Limits/Signals of a light non-standard
Higgs boson in Upsilon leptonic decays†
Miguel A. Sanchis-Lozano*
*email: mas@ific.uv.es
Department of Theoretical Physics & IFIC
University of Valencia – CSIC
Spain
† Special
thanks to the CLEO Collaboration
particularly to Jean Duboscq for many discussions
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
1
Testing Lepton Universality
BF( e+ e-) =BF( + -) = BF(  +  -)
Channel: *
BF[e+e-]
(1S)
2.52 ± 0.08 %
(1S)
(2S)
2.01 ± 0.16 %
(2S)
(3S)
1.92 ± 0.24 %
(3S)
BF[+-]
R/l
2.54 ± 0.13 %
0.01 ± 0.06
2.49 ± 0.07 % 2.54 ± 0.13 %
0.02 ± 0.05
2.11 ± 0.15 %
0.05 ± 0.11
2.03 ± 0.09 % 2.11 ± 0.15 %
0.04 ± 0.06
2.55 ± 0.24 %
0.33 ± 0.18
2.39 ± 0.12 % 2.55 ± 0.24 %
0.07 ± 0.09
BF[μ+ μ-]
*Obtained from recent but in some cases preliminary CLEO data
Statistical and systematic errors are summed in quadrature
Lepton Universality
in Upsilon decays implies
<R/l > =0
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
R /  
Miguel A. Sanchis-Lozano
IFIC-Valencia
 ( nS ) s
( em )

B  B B

1
B
B
2
Testing Lepton Universality
BF( e+ e-) =BF( + -) = BF(  +  -)
Channel: *
BF[e+e-]
(1S)
2.38 ± 0.11 %
(1S)
(2S)
1.97 ± 0.11 %
(2S)
(3S)
1.92 ± 0.24 %
(3S)
BF[+-]
R/l
2.61 ± 0.13 %
0.10 ± 0.07
2.48 ± 0.06 % 2.61 ± 0.13 %
0.05 ± 0.06
2.11 ± 0.15 %
0.07 ± 0.09
1.93 ± 0.17 % 2.11 ± 0.15 %
0.09 ± 0.12
2.55 ± 0.24 %
0.33 ± 0.18
2.18 ± 0.21 % 2.55 ± 0.24 %
0.17 ± 0.15
BF[μ+ μ-]
*Obtained from recent but preliminary CLEO data and already existing PDG values
Statistical and systematic errors are summed in quadrature
Lepton Universality
in Upsilon decays implies
<R/l > =0
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
R /  
Miguel A. Sanchis-Lozano
IFIC-Valencia
 ( nS ) s
( em )

B  B B

1
B
B
3
LU
Lepton Universality Breaking?
R/e(1S)
Hypothesis test
Null hypothesis: lepton universality
predicting <R/l> = 0
Alternative hypothesis: <R/l> > 0
R/(1S)
From this set of data
Lepton Universality can be rejected
at a level of significance of 0.5 %
(corresponding to a 2.8 sigma effect)
R/e(2S)
R/(2S)
R/e(3S)
R/(3S)
0.0
0.1
0.2
0.3
0.4
R/l
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
4
(Hidden) systematic errors?
There could be hidden systematic errors in the extraction of the muonic and tauonic
branching fractions from experimental data, e.g. use is made of lepton universality as
an intermediate step
Bee = B= B
B
~
B
  /  
~
1  3B
Defining:
 B

B   / had
~
B

1  Bee  B  B 
The muonic branching fraction would be overestimated if
Bee = B= B
~
B
B   /  
~
1  3B
Defining:
hep-ex/9409004
hep-ex/0409027
~
~
B  B

B   / had
~
B
 B 
1  Bee  B  B 
The tauonic branching fraction would be underestimated if
~
~
B  B
Besides phase space disfavors the tauonic decay mode by  1% (Van-Royen Weisskopf formula)
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
5
Searching for a light non-standard Higgs
Our conjecture to interpret a possible Lepton Universality breakdown
photon unseen! *
Van Royen-Weisskopf formula
nS  s A0(     )
nS  s b(*) ( A0*     )
2
Rn (0)
 (1  2 x )(1  4 x )1/ 2
2



M
(smoothly) decreasing function of x
with x  m2 / M 2
( em )  4 2Qb2
n=1,2,3 and A0 stands for a light (possibly CP-odd) Higgs particle
Tree-level NP process
theoretically very simple!
Key ideas:
Such NP contribution would be unwittingly ascribed to
the leptonic branching fraction (photon undetected!*)
actually only for the tauonic decay mode
A leptonic mass dependence of the decay width would
break lepton universality. Contribution from the SM
should be negligible
*Missing energy in events (neutrinos from tauonic decays)
but photons can be searched for in data recorded on tape
However, they likely would not be monochromatic!
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
Prior M1 transition

b
-
s
A0
b
+
b intermediate
state or bb continuum
Prejudice against a light Higgs?
Notice that a light non-standard Higgs
boson has not been excluded by LEP
direct searches for a broad range of
model parameters in different scenarios
6
Two-Higgs Doublet Model (II)*
Higgs sector






H
H
u
d
ˆ
ˆ



H u   0  , H d   0 
 Hu 
 Hd 
*The MSSM is a particular 2HDM(II) realization
CPC: 1 CP-odd Higgs (A0)
2 CP-even Higgses
Coupling of fermions and the CP-odd Higgs A0
In a 2HDM of type II
A0
L   tan 
m f d (i 5 ) d , d  d , s, b, e,  ,
v
A0
f f
Lint   cot 
m f u (i 5 ) u , u  u, c, t , e ,  , 
v
f f
int
h0 =-1sin +2cos
H0= 1sin -2cos
2 charged H
At tree level:
M2H  M2Ao  M2W
1
In a general 2HDM: M 2H   M 2A 0  5  4  v 2
2
tanβ stands for the 2 Higgs VEVs ratio
<Hu>/<Hd> and v=246 GeV
A large value of tanβ would imply a large A0 coupling to the bottom quark but a small coupling
(i.e. small cot β) to the charm quark. Therefore, in the quest for NP effects we will focus on
bottomonium decays and spectroscopy but not on charmonium as this is the case (see PDG)
B( '  ee)  (7.41  0.28) 103  B( '   )  (7.3  0.8) 103  B( '   )  (2.8  0.7) 103
CPV MSSM
At one-loop level, MSSM with complex parameters is not CP conserving
The three Higgs neutral bosons mix together and the resulting three physical mass eigenstates
H1, H2 and H3 (MH1 < MH2 < MH3) have mixed parities
Higgs couplings to the Z boson would vary: the H1ZZ coupling can be significantly suppressed
raising the possibility of a relatively light H1 boson having evaded detection at LEP [hep-ph/0211467]
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
7
Light Higgs windows at LEP (2HDM(II))
hep-ex/0408097
Excluded (mA,mh) region independent of the CP even Higgs
mixing angle , from flavor-independent and b-tagging
searches at LEP interpreted according to a 2HDM(II)
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
hep-ex/0408097
Excluded regions in the (mA,tanβ) plane for different choices of .
In the MSSM -/2  0 ; in a general 2HDM(II) -/2  /2
Miguel A. Sanchis-Lozano
IFIC-Valencia
8
Light Higgs windows at LEP (MSSM CPV)
hep-ex/0406057
hep-ex/0406057
Diagram illustrating the effective coupling of a Higgs
mass eigenstate H1 to the Z. Only the CP-even admixture
h and H couple to Z while the CP-odd A does not: hence
the coupling of the H1 is reduced wrt a CPC scenario.
hep-ex/0406057
CPX MSSM 95% exclusion areas using scans with different values
of arg (At,b). The region excluded by Yukawa searches, Z-width
constraints or decay independent searches is shown in red
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
CPX MSSM 95% exclusion areas using scans with different
values of the top mass
Miguel A. Sanchis-Lozano
IFIC-Valencia
9
Next-to-Minimal Supersymmetric Standard Model (NMSSM)
A new singlet superfield is added to the Higgs sector:
In general more extra SM singlets can be added: hep-ph/0405244






H
H
u
d
Hˆ u   0  , Hˆ d   0  , Sˆ
 Hu 
 Hd 
The μ-problem of the MSSM would be solved by introducing in the superpotential the term


WHiggs   Sˆ ( Hˆ u Hˆ d )  Sˆ 3  Vsoft   A S ( H u  H d )  A S 3  h.c.
3 Spontaneous breaking of the PQ symmetry 3
where   x, x  S   / 
Breaks explicitly the PQ symmetry
If к =0 → U(1) Peccei-Quinn symmetry
Spontaneous breaking  NGB (massless), an “axion” (+QCD anomaly) ruled out experimentally
If the PQ symmetry is not exact but explicitly broken  provides a mass to the (pseudo) NGB leading
to a light CP-odd scalar for small к
If λ and  zero → U(1)R symmetry; if U(1)R slightly broken → a light pseudoscalar Higgs boson too
A1  cos  A A MSMS  sin  A As
Higgs sector in the NMSSM: (seven)
2 neutral CP-odd Higgs bosons (A1,2)
3 neutral CP-even Higgs bosons (H1,2,3 )
2 charged Higgs bosons (H±)
The A1 would be the lightest Higgs:
M
2
A1
 
 3   A 

Favored decay mode: H1,2  A1 A1
hard to detect at the LHC [hep-ph/0406215]
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Coupling of A1 to down type fermions:

m 2f v
x
 ,  cos  A tan 
[hep-ph/0404220]
A  2x
v2
2
cos  A  2

,


x tan 2 
A  x
2
Miguel A. Sanchis-Lozano
IFIC-Valencia
10
Contribution from the “continuum”

Let us perform a perturbative calculation of the contribution from the “continuum”, i.e.
without formation of intermediate b b bound states assuming the Higgs boson off-shell
   l  l   
1
1
32M 2 (2 )3

2
A(   l  l  ) dm2 dm2
l+
A0
l-
M Y  M A0
where the square amplitude is
and the decay width (leading term):
 (nS )       
2
 
 Rn (0) tan 4    M 2A

  1  m2 , M A  M 

log

2 
 2
144 3v 4
  M A  M   
0
0
0
M 2 tan 4    M 2A0
R 
log
64   3v 4   M 2A 0  M 2
 
  1  m2

 
Rather large values of tanβ ( 50)
are needed to obtain R ~ 10%
m2 tan 2 
[A    ] 
M A 0 (1  4 x )1/ 2  B(A0  + -)  1
2
8 v
even for moderate tanβ
nS   A0 ( + -)
One should also consider:

0
Higgs boson on-shell M A  M Y
0
Experimentally excluded region    +”nothing”:
BF < 1.3 x10-5 for MA0 < 5 GeV
[CLEO, PRD 51, 2053 (1995)]
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL

2
M  tan 2   M 2A0
1  2
R 
2
8  v 
M
Wilczek formula
Miguel A. Sanchis-Lozano
IFIC-Valencia




MY-MAº=0.25 GeV
R=0.1
tanβ ~ 15
Somewhat higher values of tanβ
in a relativistic treatment
11
Intermediate bound states
We consider that a prior magnetic dipole transition of the Upsilon yields an intermediate
bound state, subsequently annihilating into a lepton pair via A0 (or H1)-exchange
nS  s b ( A0* (H1*)  l+ l-)
In a 2HDM(II)
Intermediate real 0-+ state
3mb4 m2 (1  4 x )1/ 2 Rn (0) tan 4 
 
( b    ) 
2 2 (M2b  M 2A 0 ) 2 v 4
2
The NP contribution would almost saturate the b decay rate
even for moderate tanβ if M is not too large !
Mass difference between
the Higgs boson and the b
A cascade decay formula should apply:
B(  s     )  B(  s b )  B( b      )
B(  s b ) 
B( b      )  1
________Y(3S)
-------------b(3S)
Allowed and hindered M1
transitions between
 600 MeV
________Y(2S)
(3S), (2S) and (1S)
-------------
b(2S)
states
 1000 MeV
________Y(1S)
-------------b(1S)
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL

1 4 I Qb k 3



3mb2
2
M1 transition probability
M  M A 0  M b
mb2 tan 4  k 3
m2
R  2

8  (1  2 x ) b v 4 M 2
M1b
P  ( b* s )  10 3  10 4
B( (nS )      )  B( (nS )   η b (n' S ))
 0.27% (95% CL)
CLEO limit
M  (1S )  M b  b  100 MeV
 100 MeV
R 
B(   s )
 1  10 %
Bee
Miguel A. Sanchis-Lozano
IFIC-Valencia
tanβ ≥ 30
ΔM=0.25 GeV
12
CP-even Higgs boson
A CP-even Higgs boson can couple to an intermediate b0 resonance after a E1 transition
of the 
Intermediate real 0++ state
Cascade decay via b0 resonances
h - bb/   - coupling : -
sin 
 sin(    )  tan  cos(    )
cos
nS   b0 ( h0  + -)
n=2,3
( H1  + -)
Decay channel only open
for (2S) and (3S) resonances
h0-mediated decay width under the assumption that sin(-β) ≈ 0
i.e. non-decoupling limit: h0 couplings deviate maximally from SM
whereas the H0 becomes a SM-like Higgs:
2
(  b 0      ) 
27mb2 m2 (1  4 x )3 / 2 Rn' (0) tan 4 
8 2 (M 2b 0  M 2A0 ) 2 v 4
Again only for the tauonic mode
would the NP contribution be significant
leading to lepton universality breaking!
   b 0  gg  
Normalizing the above partial width to
b0
setting s(M) = 0.15 and |R’n(0)|2 = 1.5 GeV5

96 s2 Rn' (0)
2
M 4b 0
 b 0  320 keV
Using ΔM=0.25 GeV, tanβ  15 → B(b0+ -) ≈ 6% and B(   b0) ≈ 3-6% (PDG) as reference values, we get:
B(   s    )  B(  s  b0 )  B(  b0     )  0.002  0.003
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
R  0.1
tanβ ~ 15
13
Possible spectroscopic consequences in bottomonium
“Is there any point to which you would wish to draw my attention?”
“To the curious incident of the dog in the night-time”
“The dog did nothing in the night-time”
“That was the curious incident”, remarked Sherlock Holmes
from “Silver Blaze” by Sir A.C.D.
Searches for
b states at CLEO:
hep-ex/0207057
No signal found!
hep-ex/0207057
hep-ex/0207057
If b decay happens to be
saturated by the
0
A -mediated channel
b A0* + A quite broad b expected!
The search for b states would
be a way of hunting
a light non-standard Higgs!
Also implying lower rates
for other decay modes:
b  J/ψ J/ψ
(3S)  ’bo  b(1S)
• Broad b states? Widths of order 50 MeV could be expected
even for moderate values of tanβ
5
mb  s (mb )
SM:  
  5 MeV
mc  s 5 (mc )
• A0-b mixing? Mass shift yielding larger hyperfine splitting than
Drees and Hikasa,
expected in the SM (i.e. more than ~100 MeV).
PRD 41, 1547 (1990)
b
c
tanβ
Most theoretical models
are ruled out !
(LEP)
hep-ex/0111010
Higgs Yukawa couplings to fermions can
be affected too, allowing quite large tanβ
for special Higgs mass values
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
14
Conclusions
Based on:
hep-ph/0510374
hep-ph/0503266
hep-ph/0407320
hep-ph/0307313
hep-ph/0210364
hep-ph/0206156
If lepton universality is found to be (slightly) broken in Upsilon decays, the simplest
explanation would require a light non-standard mediated Higgs contribution (subsequent to
a M1 transition of the  resonance) unwittingly ascribed to the tauonic decay mode
● Reasonable values of tanβ in a 2HDM(II) are needed to yield R/l of order 10%
● Different scenarios beyond the SM could explain such a LU breaking
● Further exp tests suggested: - search for soft (non-monochromatic) photons in the
+- sample of Upsilon decays
- seek after the b resonance!
- polarization studies of the tau…
If not, those light Higgs mass windows could be closed! Possibly too hard a job for the LHC
Relevance of a Super B Factory running on the  region !!!
Complementary to other searches at the LHC
Avec de l’audace, on peut tout entreprendre, on ne peut pas tout faire
Keep also an eye on the related issues:
Napoléon Ier
● g-2 muon anomaly (which might require a CP-odd light Higgs contribution in a two-loop calculation)
● Search for Bs,d  ,  decays (GIM mechanism can spoil the expected signal hep-ph/0209306)
● Searches for dark matter in the universe (McElrath, this workshop)
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
15
Light Neutralino Dark Matter
•
Neutralino dark matter is generally assumed to be relatively heavy, with a mass near
the electroweak scale
•
It has been shown, however, that the claims of dark matter detection made by DAMA
can be reconciled with null results of CDMS II if one considers a WIMP lighter than
approximately 10 GeV [hep-ph/0504010]
•
A very light CP-odd Higgs can make possible a very light neutralino to annihilate via
such a light Higgs boson exchange [hep-ph/0509024]
~ 0 ~ 0  A0  pions  muons  positrons
leading to low-enery positrons which would generate the 511 keV gamma-ray
emission observed from the galactic bulge by INTEGRAL/SPI experiment
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
16
g-2 muon anomaly in a 2HDM(II)


hep-ph/0302111

h,A
One.loop vertex correction

f
h,A



Two-loop vertex correction
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
The 2σ allowed regions in the (mA,mh) plane due to the constraints of a
and Rb for tanβ=40. The blue region is where the total 2 is less than 4
while the yellow region is where the total 2 is less than the 2 (SM)
Miguel A. Sanchis-Lozano
IFIC-Valencia
17
MSSM (CPV)
•
•
•
•
•
The Higgs potential is explicitly CP-conserving if all Higgs potential parameters are
real. Otherwise, CP is explicitly violated
CP-violation is generated by loop effects involving top and bottom squarks
CP-violation can modify drastically the tree-level couplings of the Higgs particles to
fermions and gauge bosons
The CP parities of the neutral Higgs bosons may be mixed by radiative effects
involving the third generation of squarks
Under CP violation, the mass of the CP-odd Higgs scalar is no longer an eigenvalue
but rather an entry in a general 3 x 3 neutral Higgs mass matrix
MSP2  mt4 Im(At) / v2 MSUSY2
where
 is the mixing parameter of the two Higgs superfields
At denotes the soft SUSY-breaking trilinear coupling of the Higgs bosons to top
squarks
v=246 GeV VEV of the Higgs
MSUSY stands for the common SUSY-breaking scale
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
18
Intermediate bound states
We consider that a prior magnetic dipole transition of the Upsilon yields an intermediate
bound state, subsequently annihilating into a lepton pair via A0 (or H1)-exchange
nS  s b ( A0* (H1*)  l+ l-)
In a 2HDM(II)
Intermediate real 0-+ state
3mb4 m2 (1  4 x )1/ 2 Rn (0) tan 4 
( b    ) 
2 2 (M2b  M 2A 0 ) 2 v 4
2
A cascade decay formula applies:
B(  s     )  B(  s b )  B( b      )
 
The NP contribution would almost saturate the b decay rate
even for moderate tanβ if M is not too large !
Mass difference between
the Higgs boson and the b
B(  s b ) 
M  M A 0  M b
M1b

1 4 IQb k 3



3mb2
2
M1 transition probability
hep-ph/0407320
Allowed and hindered M1 transitions
between (3S), (2S) and (1S)
P  ( b* s )  10 3  10 5
B( (nS )      )  B( (nS )   ηb (n' S ))
Range of tanβ for
a given M value
________Y(3S)
-------------b(3S)
 600 MeV
________Y(2S)
-------------b(2S)
 1000 MeV
________Y(1S)
-------------b(1S)
ΔM (GeV)
4th Int Workshop on Heavy
Quarkonium June 27-30 BNL
Miguel A. Sanchis-Lozano
IFIC-Valencia
19