-
Notifications
You must be signed in to change notification settings - Fork 0
/
Notas.tex
2376 lines (2226 loc) · 122 KB
/
Notas.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
\documentclass[a4paper,12pt]{article}
% Input encoding
% This is commented out because lualatex automatically understands UTF-8 and it
% might not only be redundant, but also can cause programs
%\usepackage[utf8]{inputenc}
% Font encoding
\usepackage[T1]{fontenc}
% Autogenerated text, the option is for it to be inserted in the right language
\usepackage[english]{babel}
\usepackage{graphicx}
% Setting margin between text and document border
\usepackage[margin=2cm]{geometry}
% Necessary math commands
\usepackage{mathtools}
% Math symbols
\usepackage{amssymb}
% Enabling the usage of [H] in the figure environment
\usepackage{float}
% Using subfigures
\usepackage{subfigure}
% Multiple columns
\usepackage{multicol}
% Various physics notations
\usepackage{physics}
% Need to look into what these do.
%\usepackage{grffile}
%\usepackage{longtable}
%\usepackage{wrapfig}
%\usepackage{rotating}
%\usepackage[normalem]{ulem}
%\usepackage{textcomp}
%\usepackage{capt-of}
% Feynman diagrams
\usepackage{tikz-feynman}
% Feynman slash notation
\usepackage{slashed}
% Proper quotes, mostly used in references
\usepackage{csquotes}
% Bibliography management
\usepackage[backend=biber,sorting=none]{biblatex}
% Bibliography database to use
\addbibresource{../TFM.bib}
% Putting hyperlinks, like those from the index to the section it's pointing to,
% but can also point to an equation from within the body of text
\usepackage{hyperref}
% Configuring hyperref so that hyperlinks are highlighted
\hypersetup{
colorlinks=true,
linkcolor=blue,
filecolor=magenta,
urlcolor=cyan,
}
\author{Agustín Matías Galante Cerviño}
\date{\today}
\title{Notes on ``The scotogenic model of neutrino masses and dark matter in warped extra dimensions''}
\hypersetup{
pdfauthor={Agustín Matías Galante Cerviño},
pdftitle={Notes on ``The scotogenic model of neutrino masses and dark matter in warped extra dimensions''},
pdfkeywords={},
pdfsubject={},
pdfcreator={},
pdflang={English}
}
% Using macros to shorthand some commands
\newcommand{\me}{\mathrm{e}}
\newcommand{\md}{\mathrm{d}}
\newcommand{\mi}{\mathrm{i}}
\newcommand{\hc}{\mathrm{H.c.}}
\newcommand{\Lagr}{\mathcal{L}}
\newcommand{\eff}{\mathrm{eff}}
\newcommand{\IR}{\mathrm{IR}}
\newcommand{\UV}{\mathrm{UV}}
\newcommand{\bulk}{\mathrm{bulk}}
\begin{document}
\maketitle
\tableofcontents
\clearpage
\section{Summary}
\label{sec:summary}
This thesis has, as a subject of study, models in theories with extra dimensions in which neutrinos
acquire masses through loops, which also contain a dark matter candidate. It will include a
calculation of the abundance of dark matter, (for example, from fermionic singlets in the scotogenic
model) and an analysis of the phenomenological consequences.
Since both the extra dimension and scotogenic models have been covered extensively in literature,
this thesis will only really introduce concepts without delving too deep into them, referencing
literature along the way. Main focus ought to be on calculating the abundance of this dark matter
candidate, an analysis of the phenomenological consequences and parameter space constraints.
\section{Preamble and terminology}
\label{sec:preamble}
\begin{itemize}
\item Note that by using upper case letters when referring to fields
\((L_L, \Phi)\), we are actually referring to gauge multiplets (unless stated otherwise, or in
section \ref{sec:kk}), so
\begin{equation}
L_{L, i} = \begin{pmatrix} \nu_{L,i} \\ l_{L,i} \end{pmatrix} \quad \Phi
= \begin{pmatrix} \phi^+ \\
\phi^0 \end{pmatrix} \quad \widetilde{\Phi} = \mi \sigma_2 \Phi^* = \begin{pmatrix}
\phi^{0*} \\
-\phi^{-} \end{pmatrix}
\end{equation}
Sub-index L indicates that they're left-handed (LH) fields, i and j are generational indices.
\item A regular (scalar, for simplicity) field destroys a particle at position x and creates
an antiparticle at the same position. The complex conjugate of a field creates a particle
and destroys an antiparticle. Furthermore, complex conjugate fields have the opposite
quantum numbers, so
\begin{equation}
(\phi^{+})^{*} = \phi^-
\end{equation}
has opposite electric charge, as well as opposite weak hypercharge and weak isospin, in the case of the Higgs.
\item Remember that boundary conditions are indeed physical for most cases, except for when one
has boundary conditions at infinity, or periodic conditions over an infinite interval.
\item Feynman rules of a theory are obtained by calculating the transition amplitude multiplied by i;
that is, crafting a final and initial state that can be completely contracted with the fields in
\(\Lagr_{int}\), calculating \(\bra{f}\Lagr_{int}\ket{i}\) and multiplying that by i; the resulting
Feynman rule is what's left after taking out the external line factors (fermion bispinors, polarization vectors...)
\item The lightest neutrino mass eigenstate could have no mass at all, so really the minimum amount
of Majorana right-handed (RH) fields that are to acquire mass is 2.
\item What people mean by ``integration by parts'' refers to this \cite{integrationparts}.
\item \textbf{Chirality}
I know about this by now, but a reminder doesn't hurt. A Dirac mass term is proportional to
\begin{equation}
\overline{\psi} \psi = \overline{\psi}_L \psi_R + \overline{\psi}_R \psi_L
\end{equation}
where
\begin{equation}
\psi_R = \frac{1}{2}(1 + \gamma_5) \psi \quad \psi_L = \frac{1}{2}(1 - \gamma_5) \psi
\end{equation}
and
\begin{equation}
\begin{aligned}
\overline{\psi}_R = \psi^{\dagger} \frac{1}{2}(1 + \gamma_5) \gamma^0 = \overline{\psi}
\frac{1}{2}(1 - \gamma_5) = \overline{\left(\psi_R\right)} = \overline{\left(\psi_R\right)} \\[7pt]
\overline{\psi}_L = \psi^{\dagger} \frac{1}{2}(1 - \gamma_5) \gamma^0 = \overline{\psi}
\frac{1}{2}(1 + \gamma_5) = \overline{\left(\psi_L\right)}
\end{aligned}
\end{equation}
\item \textbf{Numbers in between parenthesis}
Looks like those are just the labels for the representation of the group under which the fields
transform, since only SU(2), and U(1) are talked about here. For SU(2), the dimension of the
representation is shown, while for U(1), it's the ``charge''. With weak hypercharge, we will use
the convention \(Q = T_3 + Y\)
\item \textbf{Radiative mass}
A massless particle may acquire mass (or a massive particle change its mass),
not at tree level, but through loops with itself, given that, when renormalizing
its Lagrangian, mass terms appear. By ``radiative'', it means that there are loops involved.
\item \textbf{Weinberg operator}
A term of an effective Lagrangian, appearing on such seesaw models (actually it is their goal to end
up with this term), with mass dimension 5 (thus not renormalizable) which has the form \textbf{(this
needs checking)} \textbf{(checked, turns out Ma takes the H.c. of what I took to be the Higgs scalar doublets)}
\begin{equation}
\frac{g_{ij}}{\Lambda} \overline{L}_{L,i} \widetilde{\Phi} L^c_{L,j} \widetilde{\Phi} + \hc
= \frac{g_{ij}}{\Lambda} (\overline{\nu}_i \phi^{0*} - \overline{l}_i \phi^-)(\nu_j \phi^{0*} - l_j \phi^-) + \hc
\end{equation}
The reason for its appearance is that it is the operator with the smallest dimension that
can be added beyond the renormalizable terms of the SM Lagrangian, if formulating the SM
as an effective field theory. When EWSB occurs, this term generates a Majorana mass term
for left-handed (LH) neutrinos.
In radiative neutrino mass models, like the one considered here, we have that the ratio
\(g_{ij}/\Lambda\) is small, with \(\Lambda\) not being too big (\textasciitilde 1-100 TeV)
\item \textbf{Manifold}
Topological space locally resembling Euclidean space near each point.
\item \textbf{Orbifold}
Generalization of a manifold.
\item \textbf{3-brane}
four-dimensional subspace in \(n\)-dimensional spacetime.
\item \textbf{Graviton}
Within the scope of our study, we will consider the limit of small, linear perturbations of the metric,
in the following manner
\begin{equation}
g_{\mu \nu} \approx \eta_{\mu \nu} + h_{\mu \nu}
\end{equation}
(in 4 dimensions) with \(\eta_{\mu \nu}\) being the Minkowski metric and \(h_{\mu \nu}\) being our
graviton. This theory of quantum gravity is an effective field theory, valid below a certain
energy scale. This is a spin-2, massless particle.
\item \textbf{Graviscalar}
In five dimensions, the metric is still perturbed as above, however now we have additional
components of the metric beyond the ones belonging to the typical four dimensions.
The radion is the perturbation (\(h_{55}\), if you will) of the fifth diagonal component of the
metric \((g_{55})\). In four dimensions, this is indistinguishable from a scalar.
\item \textbf{Graviphoton or gravivector}
Same as the above, however there are additional components of the tensor aside from the one in
the diagonal which include the fifth dimension in the metric \((g_{\mu 5})\), so the perturbation
is now a massless vector field when seen in four dimensions \((h_{\mu 5})\).
\end{itemize}
\section{Neutrino masses}
\label{sec:neumass}
\subsection{Motivation}
It is perfectly possible to grant mass to neutrinos by extending the SM with Yukawa interaction terms
analogous to those used with right-handed quarks, as well as sterile, right-handed neutrinos.
However, many have taken issue with the extremely small couplings (\(<10^{-12}\)) necessary to fit
observations, strongly believing that such a disparity with analogous couplings indicates that the
mechanism involved is more complex than this extension. As a result, a myriad of models sprung up in
the past 40 years aiming to provide a more compelling way of resolving this problem.
\subsection{Seesaw models}
\label{sec:seesaw}
A family of models used to generate neutrino masses. Mostly taken from \cite{claude2015}, additional
section to chapter 14.
Let us take one family of neutrinos, however this can be applied to any number of them.
The fields from which we can build a theory are, \(\psi_L\), \(\psi_R\),
\((\psi^c)_R = 1/2 (1 + \gamma_5)\psi^c\),
\((\psi^c)_L\), as well as \(\overline{\psi}_L\), \(\overline{\psi}_R\),
\(\overline{(\psi^c)_R}\), and \(\overline{(\psi^c)_L}\). Remember that, as
the representation \textbf{2} of SU(2) is equivalent to its conjugate, we also have
to take into account the latter when adding Lorentz invariant terms to the Lagrangian,
so that is presumably why the charge conjugated fields are included. (I doubt this can
be done if parity is to be respected, doesn't a Majorana mass term violate parity? Doubt it
does, otherwise I'd see it mentioned).
They are not quite this (we're using the Weyl representation throughout this section),
\begin{equation}
\psi^c = C \psi^* = C \begin{pmatrix} \psi^*_L \\ \psi^*_R \end{pmatrix}
\end{equation}
but from the definition of the charge conjugation operator,
(this being a matrix in bispinor space, like parity, gamma matrices\ldots)
\begin{equation}
C \gamma_{\mu} C^{-1} = -(\gamma_{\mu})^T
\end{equation}
one can see that
\begin{equation}
\psi^c = \begin{pmatrix}(\psi^{c})_{L} \\
(\psi^{c})_{R} \end{pmatrix} = C \overline{\psi}^T = C \begin{pmatrix} \psi^*_R \\
\psi^*_L \end{pmatrix} = \begin{pmatrix} \mi \sigma^2 & 0 \\
0 & - \mi \sigma^2 \end{pmatrix} \begin{pmatrix} \psi_R^* \\
\psi_L^* \end{pmatrix} =\begin{pmatrix} \mi \sigma^2 \psi_R^* \\
- \mi \sigma^2 \psi_L^* \end{pmatrix}
\end{equation}
Mass terms containing these fields are the following,
\begin{equation}
M_D \overline{\psi}_R \psi_L + \frac{1}{2} M_L \overline{(\psi_L)^c} \psi_L + \frac{1}{2} M_R \overline{(\psi_R)^c} \psi_R + \hc
\end{equation}
It's important to note that Majorana mass terms (the two last ones) violate leptonic number by two units.
We also have that
\begin{equation}
(\psi^c)_L = \mi \sigma_2 \psi^*_R = (\psi_R)^c \quad (\psi^c)_R = -\mi \sigma_2 \psi^*_L = (\psi_L)^c
\end{equation}
so indeed the fundamental and its conjugate rep. are in the same term.
The kicker is that we can just include both these Dirac and Majorana mass terms in the same
Lagrangian, if our criterion for adding more terms is that they must be Lorentz invariant. For
arbitrary Dirac and Majorana mass parameters these fields are not the eigenstates of the
Hamiltonian, thus they do not generate physical, observable states. One needs to redefine the
fields,
\begin{equation}
f = \frac{1}{\sqrt{2}}(\psi_L + (\psi^c)_R) \quad F = \frac{1}{\sqrt{2}}(\psi_R + (\psi^c)_L)
\end{equation}
These are Majorana fields, since \(f^c = f\) and \(F^c = F\). Remember that, for a left-handed and
right-handed Majorana field, the bispinors are
\begin{equation}
\begin{pmatrix} \psi_L \\ \mi \sigma_2 \psi^*_L \end{pmatrix}
\quad \begin{pmatrix} \mi \sigma_2 \psi^*_R \\ \psi_R \end{pmatrix}
\end{equation}
One now gets the mass terms
\begin{equation}
M_L \overline{f}f + M_R \overline{F}F + M_D (\overline{f}F + \overline{F}f)
= (\overline{f} \: \overline{F}) \begin{pmatrix} M_L & M_D \\ M_D & M_R \end{pmatrix} \begin{pmatrix} f \\
F \end{pmatrix}
\end{equation}
(This setup laid down is what most seesaw models get at.)
This is, again, not physical, so diagonalizing this matrix, and finding eigenvalues
and eigenvectors will yield observable masses and fields. These are
\begin{equation}
\nu' = \frac{1}{\sqrt{2}}(F + f) \quad N = \frac{1}{\sqrt{2}}(F - f)
\end{equation}
Note that kinetic terms (ones with derivatives of the fields) are always diagonal in flavor,
and so they are not affected by field redefinitions, they are always diagonal
in whatever field is chosen \textbf{(need to check this)}. In the case \(M_L = M_R = 0\), we get
\begin{equation}
-M_D \overline{\nu}' \nu' + M_D \overline{N} N = M_D \overline{\nu} \nu + M_D \overline{N} N
\end{equation}
where \(\nu = \gamma_5 \nu'\), a redefinition made so that the mass remains
positive. Another case is \(M_L = 0, M_R >> M_D\), which
In the type-I seesaw, an additional Higgs-like scalar field singlet (under every group except
\(U(1)_{B-L}\)) is added in order to grant the RH neutrino a Majorana mass
term when this symmetry is spontaneously broken.
\section{Scotogenic model}
\label{sec:scm}
Generation of neutrino masses at one loop order, through the coupling at one loop of SM neutrino
flavour eigenstates with the Higgs through a Higgs copy. \cite{ma2006verifiable}
The scotogenic model consists in the generation of neutrino masses through the coupling at one loop
of neutrinos with the SM Higgs through various new particles, these being a Higgs-like scalar
doublet under \(SU(2)_{L}\), \(\eta\), and fermionic singlets, \(N_{i}\). In order to accomplish
this, an exact \(Z_2\) symmetry is introduced as well. If the Higgs potential, now containing both
the SM Higgs and this doublet, is to respect this symmetry, then the bilinear term mixing both
fields, \(\mu^{2}\Phi^{\dagger}\eta\), has to be forbidden. This term would otherwise have been able
to give neutrinos mass at tree level. Not only does this model attempt to explain the origin of
neutrino masses, but it also happens to provide a good candidate to dark matter, given that there is
a lightest stable particle that could either be the lightest \(N_{i}\) or
\(\sqrt{2} \: \mathrm{Re} \: \eta^{0}\).
In other words, it distinguishes the ``neutrinophobic'' regular Higgs doublet from the ``neutrinophilic''
\(\eta\). If \(Z_2\) is exact, the lightest (charged) RH neutrino is stable and a dark matter
candidate, Representations (of \(SU(2)_L \times U(1)_Y \times Z_{2}\)) of relevant particles are the
following; we first have the SM lepton LH doublet and RH singlet,
\begin{multicols}{2}
\begin{itemize}
\item \(L_i\) = \((\nu_i, l_i)^T\) \(\sim\) \((2, -1/2, +)\)
\item \(l^c_i\) \(\sim\) \((1, -1, +)\)
\end{itemize}
SM Higgs doublet,
\begin{itemize}
\item \(\Phi\) = \((\phi^{+}, \phi^{0})^T\) \(\sim\) \((2, 1/2, +)\)
\item \(\widetilde{\Phi}\) = \((\phi^{0*}, -\phi^{-})^T\) \(\sim\) \((2, -1/2, +)\)
\end{itemize}
%\vfill\null
%\columnbreak
new Higgs-like scalar doublet,
\begin{itemize}
\item \(\eta\) = \((\eta^{+}, \eta^{0})^T\) \(\sim\) \((2, 1/2, -)\)
\item \(\widetilde{\eta}\) = \((\eta^{0*}, -\eta^{-})^T\) \(\sim\) \((2, -1/2, -)\)
\end{itemize}
and finally, the fermionic singlets,
\begin{itemize}
\item \(N_{i}\) \(\sim\) \((1, 0, -)\).
\end{itemize}
\end{multicols}
The Yukawa interactions of this model are (all leptonic fields are taken to be LH so we drop the L sub-index)
\begin{equation}
f_{ij} \overline{L}_{i} \Phi l^c_{j} + h_{ij} \overline{L}_{i} \widetilde{\eta} N_j + \hc =
f_{ij} (\overline{\nu}_{i}\phi^{+} + \overline{l}_{i}\phi^{0}) l^c_{j} +
h_{ij} (\overline{\nu}_{i}\eta^{0*} - \overline{l}_{i}\eta^{-}) N_j + \hc
\end{equation}
and the most general Higgs potential allowed is
\begin{equation}
\begin{aligned}
V(\Phi, \eta) = m^{2}_{1} \Phi^{\dagger}\Phi + m^{2}_{2} \eta^{\dagger}\eta
+ \frac{1}{2}\lambda_{1}(\Phi^{\dagger}\Phi)^{2}
+ \frac{1}{2}\lambda_{2}(\eta^{\dagger}\eta)^{2} \\
+ \lambda_{3}(\Phi^{\dagger}\Phi)(\eta^{\dagger}\eta)
+ \lambda_{4}(\Phi^{\dagger}\eta)(\eta^{\dagger}\Phi)
+ \frac{1}{2}\lambda_{5}[(\Phi^{\dagger}\eta)^{2} + \hc]
\end{aligned}
\end{equation}
The fermionic singlets also admit a Majorana mass term,
\(\frac{1}{2}M_{i}\overline{N_{i}}^{c}N_{i}\). In the original paper, this mass was put in
by hand rather than having a dynamical origin like all the other masses.
Putting all this together allows for the loop in figure \ref{fig:scmmass1} to generate Majorana
mass terms when electroweak symmetry breaking (EWSB) takes place,
\begin{figure}[H]
\centering
\begin{tikzpicture}
\begin{feynman}
\vertex (i) {\(\nu_{i}\)};
\vertex[right=1.6cm of i] (a);
\vertex[right=2.2cm of a] (b);
\vertex[right=1.6cm of b] (f) {\(\nu_{j}\)};
\vertex[right=1.1 of a] (aux);
\vertex[above=1.65cm of aux] (c);
\vertex[above=3.3cm of a] (d) {\(\phi^0\)};
\vertex[above=3.3cm of b] (e) {\(\phi^0\)};
\diagram*{
(i) -- [plain] (a) -- [plain, edge label'=\(N_{s}\)] (b) -- [plain] (f),
(a) -- [anti charged scalar, edge label=\(\eta^{0}\)] (c) -- [anti charged scalar] (e),
(b) -- [anti charged scalar, edge label'=\(\eta^{0}\)] (c) -- [anti charged scalar] (d),
};
\end{feynman}
\end{tikzpicture}
\caption{Feynman diagram of the process responsible for giving mass to neutrinos}
\label{fig:scmmass1}
\end{figure}
\subsection{Neutrino mass matrix}
We shall now lay down the calculation for said mass matrix. Let us start with considering the
Lagrangian after EWSB, so we will take \(\Phi = (0, v)^{T}\), with \(v\) being the known
VEV of the SM Higgs. The Higgs potential detailed above becomes
\begin{equation}
\begin{aligned}
&= m^{2}_{2}(\eta^{+*} \eta^{+} + \eta^{0*} \eta^{0}) + \frac{\lambda_{2}}{2} (\eta^{+*} \eta^{+} + \eta^{0*} \eta^{0})^{2} \\
&+ \lambda_{3}v^{2}(\eta^{+*} \eta^{+} + \eta^{0*} \eta^{0}) + \lambda_{4}v^{2}\eta^{0*} \eta^{0} + \frac{1}{2}\lambda_{5}v^{2}((\eta^{0})^{2} + (\eta^{0*})^{2}) \\
&= (m^{2}_{2} + \lambda_{3}v^{2}) \eta^{+*} \eta^{+} + \frac{\lambda_{2}}{2} (\eta^{+*} \eta^{+})^{2} \\
& + v^{2}(\lambda_{3} + \lambda_{4}) \eta^{0*} \eta^{0} + \frac{\lambda_{5}v^{2}}{2}((\eta^{0})^{2} + (\eta^{0*})^{2}) + \frac{\lambda_{2}}{2} (\eta^{0*} \eta^{0})^{2} \\
& + \lambda_{2} (\eta^{+*} \eta^{+} \eta^{0*} \eta^{0})
\end{aligned}
\end{equation}
We see that, while \(\eta^{+}\)'s mass terms are obtained straightfowardly just by looking at the
potential, \(\eta^{0}\) has quadratic couplings not corresponding to a mass term, and as such
this is not the mass eigenstate. We will need to redefine this field in order to use the right
propagators in our one-loop calculation. So, we use two real fields, which are the
real and imaginary part of \(\eta^{0}\), so \(\eta^{0} = \eta^{0}_{R} + \mi \eta^{0}_{I}\)
and \(\eta^{0*} = \eta^{0}_{R} - \mi \eta^{0}_{I}\).
% Right after this, lay down the potential as a function of these fields
The masses we obtain are
\begin{equation}
\begin{aligned}
m^{2}(\eta^{\pm}) \equiv m^2_{\eta} &= m^{2}_{2} + \lambda_{3} v^{2}\\
m^{2}(\sqrt{2} \eta^{0}_{R}) \equiv m^{2}_{R} &= m^{2}_{2} + (\lambda_{3} + \lambda_{4} + \lambda_{5}) v^{2}\\
m^{2}(\sqrt{2} \eta^{0}_{I}) \equiv m^{2}_{I} &= m^{2}_{2} + (\lambda_{3} + \lambda_{4} - \lambda_{5}) v^{2}\\
\end{aligned}
\end{equation}
Having settled this matter, we shift our focus to the loop itself. After EWSB, we may ignore
the coupling to the surviving Higgs field and merely consider \(\eta^{0}_{R}\) and \(\eta^{0}_{I}\)
with their new masses, so we end up dealing with 2 \(\cdot\) \(N_{fs}\) self-energy diagrams.
This is because we have two scalar fields, and \(N_{fs}\) fermionic singlets. Now, we take
a look at how the Yukawa interactions look like with the new definitions,
\begin{equation}
\begin{aligned}
h_{ij} \overline{\nu}_{i}\eta^{0*} N_j + h_{ij} \overline{N}_{j}\eta^{0} \nu_i
= h_{ij} \overline{\nu}_{i}\eta^{0}_{R} N_j + h_{ij} \overline{N}_{j}\eta^{0}_{R} \nu_i
- \mi h_{ij} \overline{\nu}_{i}\eta^{0}_{I} N_j + \mi h_{ij} \overline{N}_{j}\eta^{0}_{I} \nu_i
\label{eq:yukscm}
\end{aligned}
\end{equation}
and we see that not much changes, aside from a relative sign and a couple of factors of i.
These latter factors will play out a critical role in the following.
\vspace{-2ex}
\begin{figure}[H]
\centering
\feynmandiagram [large, baseline=(d.base), layered layout, horizontal=b to c] {
a [particle=\(\nu_i\)] -- [fermion, momentum=\(p\)] b -- [scalar, half left, looseness=1.5, momentum=\(p-k\), edge label'=\(\eta^0_R/\eta^0_I\)] c
-- [anti fermion, half left, looseness=1.5, momentum=\(k\), edge label'=\(N_s\)] b,
c -- [anti fermion, momentum=\(p\)] d [particle=\(\nu_j\)],
};
\caption{Diagram of the resulting self-energy process after EWSB from fig. \ref{fig:scmmass1}}
\label{fig:scmmass2}
\end{figure}
Calculating the reduced amplitude through the use of the Feynman diagram in
figure \ref{fig:scmmass2} yields
\begin{equation}
\begin{aligned}
\begin{gathered}
\mi \mathcal{M} = \overline{u}_{\nu_j}(p, r)\sum_{s}\int \frac{\mu^{4-D}\md^{D}k}{(2\pi)^{D}}
\Bigg(\frac{\mi}{(p - k)^{2} - m^{2}_{R}} (\mi h_{js} P_L) \frac{\mi (\slashed{k} + M_{s})}{k^{2} - M^{2}_{s}} (\mi h_{is} P_L) \\
+ \frac{\mi}{(p - k)^{2} - m^{2}_{I}} (h_{js} P_L) \frac{\mi (\slashed{k} + M_{s})}{k^{2} - M^{2}_{s}}(h_{is} P_L)\Bigg)
u_{\nu_i}(p, r) \\
= \overline{u}_{\nu_j}(p, r) [-\mi \Sigma(p^{2})] u_{\nu_i}(p, r)
\end{gathered}
\end{aligned}
\end{equation}
with \(D = 4 - \epsilon\), \(\epsilon \rightarrow 0\) as we're using dimensional regularization.
So the radiative correction to the neutrino propagator is this amplitude without the external line factors,
these being the neutrino bispinors,
\begin{equation}
- \mi \Sigma(p^{2}) = - P_L \sum_{s}h_{is}h_{js} \int \frac{\mu^{4-D}\md^{D}k}{(2\pi)^{D}}
\left( \frac{\mi}{(p - k)^{2} - m^{2}_{R}} \frac{\mi M_{s}}{k^{2} - M^{2}_{s}}
- \frac{\mi}{(p - k)^{2} - m^{2}_{I}} \frac{\mi M_{s}}{k^{2} - M^{2}_{s}}\right)
\end{equation}
So we see that we really only have to solve one integral, up to the mass of the scalar in the loop.
This integral is
\begin{equation}
\int \frac{\mu^{4-D}\md^{D}k}{(2\pi)^{D}} \frac{M_{s}}{[(p - k)^{2} - m^{2}_{R}][k^{2} - M^{2}_{s}]}
\end{equation}
so now we make use of Feynman parametrization,
\begin{equation}
\frac{1}{AB} = \int^1_0 \md x \: \frac{1}{(xA + (1-x)B)^2}; \quad A = (p-k)^2 - m^2_R, \quad B = k^2 - M^2_s
\end{equation}
which yields
\begin{equation}
\int \frac{\mu^{4-D}\md^{D}k}{(2\pi)^{D}}\int^1_0 \md x \: \frac{M_{s}}{[((p-k)^2 - m^2_R)x + (k^2 - M^2_s)(1-x)]^2}
\end{equation}
Now, we need to manipulate the denominator a bit,
\begin{equation}
\begin{aligned}
\begin{gathered}
((p-k)^2 - m^2_R)x + (k^2 - M^2_s)(1-x) = p^2 x + k^2 x - 2(pk)x - m^2_R x + k^2 - k^2 x - M^2_s + M^2_s x \\
= k^2 - 2(pk)x + p^2 x^2 - p^2 x^2 + p^2 x - m^2_R x - M^2_s + M^2_s x \\
= (k - px)^2 - p^2 x^2 + p^2 x - m^2_R x - M^2_s + M^2_s x \\
= (k - px)^2 - p^2 x(x-1) + M^2_s(x-1) - m^2_R x \equiv l^2 - \Delta
\end{gathered}
\end{aligned}
\end{equation}
where we defined \(l = k - px\) and \(\Delta = m^{2}_{R}x - (p^{2}x - M^{2}_{s})(1-x)\). The resulting
integral is
\begin{equation}
\begin{aligned}
\begin{gathered}
M_{s}\mu^{4-D} \int^1_0 \md x \: \int \frac{\md^{D}k}{(2\pi)^{D}} \frac{1}{(l^2 - \Delta)^2}
\end{gathered}
\end{aligned}
\end{equation}
Knowing that
\begin{equation}
\int \md^{D}k \frac{1}{(l^2 - \Delta)^N} = \mi (-1)^N \pi^{\frac{D}{2}} \frac{\Gamma(N - \frac{D}{2})}{\Gamma(N)}\Delta^{\frac{D}{2}-N}
\label{eq:intrenor}
\end{equation}
we evaluate the integral with respect to \(k\) and press on,
\begin{equation}
\begin{aligned}
\begin{gathered}
= \frac{\mi M_{s}}{(4\pi)^{D/2}}\mu^{4-D} \Gamma \left(2 - \frac{D}{2}\right) \int^{1}_{0} \md x \Delta^{D/2 - 2}\\
= \left[\Gamma\left(2 - \frac{D}{2}\right) = \Gamma\left(\frac{\epsilon}{2}\right) \approx \frac{2}{\epsilon} - \gamma_{E}; \quad \Delta^{D/2 - 1} = \Delta^{-\epsilon/2} \approx 1 - \frac{\epsilon}{2}\ln \Delta\right] =\\
= \mi M_{s} (4\pi)^{-\frac{1}{2}(4-\epsilon)}\mu^{\epsilon}\left(\frac{2}{\epsilon} - \gamma_{E}\right) \int^{1}_{0} \md x \Delta^{-\epsilon/2}\\
\approx \frac{\mi}{(4\pi)^{2}} \left(1 + \frac{\epsilon}{2}\ln 4\pi \right) (1 + \epsilon \ln \mu)
\left(\frac{2}{\epsilon} - \gamma_{E}\right) \int^{1}_{0} \md x M_{s}\left(1 - \frac{\epsilon}{2} \ln \Delta\right)\\
\approx \frac{\mi M_{s}}{(4\pi)^{2}} \left(\frac{2}{\epsilon} - \gamma_{E} + \ln 4\pi \mu^2 \right)\int^{1}_{0} \md x \left(1 - \frac{\epsilon}{2} \ln \Delta\right)\\
\approx \frac{\mi M_{s}}{(4\pi)^{2}} \left(\frac{2}{\epsilon} - \gamma_{E} + \ln 4\pi\mu^2 \right)
- \frac{\mi M_{s}}{(4\pi)^{2}} \int^{1}_{0} \md x \ln \Delta
\label{eq:massmint}
\end{gathered}
\end{aligned}
\end{equation}
Let's stop here for a second and appreciate that the divergent contribution to this correction does
not depend on the mass of the scalar, which means that the two sorts of diagrams here, each with a
different scalar and opposite sign, have their divergent parts (and other terms not depending on this mass)
cancel each other out without resorting to renormalization. Let's remember that we are attempting to
correct the mass, so this means that, if terms proportional to \(\slashed{p}\) survived, we could
safely disregard them, however they were eliminated, as being sandwiched between two identical
chirality projectors means that it doesn't contribute, since \(P_L \slashed{k} P_L = P_L P_R \slashed{k} = 0\)
because \(\{\gamma_5, \gamma^{\mu}\} = 0\). Not only that, we shall set \(p^{2} = 0\), since neutrinos are massless at tree level.
\textbf{(Is this the actual reason? I'd say so but I could use feedback)}.
\begin{equation}
\begin{aligned}
\begin{gathered}
-\frac{\mi M_{s}}{(4\pi)^{2}} \int^{1}_{0} \md x \ln (m^{2}_{R}x + M^{2}_{s}(1-x))\\
= -\frac{\mi M_{s}}{(4\pi)^{2}} \left(\frac{M^{2}_{s}\ln M^{2}_{s} - m^{2}_{R}\ln m^{2}_{R} + m^{2}_{R} - M^{2}_{s}}{M^{2}_{s} - m^{2}_{R}}\right)\\
% We need to clarify this last step.
= -\frac{\mi M_{s}}{(4\pi)^{2}} \frac{m^{2}_{R}}{m^{2}_{R} - M^{2}_{s}}\ln \frac{m^{2}_{R}}{M^{2}_{s}}
\label{eq:asss}
\end{gathered}
\end{aligned}
\end{equation}
Now, all we need to obtain the resulting correction to the mass is to do this again, but with
\(\eta^0_I\) rather than \(\eta^0_R\) as we have just done, and to sum their contributions.
The resulting mass matrix is this times i,
\begin{equation}
\begin{aligned}
\begin{gathered}
(m_{\nu})_{ij} = \sum_{s}\frac{h_{is}h_{js}M_{s}}{16\pi^{2}}\left( \frac{m^{2}_{R}}{m^{2}_{R} - M^{2}_{s}}\ln \frac{m^{2}_{R}}{M^{2}_{s}} -
\frac{m^{2}_{I}}{m^{2}_{I} - M^{2}_{s}}\ln \frac{m^{2}_{I}}{M^{2}_{s}} \right)
\label{eq:massmat}
\end{gathered}
\end{aligned}
\end{equation}
\subsection{Lepton flavor violation}
As Majorana neutrinos do not possess a global lepton flavor symmetry, charged leptons
may transmute into others violating lepton flavor number conservation. The most sought after
process is \(\mu \rightarrow e \gamma\), having very strong upper limits for its decay rate.
The effective Lagrangian describing this decay is (\cite{Beneke_2016}, page 6, eq. (10))
\begin{equation}
\begin{aligned}
\begin{gathered}
\Lagr_{l_i \rightarrow l_j \gamma} = A_R m_i \overline{l}_j \sigma^{\mu \nu}F_{\mu \nu} P_R l_i
+ A_L m_i \overline{l}_j \sigma^{\mu \nu}F_{\mu \nu} P_L l_i + \hc \\
= A_R m_i \overline{l}_{j,L} \sigma^{\mu \nu}F_{\mu \nu} l_{i,R}
+ A_L m_i \overline{l}_{j,R} \sigma^{\mu \nu}F_{\mu \nu} l_{i,L} + \hc
\end{gathered}
\end{aligned}
\end{equation}
where subindex \(i\) denotes the heavier lepton, in our case the muon.
\begin{equation}
A_R = - \frac{e(F_2 (0) - \mi F_3(0))}{4m^2_{i}}; \quad A_L = - \frac{e(F_2 (0) + \mi F_3(0))}{4m^2_{i}}
\label{eq:aral}
\end{equation}
\begin{figure}[H]
\centering
\begin{tikzpicture}
\begin{feynman}
\vertex (i) {\(l_i\)};
\vertex[right=2.5cm of i] (a);
\vertex[right=3.2cm of a] (aux1);
\vertex[above=1.6cm of aux1] (b);
\vertex[below=1.6cm of aux1] (c);
\vertex[right=1.8cm of aux1] (aux2);
\vertex[above=3.5cm of aux2] (f1) {\(\gamma\)};
\vertex[below=3.5cm of aux2] (f2) {\(l_j\)};
\diagram*{
(i) -- [fermion, momentum=\(q_1\)] (a) -- [plain, edge label=\(N_{s}\), rmomentum'=\(k-q_1\)] (c) -- [fermion, momentum=\(q_2\)] (f2),
(a) -- [charged scalar, edge label'=\(\eta^{-}\), momentum=\(k\)] (b) -- [photon, rmomentum=\(p\)] (f1),
(b) -- [charged scalar, edge label'=\(\eta^{\pm}\), momentum=\(k+p\)] (c),
};
\end{feynman}
\end{tikzpicture}
\caption{One-loop diagram responsible for \(l_i \rightarrow l_j \gamma\)}
\label{fig:egamma}
\end{figure}
The Feynman rule for the uppermost interaction in fig. \ref{fig:egamma} is \(\mi e (k - (-k - p))_{\mu} = \mi e (2k + p)_{\mu}\),
taken from \cite{Romao_2012}, page 15; even though those rules are for the SM Higgs, \(\eta\)'s covariant derivative
is identical, since their quantum numbers are the same.\footnote{Also, according to this paper, we have chosen \(\eta_e = -1\)}
The reduced amplitude for this process is (making use of dimensional regularization from the get-go)
\begin{equation}
\begin{aligned}
\mi \mathcal{M} = \varepsilon^*_{\mu}(p, \lambda) \overline{u}_{l_j} (q_2, r_2) \Bigg(\sum_s \int \frac{\mu^{4 - D}\md^D k}{(2\pi)^D}
\frac{\mi}{k^2 - m^2_{\eta}}(\mi e (2k^{\mu} + p^{\mu}))\frac{\mi}{(k + p)^2 - m^2_{\eta}} \\
\cdot (\mi h_{js} P_L) \frac{\mi(\slashed{k} - \slashed{q_1} + M_s)}{(k-q_1)^2 - M^2_s} (\mi h_{is} P_R) \Bigg) u_{l_i} (q_1, r_1)
\end{aligned}
\end{equation}
so the vertex integral is said amplitude without the external line factors, that is, without the charged lepton bispinors and
photon polarization vector,
\begin{equation}
\begin{aligned}
\Gamma^{\mu} = - e P_L \sum_s h_{is} h_{js} \int \frac{\mu^{4 - D}\md^D k}{(2\pi)^D}
\frac{(2k^{\mu} + p^{\mu})(\slashed{k} - \slashed{q_1})}{[k^2 - m^2_{\eta}][(k + p)^2 - m^2_{\eta}][(k-q_1)^2 - M^2_s]}
\end{aligned}
\end{equation}
We are going to somewhat mirror the discussion in \cite{peskshro}, chapter 6.3, beginning at page
189. We shall now make use of Feynman parametrization, like so,
\begin{equation}
\begin{aligned}
\frac{1}{ABC} &= \int^1_0 \md x \int^1_0 \md y \int^1_0 \md z \: \delta(x + y + z - 1) \frac{2}{(xA + yB + zC)^3} \: ; \\[7pt]
A &= k^2 - m^2_{\eta} \: , B = (k + p)^2 - m^2_{\eta} \: , C = (k-q_1)^2 - M^2_s
\label{eq:feynparam3}
\end{aligned}
\end{equation}
so we take the resulting denominator and manipulate it to make it a bit more palatable,
\begin{equation}
\begin{aligned}
\begin{gathered}
x[k^2 - m^2_{\eta}] + y[(k + p)^2 - m^2_{\eta}] + z[(k-q_1)^2 - M^2_s] \\
= xk^2 - x m^2_{\eta} + yk^2 + 2y(kp) + yp^2 - y m^2_{\eta} + zk^2 - 2z(kq_1) + zq^2_1 - zM^2_s \\
= (x + y + z)k^2 + 2k(yp - zq_1) + yp^2 + zq^2_1 - (x+y) m^2_{\eta} - zM^2_s \\
= k^2 + 2k(yp - zq_1) + y^2 p^2 - y^2 p^2 + z^2 q^2_1 - z^2 q^2_1 - 2yz(pq_1) + 2yz(pq_1) \\
+ yp^2 + zq^2_1 - (x+y) m^2_{\eta} - zM^2_s \\
= (k + yp - zq_1)^2 - y^2 p^2 - z^2 q^2_1 + 2yz(pq_1) + yp^2 + zq^2_1 + (z-1) m^2_{\eta} - zM^2_s \\
\equiv l^2 - \Delta
\end{gathered}
\end{aligned}
\end{equation}
where we used that \(x + y + z = 1\) as per the Dirac delta in eq. (\ref{eq:feynparam3}). The integral
now takes the following form after changing variables,
\begin{equation}
\begin{aligned}
- 2 e P_L \sum_s h_{is} h_{js} \int^1_0 \md x \: \md y \: \md z \: \delta(x + y + z - 1) \int \frac{\mu^{4 - D}\md^D l}{(2\pi)^D}
\frac{(2l^{\mu} - 2yp^{\mu} + 2zq^{\mu}_1 + p^{\mu}) (\slashed{l} - y\slashed{p} + (z - 1)\slashed{q_1})}{(l^2 - \Delta)^3}
\end{aligned}
\end{equation}
We now need to massage this numerator a bit in order to carry out the calculation,
\begin{equation}
\begin{aligned}
\begin{gathered}
(2l^{\mu} - 2yp^{\mu} + 2zq^{\mu}_1 + p^{\mu}) (\slashed{l} - y\slashed{p} + (z - 1)\slashed{q_1}) \\
= \gamma_{\nu} (2l^{\mu}l^{\nu} - 2y l^{\mu} p^{\nu} + 2(z-1)l^{\mu}q^{\nu}_1
+ (1 - 2y)(p^{\mu}l^{\nu} - yp^{\mu}p^{\nu} + (z-1)p^{\mu}q^{\nu}_1) \\
+ 2zq^{\mu}_1 l^{\nu} - 2yz q^{\mu}_1 p^{\nu} + 2z(z-1) q^{\mu}_1 q^{\nu}_1)
\label{eq:numerator1}
\end{gathered}
\end{aligned}
\end{equation}
We know that
\begin{equation}
\begin{aligned}
\begin{gathered}
\int \md^D k \frac{k_{\mu_1}k_{\mu_2}\cdots k_{\mu_{2n+1}}}{(k^2-\Delta)^N} = 0
\end{gathered}
\end{aligned}
\end{equation}
so terms in eq. (\ref{eq:numerator1}) which are proportional to \(l^{\mu}\) vanish when integrating.
Making use of Lorentz invariance, as such,
\begin{equation}
\begin{aligned}
\int \frac{\md^D l}{(2\pi)^D} l^{\mu} l^{\nu} = \int \frac{\md^D l}{(2\pi)^D} \frac{l^2 \eta^{\mu \nu}}{D}
\end{aligned}
\end{equation}
we now get
\begin{equation}
\begin{aligned}
\begin{gathered}
\frac{1}{2}l^2\gamma^{\mu} + (1 - 2y)(- yp^{\mu}p^{\nu} + (z-1)p^{\mu}q^{\nu}_1)
- 2yz q^{\mu}_1 p^{\nu} + 2z(z-1) q^{\mu}_1 q^{\nu}_1
\end{gathered}
\end{aligned}
\end{equation}
If we reintroduce the gamma matrix, \(\gamma_{\nu}\), that was omitted in the previous
equation in all but the first term, we get
\begin{equation}
\begin{aligned}
\begin{gathered}
\frac{1}{2}l^2\gamma^{\mu} + (1 - 2y)(- yp^{\mu}\slashed{p} + (z-1)p^{\mu}\slashed{q_1})
- 2yz q^{\mu}_1 \slashed{p} + 2z(z-1) q^{\mu}_1 \slashed{q_1}
\end{gathered}
\end{aligned}
\end{equation}
and if we apply the equations of motion for the bispinors, \(\slashed{q_1}u_{l_i}(q_1, r_1) = m_i u_{l_i}(q_1, r_1)\)
and \(\slashed{q_2}u_{l_j}(q_2, r_2) = m_j u_{l_j}(q_2, r_2)\), we can get rid of these slashed momenta,
replacing them for their corresponding masses. It is worth noting that one of the differences with respect to chapter
6.3 of \cite{peskshro} is that the mass corresponding to each bispinor is not the same, thus \(\slashed{p}\) does
not vanish here.
\begin{equation}
\begin{aligned}
\begin{gathered}
\frac{1}{2}l^2\gamma^{\mu} + (1 - 2y)(- y(m_j - m_i) + (z-1)m_i) p^{\mu}
- 2yz q^{\mu}_1 (m_j - m_i) + 2z(z-1) q^{\mu}_1 m_i
\end{gathered}
\end{aligned}
\end{equation}
We shall also plug in \(q_1 = \frac{1}{2}q_1 + \frac{1}{2}q_1 = \frac{1}{2}q_1 + \frac{1}{2}(q_2 - p)
= \frac{1}{2}(q_1 + q_2 - p)\), so
\begin{equation}
\begin{aligned}
\begin{gathered}
\frac{1}{2}l^2\gamma^{\mu} + (1 - 2y)(- y(m_j - m_i) + (z-1)m_i) p^{\mu}
+ 2z(- y (m_j - m_i) + (z-1) m_i)q^{\mu}_1 \\
= \frac{1}{2}l^2\gamma^{\mu} + (1 - 2y - z)(- y(m_j - m_i) + (z-1)m_i) p^{\mu}
+ z(- y (m_j - m_i) + (z-1) m_i)(q_1 + q_2)^{\mu}
\label{eq:numerator2}
\end{gathered}
\end{aligned}
\end{equation}
If we take a look at the coefficient multiplying \(p^{\mu}\), reverting the substitution
of \(x\) for a bit,
\begin{equation}
\begin{aligned}
\begin{gathered}
(1 - 2y - z)(- y(m_j - m_i) + (z-1)m_i) = (x + y - 2y)(- y(m_j - m_i) - (x + y)m_i) \\
= -(x - y)(xm_i + ym_j) = -(x - y)\left(\frac{1}{2}(x + x)m_i + \frac{1}{2}(y + y)m_j\right) \\
= -(x - y)\left(\frac{1}{2}(x + x - y + y)m_i + \frac{1}{2}(y + y - x + x)m_j\right) \\
= -(x - y)\left(\frac{1}{2}(x + y)(m_i + m_j) + \frac{1}{2}(x - y)(m_i - m_j)\right)
\end{gathered}
\end{aligned}
\end{equation}
We run across a bit of a problem; it doesn't vanish when integrating with respect to
\(x\) and \(y\) as the term that is symmetric when exchanging both these variables
survives when integrating with respect to them. For now, though, this will be put
in the back burner.
Up until now we have ignored \(P_L\); when we multiply eq. (\ref{eq:numerator2}) with \(P_L\), we
will have a vector and an axial vector contribution. We shall now focus on the vector contribution,
and within it, calculate the integral which is a factor to \((q_1 + q_2)^{\mu}\), which is \(- \frac{\mi e}{2m_j}F_2(p^2)\).
This is as seen in \cite{Beneke_2016}, page 6, eq. (11), this being the template for our vertex integral,
shown here below
\begin{equation}
\begin{aligned}
\Gamma^{\mu} = - \mi e \overline{u}_{l_j} (q_2, r_2) \Bigg[\gamma^{\mu} F_1(p^2) +
\frac{\mi \sigma^{\mu \nu}p_{\nu}}{2 m_i}F_2(p^2) + \frac{\sigma^{\mu \nu}p_{\nu}}{2 m_i}\gamma_5 F_3(p^2)
+ (p^2 \gamma^{\mu} - \slashed{p} p^{\mu}) \gamma_5 F_4(p^2)\Bigg]u_{l_i} (q_1, r_1)
\label{eq:template}
\end{aligned}
\end{equation}
The original Gordon identity\footnote{The axial vector equivalent can be found in \cite{cheng1994gauge}, page 421, eq. (13.73)},
lets us swap \((q_1 + q_2)^{\mu}\) for \(\mi \sigma^{\mu \nu}p_{\nu}\),
\begin{equation}
\begin{aligned}
\begin{gathered}
\overline{u}(q_2) \gamma^{\mu} u(q_1) = [\slashed{p}u(p) = m u(p); \quad \overline{u}(p)\slashed{p} = \overline{u}(p) m] \\
= \frac{1}{2m} \overline{u}(q_2) \gamma^{\mu}\slashed{q_1}u(q_1) + \frac{1}{2m}\overline{u}(q_2)\slashed{q_2} \gamma^{\mu}u(q_1)
= \frac{1}{2m}\overline{u}(q_2) (q^{\mu}_1 - \mi \sigma^{\mu \nu} q_{1, \nu} + q^{\mu}_2 + \mi \sigma^{\mu \nu} q_{2, \nu}) u(q_1) \\
= \overline{u}(q_2) \left(\frac{q^{\mu}_1 + q^{\mu}_2}{2m} + \frac{\mi \sigma^{\mu \nu}p_{\nu}}{2m}\right) u(q_1)
\label{eq:gordon1}
\end{gathered}
\end{aligned}
\end{equation}
where again we have \(p = q_2 - q_1\), as well as having used \(\gamma^{\mu}\gamma^{\nu} = \eta^{\mu \nu} - \mi \sigma^{\mu \nu}\).
However, we have bispinors with different masses, and so we will use the following identities,
\begin{equation}
\begin{aligned}
\begin{gathered}
\overline{u}(q_2) \gamma^{\mu} u(q_1) = \frac{1}{m_1 + m_2}\overline{u}(q_2)[(q_1 + q_2)^{\mu} - \mi \sigma_{\mu \nu}p^{\nu}]u(q_1)\\
\overline{u}(q_2) \gamma_5 \gamma^{\mu} u(q_1) = \frac{1}{m_1 - m_2}\overline{u}(q_2) \gamma_5[(q_1 + q_2)^{\mu} - \mi \sigma_{\mu \nu}p^{\nu}]u(q_1)
\label{eq:gordon3}
\end{gathered}
\end{aligned}
\end{equation}
We must find the form factors evaluated at \(p^2 = 0\), as can be seen in eq. (\ref{eq:aral}).
Setting this greatly simplifies the final expression for the integral over \(y\).
The aforementioned integral is
\begin{equation}
\begin{aligned}
\begin{gathered}
- 2 e \sum_s h_{is} h_{js} \int^1_0 \md z \int^{1-z}_0 \md y \int \frac{\mu^{4 - D}\md^D l}{(2\pi)^D}
\frac{z(y (m_i - m_j) + (z-1) m_i)}{(l^2 - \Delta)^3}
\end{gathered}
\end{aligned}
\end{equation}
Using eq. (\ref{eq:intrenor}), we see that
\begin{equation}
\begin{aligned}
\mu^{4-D} \int \frac{\md^D l}{(2\pi)^D} \frac{1}{(l^2 - \Delta)^3} = \frac{\mu^{4 - D} (-1)^3 \pi^{D/2}}{(2\pi)^D}
\frac{\Gamma\left(3 - \frac{D}{2}\right)}{\Gamma(3)}\Delta^{\frac{D}{2} - 3}
\end{aligned}
\end{equation}
If we make \(D \rightarrow 4\), through the usual parametrization, \(D = 4 - \epsilon\), \(\epsilon \rightarrow 0\),
we obtain
\begin{equation}
\begin{aligned}
- \mi \mu^{\epsilon} \pi^{\frac{\epsilon}{2} - 2} 2^{\epsilon - 5} \Gamma\left(1 - \frac{\epsilon}{2}\right)\Delta^{-1-\frac{\epsilon}{2}}
= - \frac{\mi}{32\pi^2} \frac{1}{\Delta}
\label{eq:firkind}
\end{aligned}
\end{equation}
There is no divergence here either, since the only source of this was the evaluation of the gamma function at one
of its poles, which does not take place in this case. So, the whole integral is now
\begin{equation}
\begin{aligned}
\begin{gathered}
- 2 e \sum_s h_{is} h_{js} \int^1_0 \md z \int^{1-z}_0 \md y \: z(y (m_i - m_j) + (z-1) m_i)
\left(- \frac{\mi}{32\pi^2} \frac{1}{\Delta}\right) \\
\approx \frac{\mi e}{16\pi^2} \sum_s h_{is} h_{js} \int^1_0 \md z \int^{1-z}_0 \md y \: \frac{z(y (m_i - m_j) + (z-1) m_i)}{m^2_{\eta} + z(M^2_s - m^2_{\eta})} \\
\approx -\frac{\mi e m_i}{32\pi^2 m^2_{\eta}} \sum_s h_{is} h_{js} \int^1_0 \md z \: \frac{z(1-z)^2}{x_s z - z + 1} \\
= \frac{\mi e m_i}{16\pi^2 m^2_{\eta}} \sum_s h_{is} h_{js} \left(\frac{x_s^{2}}{x_s^{3} - 3 x_s^{2} + 3 x_s - 1} - \frac{x_s^{2} \ln x_s}{(x_s - 1)^{4}}
- \frac{2 x_s}{2 x_s^{2} - 4 x_s + 2} + \frac{1}{2 x_s^{2} - 4 x_s + 2} + \frac{1}{3 x_s - 3}\right) \\
\end{gathered}
\end{aligned}
\end{equation}
where \(x_s \equiv M^2_s/m^2_{\eta}\). Let us take this function of \(x_s\) and manipulate it further,
\begin{equation}
\begin{aligned}
\begin{gathered}
\frac{x_s^{2}}{x_s^{3} - 3 x_s^{2} + 3 x_s - 1} - \frac{x_s^{2} \ln x_s}{(x_s - 1)^{4}}
- \frac{2 x_s}{2 x_s^{2} - 4 x_s + 2} + \frac{1}{2 x_s^{2} - 4 x_s + 2} + \frac{1}{3 x_s - 3} \\
= -\frac{x_s^{2}}{(1-x_s)^3} - \frac{x_s^{2} \ln x_s }{(x_s - 1)^{4}}
- \frac{2 x_s}{2(1-x_s)^2} + \frac{1}{2(1-x_s)^2} - \frac{1}{3(1 - x_s)} \\
= \frac{- 6x_s^2(1-x_s) - 6x_s^2 \ln x_s + 3(1 - 2x_s)(1-x_s)^2 - 2(1-x_s)^3}{6(1-x_s)^4} \\
= \frac{- 6x_s^2 + 6x_s^3 - 6x_s^2 \ln x_s + 3(1-2x_s)(1-2x_s+x_s^2) - 2(x_s^3 - 3x_s^2 + 3x_s - 1)}{6(1-x_s)^4} \\
= \frac{- 6x_s^2 + 6x_s^3 - 6x_s^2 \ln x_s + 3(1 - 2x_s + x_s^2 - 2x_s + 4x_s^2 - 2x_s^3) - 2(x_s^3 - 3x_s^2 + 3x_s - 1)}{6(1-x_s)^4} \\
= \frac{1 - 6x_s + 3x_s^2 + 2x_s^3 - 6x_s^2 \ln x_s}{6(1-x_s)^4} \\
\end{gathered}
\end{aligned}
\end{equation}
which is \(F_2(x)\) as used in \cite{Vicente_2015}, eq. (13), page 7 and seen in appendix A. \(F_2(p^2=0)\) and \(F_3(p^2=0)\)
as outlined in eq. (\ref{eq:template}) are finally
\begin{equation}
\begin{aligned}
\begin{gathered}
F_2(p^2 = 0) = -\frac{m^2_j}{8\pi^2 m^2_{\eta}} \sum_s h_{is} h_{js} \frac{1 - 6x_s + 3x_s^2 + 2x_s^3 - 6x_s^2 \ln x_s}{6(1-x_s)^4} \\
F_3(p^2) = \mi F_2(p^2)
\end{gathered}
\end{aligned}
\end{equation}
% Cite these papers when discussing future research
%https://inspirehep.net/literature/1837140 https://inspirehep.net/literature/1863345
%%- la regla de Feynman del vertice foton escalar escalar no esta bien. En esa referencia que citas
%los dos escalares se aniquilan. Aqui uno entra y otro sale, de acuerdo a tu convencion de momentos.
%Entonces la regla de Feynman seria: i e (2k + q2 - q1)
%%- el paso de la 40 a la 41 no es correcto. Ademas debes de poner el proyector PL dentro de la
%integral para acordarte.
%%- Por otra parte, puedes usar EOM: barq1 u1 = m1 u1, u2 barq2 = u2 m2.
%%- Ademas en todo el calculo puedes sustituir p = q2 - q1, y tomar (dentro de Delta): p^2=0, q1^2=0, q2^2=0
%%- Nota que si hubiera que usar Gordon, no puedes tomar las masas de los fermiones iguales. Y debes tener en cuenta el proyector.
The diagram shown in fig. \ref{fig:egamma} is not the only one yielding \(l_i \rightarrow l_j\gamma\), we have two more
that do this, through a loop in one of the leptonic external lines.
\begin{figure}[H]
\centering
\begin{tikzpicture}
\begin{feynman}
\vertex (i) {\(l_i\)};
\vertex[right=1.75cm of i] (loop1);
\vertex[right=2.5cm of loop1] (loop2);
\vertex[right=1.75cm of loop2] (qed);
\vertex[right=2cm of qed] (aux1);
\vertex[above=2cm of aux1] (f1) {\(\gamma\)};
\vertex[below=2cm of aux1] (f2) {\(l_j\)};
\diagram*{
(i) -- [fermion, momentum=\(q_1\)] (loop1) -- [plain, edge label=\(N_{s}\), rmomentum'=\(k\)] (loop2) -- [fermion, momentum=\(q_1\)] (qed) -- [photon, rmomentum=\(p\)] (f1),
(qed) -- [fermion, momentum=\(q_2\)] (f2),
(loop1) -- [charged scalar, quarter left, out=90, in=90, looseness=2.0, edge label'=\(\eta^{-}\), momentum=\(q_1-k\)] (loop2),
};
\end{feynman}
\end{tikzpicture}
\begin{tikzpicture}
\begin{feynman}
\vertex (i) {\(l_i\)};
\vertex[right=2.5cm of i] (qed);
\vertex[right=1cm of qed] (aux1);
\vertex[right=3cm of qed] (aux2);
\vertex[right=4.2cm of qed] (aux3);
\vertex[below=0.9cm of aux1] (loop1);
\vertex[below=2.5cm of aux2] (loop2);
\vertex[above=4cm of aux3] (f1) {\(\gamma\)};
\vertex[below=3.2cm of aux3] (f2) {\(l_j\)};
\diagram*{
(i) -- [fermion, momentum=\(q_1\)] (qed) -- [photon, rmomentum=\(p\)] (f1),
(qed) -- [fermion, momentum=\(q_2\)] (loop1) -- [plain, edge label=\(N_{s}\), rmomentum'=\(k\)] (loop2) -- [fermion, momentum=\(q_2\)] (f2),
(loop1) -- [charged scalar, quarter left, out=90, in=90, looseness=2.0, edge label'=\(\eta^{-}\), momentum=\(q_2-k\)] (loop2),
};
\end{feynman}
\end{tikzpicture}
\caption{Additional one-loop diagrams responsible for \(l_i \rightarrow l_j \gamma\)}
\label{fig:egammaprime}
\end{figure}
The reduced amplitude corresponding to the first diagram is
\begin{equation}
\begin{aligned}
\begin{gathered}
\mi \mathcal{M} = \varepsilon^*_{\mu}(p, \lambda) \overline{u}_{l_j} (q_2, r_2)(-\mi e \gamma^{\mu})
\frac{\mi(\slashed{q_1}+m_j)}{q^2_1 - m^2_j} \\
\cdot \Bigg(\sum_{s}
\int \frac{\mu^{4-D}\md^{D}k}{(2\pi)^{D}} \frac{\mi}{(q_1 - k)^{2} - m^{2}_{\eta}} (-\mi h_{js} P_L)
\frac{\mi (\slashed{k} + M_{s})}{k^{2} - M^{2}_{s}} (-\mi h_{is} P_L)\Bigg)
u_{l_i} (q_1, r_1)
\end{gathered}
\end{aligned}
\end{equation}
and the one corresponding to the second diagram is
\begin{equation}
\begin{aligned}
\begin{gathered}
\mi \mathcal{M} = \varepsilon^*_{\mu}(p, \lambda) \overline{u}_{l_j} (q_2, r_2) \Bigg(\sum_{s}
\int \frac{\mu^{4-D}\md^{D}k}{(2\pi)^{D}} \frac{\mi}{(q_2 - k)^{2} - m^{2}_{\eta}} (-\mi h_{js} P_L)
\frac{\mi (\slashed{k} + M_{s})}{k^{2} - M^{2}_{s}} (-\mi h_{is} P_L)\Bigg) \\
\cdot \frac{\mi(\slashed{q_2}+m_i)}{q^2_2 - m^2_i} (-\mi e \gamma^{\mu}) u_{l_i} (q_1, r_1)
\end{gathered}
\end{aligned}
\end{equation}
The factor corresponding to the loop is nearly identical to what was discussed when obtaining the
mass matrix, as shown in eq. (\ref{eq:massmint}), we only have to swap \(m^2_R\) or \(m^2_I\) for
\(m^2_{\eta}\). \(q^2_i\) is however not 0, so we will have to calculate the integral shown in the
aforementioned equation again taking this into account. Instead of calculating it symbolically
through the use of SymPy and Python as done before, we shall use the result shown in
\cite{tHooft_1979}, page 7, section 4. We have that
\begin{equation}
\begin{aligned}
\begin{gathered}
\int^1_0 \md x \ln \Delta = \int^1_0 \md x \ln (q^2_1 x(x-1) - M^2_s(x-1) + m^2_{\eta}x) \\
= \ln q^2_1 - 2 + \ln \Bigg(1 - \frac{m^2_{\eta} - M^2_s}{q^2_1}\Bigg)
- \Bigg(\frac{m^2_{\eta} - M^2_s}{q^2_1}\Bigg) \ln \frac{\frac{m^2_{\eta} - M^2_s}{q^2_1} - 1}{\frac{m^2_{\eta} - M^2_s}{q^2_1}}
\end{gathered}
\end{aligned}
\end{equation}
where we have taken the limit \(q^2_1 \ll m^2_{\eta}\), \(q^2_1 \ll M^2_s\) when obtaining
\(\Delta\)'s roots. Thus, the vertex, in the first case, is
\begin{equation}
\begin{aligned}
\begin{gathered}
\Gamma^{\mu} = (-\mi e \gamma^{\mu}) \frac{\mi(\slashed{q_1}+m_j)}{q^2_1 - m^2_j}
P_L \sum_s \frac{\mi M_{s}}{16\pi^{2}} h_{is}h_{js} \Bigg[\frac{2}{\epsilon} - \gamma_{E} + \ln 4\pi\mu^2
- \ln q^2_1 + 2 \\
- \ln \Bigg(1 - \frac{m^2_{\eta} - M^2_s}{q^2_1}\Bigg) + \Bigg(\frac{m^2_{\eta} - M^2_s}{q^2_1}\Bigg)
\ln \Bigg( \frac{\frac{m^2_{\eta} - M^2_s}{q^2_1} - 1}{\frac{m^2_{\eta} - M^2_s}{q^2_1}} \Bigg)\Bigg]
\end{gathered}
\end{aligned}
\end{equation}
the second case has
\begin{equation}
\begin{aligned}
\begin{gathered}
\Gamma^{\mu} = P_L \sum_s \frac{\mi M_{s}}{16\pi^{2}} h_{is}h_{js} \Bigg[\frac{2}{\epsilon}
- \gamma_{E} + \ln 4\pi\mu^2 - \ln q^2_2 + 2 \\
- \ln \Bigg(1 - \frac{m^2_{\eta} - M^2_s}{q^2_2}\Bigg) + \Bigg(\frac{m^2_{\eta} - M^2_s}{q^2_2}\Bigg)
\ln \Bigg( \frac{\frac{m^2_{\eta} - M^2_s}{q^2_2} - 1}{\frac{m^2_{\eta} - M^2_s}{q^2_2}} \Bigg)\Bigg]
\frac{\mi(\slashed{q_2}+m_i)}{q^2_2 - m^2_i}(-\mi e \gamma^{\mu})
\end{gathered}
\end{aligned}
\end{equation}
\section{Extra dimensions}
\label{sec:extradim}
%\subsection{Einstein field equations}
%\label{sec:efe}
%
%They are a set of 10 (in 1+3 dimensional spacetime) highly non-linear equations,
%\begin{equation}
%G_{\mu \nu} + \Lambda g_{\mu \nu} = \kappa T_{\mu \nu}
%\end{equation}
%where \(T_{\mu \nu}\) is the energy-momentum (or stress-energy) tensor, \(g_{\mu \nu}\) the metric
%tensor, \(\Lambda\) the cosmological constant, \(\kappa = 8\pi G/c^4\), and the first term is
%\begin{equation}
%G_{\mu \nu} = R_{\mu \nu} - \frac{1}{2} R g_{\mu \nu}
%\end{equation}
%\(R_{\mu \nu}\) is the Ricci tensor and \(R\) is the scalar curvature. The former is defined as the
%contraction of the first and third indices of the Riemann curvature tensor, and the latter is,
%respectively,