisWiki
Jump to: navigation, search
  • Which of 222Rn and 222Fr has lower energy?
  • Which of the state and the state of 214At has lower energy? (levels of 214At)
  • Which of 247Cm and the state of 247Bk has lower energy? (levels of 247Bk)
  • Which of 259Fm and 259Md has lower energy?

Note the error margins! These questions do not have decisive answers by now.

Important: Please kindly contact me by jianing.song@polytechnique.edu. My old email address sjn1508@163.com is no longer in use.

Table of minimal polynomial of -2 cos(2π/n) (note the minus sign!)

PARI code giving the minimal polynomial of :

P(n) = prod(m=1, n/2, (x + 2*cos(2*m*Pi/n))^(gcd(n,m)==1))

This is not most satisfactory because it doesn't give an integral polynomial, but it gives the correct result better than the apparently "obvious" solution

P(n) = algdep(-2*cos(2*Pi/n), eulerphi(n)/2)

which works for n ≤ 25, but then breaks down because a false polynomial nullifies with smaller floating error.

n Pn Pn(k) for k = 3 k = 4 k = 6
3 2 3 5
4 3 4 6
5 5 11 29
6 4 5 7
7 13 41 169
8 7 14 34
9 17 51 197
10 11 19 41
11 89 571 5741
12 6 13 33
13 233 2131 33461
14 29 71 239
15 61 241 1345
16 47 194 1154
17 1597 29681 1136689
18 19 53 199
19 4181 110771 6625109
20 41 181 1121

Let p be an odd prime not dividing k2 - 4, then there exists a unique number of the form such that p divides Pn(k). The parity of m is determined by the value of . Consider the case k = 4, so k2 - 4 = 12 and -(k - 2) = -2:

p 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97
corresponding m 1 1 2 1 2 2 1 1 1 1 6 2 1 3 2 1 4 5 2 1 2 2 6
-1 -1 1 -1 1 1 -1 -1 -1 -1 1 1 -1 -1 1 -1 1 -1 1 -1 1 1 1

Replacing and in linear algebra

Traditionally, various topics on linear algebra assume that the base field of study is either or . But wait, we put too many restrictions on ourselves!

Definition. An ordered field is called archimedean if the integers are unbounded. An ordered field is called an Euclidean field if every positive element is a square. A field is called quadratically closed if every element is a square. A field is called real-closed if is not a square , and is algebraically closed.

We note that a field can be turned into an Euclidean field if and only if is not a square in , and is quadratically closed.

Theory on inner product: can be replaced by any archimedean Euclidean field. Technically inner products can be defined on every ordered field, but in order to allow Gram–Schmidt process we need that every positive element is a square, so it's better if the base field is Euclidean. Moreover, we want that a norm satisfying parallelogram law is induced by an inner product, so we want to be dense in the base field, which means that it's best if the base field is also archimedean. (If we do not care about norms, of course an Euclidean field is enough.) I didn't figure out how we could replace . I presume that it can be replaced by an arbitrary quadratically closed field with characteristic 0 and cardinality , but must such a field have an archimedean Euclidean subfield of index 2? (Anyway, must a quadratically closed field with characteristic 0 have an Euclidean subfield of index 2? Of course this is true for an algebraically closed field with characteristic 0: see the next paragraph.)

Theory on orthogonal/unitary diagonalization: can be replaced by any real-closed field, on which every symmetric matrix is diagonalizable. In fact real-closedness is a sufficient but not necessary condition (see here), but for the sake of simplicity we will assume it. can be replaced by any algebraically closed field with characteristic 0, because such a field must have a subfield of index 2 (which must be real closed; see here), so an automorphism of order 2 which looks like the ordinary complex conjugation.

Theory on binary forms: can be replaced by any Euclidean field, which is precisely one of the two kinds of fields on which the Sylvester law of inertia holds (see here). This means that every symmetric matrix on an Euclidean field is congruent to a diagonal matrix consisting of 1s, (-1)'s, and 0s, and the occurence of each number is independent of the congruence factor. can be replaced by any quadratically closed field (the other kind of fields on which the Sylvester law of inertia is valid; see the link aformentioned in this paragraph): every symmetric matrix on such a field is is congruent to a diagonal matrix consisting of 1s and 0s, where the number of 1s is equal to the rank of the matrix.

Character values of representations of finite group

Let be a field. What is the set of all numbers of the form , where is a representation of a finite group in , and is an element in ?

Lemma. Let be an algebraically closed field, and be the set of elements in that can be expressed as a sum of roots of unity. If is of characteristic , then , where is the prime field of and is the algebraic closure of within . If is of characteristic 0, then , where is the prime field of , is the extension of generated by all roots of unity, and is the set of algebraic integers in ,

Proof. The inclusion is clear. For the case characteristic , every element in is automatically a root of unity, so lies in . For characteristic 0, each element in lies in some , so it is a sum of roots of unity (possibly with repetitions).

Since is of finite order, is also of finite order, so the eigenvalues of in (an algebraic closure of ) must be roots of unity, which means that a character value must be an element in which can be expressed as a sum of roots of unity in .

If is of characteristic , let be the prime field of , then the lemma above tells us that a character value lies in , which is the maximal algebraic subextension of within . Since every element in is itself a root of unity, every element in is automatically a character value of a representation of dimension 1 of a finite cyclic group in .

If is of characteristic 0, let be the prime field of , then the lemma above tells us that a character value lies in (note that is the maximal abelian subextension of within ). It remains to show that if an element in which can be expressed as a sum of roots of unity in , then it must occur as a character value. ...

Alpha-emitters with Z ≤ 83 and half-lives lower than 1020 years

No alpha decay with a half-life of at least 1020 years is known, nor do I think we will be able to confirm one in this century.

A half-life of at least 1013 years would only have theoretical values, meaning that the alpha decay of nuclides from 184Os on can be ignored. Nuclides with half-lives not exceeding that of 146Sm are too short-lived to exist in nature. So 147Sm (1.066×1011 years) and 190Pt (4.83×1011 years) are the only primordial nuclides with Z ≤ 83 that emit a noticeable amount of alpha particles (still insignificant compared to 232Th, 235U, and 238U). In comparison, natural beta-emitters are more common - there are 87Rb, 138La, 176Lu, 187Re, and most notably, 40K.

Nuclide Z N log Tα (yr) Qβ+ (keV) Q+ Q+ Q+
144Nd 60 84 15.36 - - - -
146Sm 62 84 7.96 - - - -
147Sm 62 85 11.03 - - - -
148Sm 62 86 15.80 - - - -
149Sm 62 87 17.92 - 18.99 (not yet observed) - - - -
151Eu 63 88 18.66 - - - -
148Gd 64 84 1.94 28 3067 597 1139
150Gd 64 86 6.25 - 1287 - -
152Gd 64 88 14.03 - 54 - -
154Dy 66 88 6.15 - 3312 1344 2062
174Hf 72 102 16.85 - 1103 - -
180W 74 106 18.26 - 144 - -
184Os 76 108 13.05 - 1451 - -
186Os 76 110 15.30 - - - -
187Os 76 111 16.61 - 19.65 (not yet observed) - - - -
190Pt 78 112 11.68 - 1383 - -
209Bi 83 126 19.30 - - - -
210mBi 83 127 6.48 208 - - -

Nuclides with the lowest mass among isobars of mass numbers 141 ~ 209

For 141 ≤ A ≤ 209, the nuclide with the lowest mass among isobars of A (141Pr, 142Nd, 143Nd, 144Nd, 145Nd, 146Sm, ..., 209Bi) are of particular interest. The following table lists their predicted half-lives here:

A Nuclide log Tα (yr)
146 146Sm 7.96
147 147Sm 11.03
186 186Os 15.30
144 144Nd 15.36
148 148Sm 15.80
187 187Os 17.66
149 149Sm 18.47
151 151Eu 18.66
209 209Bi 19.30
176 176Hf 20.51
177 177Hf 22.00
145 145Nd 22.93
192 192Pt 22.96
178 178Hf 23.64
185 185Re 25.02
188 188Os 26.28
150 150Sm 28.05
191 191Ir 29.09
170 170Yb 29.24
189 189Os 31.18
179 179Hf 31.31
182 182W 32.82
175 175Lu 35.28
184 184W 35.97
204 204Pb 36.01
171 171Yb 36.50
181 181Ta 38.78
183 183W 39.46
172 172Yb 42.34
194 194Pt 44.47
180 180Hf 45.84
190 190Os 47.39
169 169Tm 47.63
198 198Hg 52.15
154 154Gd 52.60
173 173Yb 62.37
195 195Pt 64.13
166 166Er 64.52
206 206Pb 66.55
193 193Ir 67.00
197 197Au 71.78
174 174Yb 75.31
203 203Tl 80.57
167 167Er 80.96
196 196Pt 82.67
199 199Hg 83.82
143 143Nd 88.12
168 168Er 91.90
200 200Hg 95.46
153 153Eu >100
160 160Dy >100
152 152Sm >100
155 155Gd >100
156 156Gd >100
208 208Pb >100
161 161Dy >100
157 157Gd >100
207 207Pb >100
159 159Tb >100
158 158Gd >100
165 165Ho >100
162 162Dy >100
201 201Hg >100
163 163Dy >100
142 142Nd >100
202 202Hg >100
164 164Er >100
205 205Tl >100
141 141Pr >100

For nuclides in the table with predicted alpha half-lives longer than 10100 years (including being stable to alpha decay), cluster decay modes (for example 16O emission) must be taken into account when modeling the half-lives. It seems that there is no concensus on predicting cluster decay half-lives, in contrast to alpha decay.

Beta-stable nuclides around technetium and promethium

It is well-known that Tc and Pm have no beta-stable isotopes, which is to say, all their isotopes undergo beta decays to form adjecant elements. Let's draw the table of beta-stable nuclides around Tc (which have an even number of protons and 54, 55, or 56 neutrons), and those around Pm (which have an even number of protons and 84, 85, or 86 neutrons):

Beta-stable nuclides around Tc (actually 96Zr is not stable to )
Z = 40 Z = 42 Z = 44 Z = 46
N = 54 94Zr 96Mo 98Ru
N = 55 97Mo 99Ru
N = 56 96Zr 98Mo 100Ru 102Pd
Beta-stable nuclides around Pm (actually 148Gd is not stable to )
Z = 58 Z = 60 Z = 62 Z = 64
N = 84 142Ce 144Nd 146Sm 148Gd
N = 85 145Nd 147Sm
N = 86 146Nd 148Sm 150Gd

There are some patterns of symmetry here. 96Zr is energetically unstable to , but its half-life is too long (> 1020 years) to be measured. The nuclide lying at the opposite corner of the second table, 148Gd, has the same property: is energetically unstable to , but the half-life of this process (which has a spin change of ) is way too long. (Note that 148Gd does dacay via alpha emission to 144Sm, having a short half-life of only 86.9 years. Its decay releases an energy of 3271.21 keV. For the sake of symmetry let's imagine 96Zr absorbing an alpha particle to form 100Mo; this process would release an energy of 3179.1 keV).

The nuclides in bold have the lowest energy among its isobars. There is one breaking of symmetry: the nuclides with lowest energy for A = 98 and A = 146 do not lie at opposite positions of the two tables. Yet, the energy differences of 98Mo/98Ru (112.75 keV) and 146Nd/146Sm (70.83 keV) are small. (Although those of 152Sm/152Gd (54.16 keV) and 164Dy/164Er (23.33 keV) are smaller).

The nuclide 94Zr is stable to decay by 901.7 keV, meaning that it is 901.7 keV lower in energy than 94Nb. Likewise, the corresponding nuclide lying at the opposite corner of the second table, 150Gd, is stable to decay by 972 keV, meaning that it is 972 keV lower in energy than 150Eu.

Factoring rational primes on the quadratic number field with discriminant