Prepared for submission to JHEP CPHT-RR084.

112017
LPT-Orsay-17-48
IFIC/17-49

Update of global Two-Higgs-Doublet model fits

arXiv:1711.02095v1 [hep-ph] 6 Nov 2017

Debtosh Chowdhury,a,b,c Otto Eberhardtd

a
Istituto Nazionale di Fisica Nucleare, Sezione di Roma, Piazzale Aldo Moro 2, I-00185 Roma,
Italy
b
Centre de Physique Theorique, Ecole Polytechnique, F-91128 Palaiseau Cedex, France
c
Laboratoire de Physique Theorique, Universite Paris-Sud, F-91405 Orsay Cedex, France
d
Instituto de Fsica Corpuscular, Parque Cientfico, C/Catedratico Jose Beltran, 2, E-46980 Pa-
terna, Spain
E-mail: [email protected],
[email protected]

Abstract: We perform global fits of Two-Higgs-Doublet models with a softly broken

Z2 symmetry to recent results from the LHC detectors CMS and ATLAS, that is signal

strengths and direct search limits obtained at s = 8 TeV and s = 13 TeV. We combine
all available ATLAS and CMS constraints with the other relevant theoretical and exper-
imental bounds and present the latest limits on the model parameters. We obtain that
deviations from the so-called alignment limit = /2 cannot be larger than 0.079
in type I and have to be smaller than 0.029 in the remaining three types. For the latter
we also observe lower limits on the heavy Higgs masses in the global fit. The splittings
between these masses cannot exceed 200 GeV in the types I and X and 130 GeV in the
types II and Y. Finally, we find that the decay widths of the heavy Higgs particles cannot
be larger than 7% of their masses if they are lighter than 1.5 TeV.
Contents

1 Introduction 1

2 Model 2

3 Constraints and fitting set-up 3

4 h signal strengths 9

5 Heavy Higgs searches 12

6 Combination of all constraints 25

7 Conclusions 28

A Prior dependence of the massive parameters 31

1 Introduction

The discovery of a new scalar resonance with a mass around 125 GeV [1, 2] in the Run
1 phase of the Large Hadron Collider (LHC) has paved the way for new directions in
high-energy particle physics. Analyzing the properties of this particle has suggested strong
evidence that it is the Higgs boson of the Standard Model (SM), i.e. a scalar CP-even
state which have SM-like coupling to the other particles. Currently the combined analysis
based on the Run 1 (7 and 8 TeV) LHC data shows that its couplings with the vector
bosons are found to be compatible with those expected from the SM within a 10%
uncertainty, whereas the coupling to the third generation fermions (top, bottom quarks
and the lepton) is compatible within an uncertainty of 15 20% [3]. Thus the current
status of the Higgs properties still allows to explore new interpretations of the observation
coming from new physics of different underlying structures.
The Two-Higgs-doublet model (2HDM) [46] is one of such extensions of the Standard
Model. Alike other popular NP models, it gets more and more constrained by recent
experimental progress, especially by the LHC data [732]. As the name suggests the
2HDM has two Higgs doublets in contrast to the single Higgs doublet in the SM. This
extension of the Higgs sector leads to the existence of five scalar bosons, namely a heavy
and light CP-even Higgs boson, H and h, a CP-odd Higgs boson, A, and a pair of charged
Higgs bosons, H . Whether the scalar boson observed in the Run 1 of LHC is a part of
an extended Higgs sector is an outstanding question and is at the cynosure of attention of
the current Run 2 (13 TeV) phase of the LHC.

1
 The questions we ask is: Which parts of the 2HDM parameter space are favoured
after imposing the latest experimental data from the LHC? Compared to the Run 1 phase
of the LHC, in Run 2 the situation has changed in several respects: Alongside the latest
Higgs signal strength data by the ATLAS and CMS collaborations also many of the recent
results of searches for additional heavy Higgs bosons are more constraining than the Run
1 data. Both experiments have performed dedicated searches for new signatures in various
possible final states at the LHC: besides the fermionic final states tt, bb, + , tb and
+ they include gauge bosons (, Z, ZZ, W + W ) and Higgs particles (hh, hZ) as
the decay products of a heavy resonance. So far, these searches for heavy resonances have
remained elusive in the ATLAS and CMS data, and thus the measurements put model-
independent 95% C.L. upper limits on the production cross section times branching ratios
for different production processes and decay modes. In the present work, we assess the
status of the all four types of softly broken Z2 symmetric 2HDM when all the experimental
constraints coming from the latest LHC data are taken into account. We confront these
with the theoretical constraints on these models (positivity, stability and next-to-leading
order unitarity). Furthermore, we perform global Bayesian fits to all relevant constraints
on these models, which also include electroweak precision and flavour observables, and
highlight the complementarity between them.
This paper is organized as follows: The 2HDM is defined in Section 2. In Section 3
we list all relevant constraints and explain the fitting set-up. The results are presented in
the subsequent sections, first taking into account the Higgs signal strengths in Section 4
and the direct searches in Section 5, before combining them with the other constraints in
Section 6. We conclude in Section 7. In Appendix A we explain how we treat the prior
dependence of the massive parameters.

2 Model

The Two-Higgs-Doublet model with a softly broken Z2 symmetry is characterized by the

following scalar potential:

V = m211 1 1 + m222 2 2 m212 (1 2 + 2 1 ) + 12 1 (1 1 )2 + 12 2 (2 2 )2

h i
+ 3 (1 1 )(2 2 ) + 4 (1 2 )(2 1 ) + 21 5 (1 2 )2 + (2 1 )2 , (2.1)

where 1 and 2 are the two Higgs doublets. While writing the potential we have assumed
that the scalar potential is CP conserving. Instead of the eight potential parameters from
Eq. (2.1) we will use the physical parameters in the rest of this article. They consist of the
vacuum expectation value v, the CP-even Higgs masses mh and mH , the CP-odd Higgs
mass mA , the mass of the charged Higgs, mH + , the two diagonalization angles and ,
and the soft Z2 breaking parameter m212 . Assuming the observed scalar of mass 125
GeV at the LHC to be the lighter CP-even Higgs h, the first two of these can be treated
as fixed. The rest of the scalar masses could in general even be lighter than 125 GeV, they
are not necessarily in the decoupling limit [5]. Keeping in mind the discovery potential of
the HL-LHC, in the following we will consider them to be in the range between 130 GeV

2
and 1.6 TeV, that is beyond the region where the 125 GeV scalar was found. Moreover,
we trade the angles and with and tan , since they can be directly related to
physical observables. All SM parameters were fixed to their best-fit values [33].
After neglecting the first two generations of fermions, the Yukawa part of the 2HDM
Lagrangian reads as follows:

LY = Yt QL i2 2 tR Yb,1 QL 1 bR Yb,2 QL 2 bR Y,1 LL 1 R Y,2 LL 2 R + h.c.

In the above Lagrangian, by convention the top quark only couples to 2 ; its Yukawa cou-
pling is related to the SM value YtSM by Yt = YtSM / sin . With an unbroken Z2 symmetry
in the Yukawa sector, there are only four possibilities through which the Higgs fields couple
to the bottom quark and tau lepton at tree-level. They are called type I, type II, type X
or lepton specific and type Y or flipped. In Table 1 we categorize the corresponding
Yukawa coupling assignments.

Table 1. Yukawa coupling assignments in the four possible Z2 symmetric 2HDM types.

Type I Type II Type X (lepton specific) Type Y (flipped)

Yb,1 = Y,1 = 0 Yb,2 = Y,2 = 0 Yb,1 = Y,2 = 0 Yb,2 = Y,1 = 0
Yb,2 = YbSM / sin Yb,1 = YbSM / cos Yb,2 = YbSM / sin Yb,1 = YbSM / cos
Y,2 = YSM / sin Y,1 = YSM / cos Y,1 = YSM / cos Y,2 = YSM / sin

3 Constraints and fitting set-up

Our statistical analysis of the 2HDM is a Bayesian fit, in which the following priors are
used for the previously defined parameters:

1.1 log(tan ) 1.7 (equivalent to 0.08 tan 50),

0 ,
130 GeV mH , mA , mH + 1.6 TeV,
(1.6 TeV)2 m212 (1.6 TeV)2

Extreme tan values outside the chosen prior are expected to be excluded due to the
absence of strong 2HDM effects in certain flavour observables (see e.g. reference [34]); the
aforementioned interval is a very conservative estimate. The only implicit assumption we
make is that the 125 GeV scalar is the light CP-even Higgs particle of the 2HDM and that
the other scalars should be heavier, yet in LHC reach.
The focus of this article is on LHC Higgs observables, that is h signal strengths and searches
for H, A and H + . Most details of the implementation of the corresponding observables
can be found in our last article [31]. The modifications to this will be explained in the
following.

3
 Signal Value Correlation matrix
strength

ggF 1.10 0.23 1 -0.25 0 -0.14 0

VBF 1.3 0.5 -0.25 1 0 0 0

Wh 0.5 1.3 0 0 1 -0.64 0

Zh 0.5 2.8 -0.14 0 -0.64 1 -0.11

tth 2.2 1.5 0 0 0 -0.11 1
ZZ
ggF 1.13 0.33 1 -0.26
ZZ
VBF 0.1 0.9 -0.26 1
W W
ggF 0.84 0.17 1 -0.16
W
VBF
W 1.2 0.4 -0.16 1
ggF
1.0 0.6 1 -0.37 0 -0.25 0 -0.21
VBF
1.3 0.4 -0.37 1 0 0 0 0
W
Wh
W 1.6 1.1 0 0 1 -0.12 -0.12 0
Wh 1.4 1.4 -0.25 0 -0.12 1 0 0
W
Zh
W 5.9 2.4 0 0 -0.12 0 1 0
Zh 2.2 2.0 -0.21 0 0 0 0 1
W
tth
W 5.0 1.8 1 -0.47
tth
1.9 3.5 -0.47 1
bb
Wh 1.0 0.5
bb
Zh 0.4 0.4
bb
tth 1.1 1.0

pp 0.1 2.5

Table 2. h signal strengths from Table 8, Table 13 and Figure 27 of the official ATLAS and CMS
combination for Run 1 [3], based on 25 fb1 of integrated luminosity. We neglect correlations below
0.1. The colours in the first column indicate the decay category in Figures 1 and 2.

For the signal strengths, we define decay

production , where production stands for the ggF,
VBF, Vh, Zh, Wh, tth or pp production channels of the h, while decay denotes the
subsequent h decay products , ZZ, W W , , bb, or Z.1 For the last one, only
upper limits are available; we assign to this signal strength a central value of 0 and adjust
the Gaussian error such that the likelihood distribution has the 95% limit at the value
provided by the experimental collaborations. All h couplings are calculated at leading
order: While the fermionic decays and the bosonic decays to W W and ZZ are possible at
1
In order to improve readability, we drop charge or conjugation labels when there is no ambiguity.

4
 Signal Value Correlation matrix L Source
strength [fb1 ]

ggF 0.80 0.19 1 -0.29 -0.22 0

VBF 2.1 0.6 -0.29 1 0 0 [36, 37]
36.1

Vh 0.7 0.9 -0.22 0 1 -0.14

tth 0.5 0.6 0 0 -0.14 1
ZZ
ggF 1.11 0.24 1 -0.29 -0.22 0
ZZ
VBF 4.0 1.6 -0.29 1 0 0 [37, 38]
36.1
ZZ
Vh 0 1.9 -0.22 0 1 -0.14
ZZ
tth 0 3.9 0 0 -0.14 1
W W
VBF 1.7 1.0 5.8 [39]
W
Wh
W 3.2 4.3 5.8 [39]
bb
Vh 1.20 0.39 36.1 [40]
bb
tth 2.1 1.0 13.2 [41]

pp 0.1 1.5 36.1 [42]
Z
pp 0 3.4 36.1 [43]

Table 3. Run 2 h signal strengths measured by ATLAS. Again, correlations below 0.1 were treated
to be 0. The colours in the first column correspond to the ones in Figures 1 and 2. In the fifth
column, we highlight the underlying integrated luminosity with red, yellow and green, depending
on whether the measurement is based on few, moderate or the full Run 2 data.

tree-level, we apply one-loop expressions for the decays into final states including massless
bosons (that is gg, and Z) [35]. A list of the available experimental signal strength
values from LHC Run 1 and 2 can be found in the Tables 2 (ATLAS and CMS combination
for Run 1), 3 (ATLAS numbers for Run 2) and 4 (CMS measurements for Run 2). For
the Run 2 data, we also list the corresponding integrated luminosities L. The numbers
for the correlations in Tables 2 and 3 can be found in the mentioned documents. ATLAS
provides correlations only for the combination of the and ZZ decays; observing very
similar numbers in the corresponding Run 1 data, we assume identical correlations for the
and ZZ final states. CMS has not published the numerical values for the correlations.
We reconstructed them from the signal strength contours (or cross section times branching
ratio contours) in the plane of VBF vs. ggF production. In Section 4 we discuss the
individual impact of the signal strengths on the 2HDM parameters, ordered by the decay
products.
Concerning the direct searches for the heavy CP-even, the CP-odd and the charged
Higgs, we have updated the number of used LHC analyses from 16 in [31] to 48 from Run
1 and 2 in the present article. We calculate the product of the production cross section

5
 Signal Value Correlation matrix L Source
strength [fb1 ]

ggF 1.11 0.19 1 -0.32
35.9 [44]

VBF 0.5 0.6 -0.32 1

Vh 2.3 1.1 35.9 [44]

tth 2.2 0.9 35.9 [44]
ZZ
ggF 1.20 0.22 1 -0.43
35.9 [45]
ZZ
VBF 0.06 1.03 -0.43 1
ZZ
Vh,h 0 2.85 35.9 [45]
ZZ
Vh,l 0 2.78 35.9 [45]
ZZ
tth 0 1.19 35.9 [45]
W W
ggF 1.02 0.27 1 -0.24
15.2 [46]
W W
VBF+Vh 0.89 0.67 -0.24 1
bb
pp 2.3 1.7 35.9 [47]
ggF
1.05 0.48 35.9 [48]
VBF+Vh
1.07 0.44 35.9 [48]
tth
0.72 0.58 35.9 [49]

Table 4. Run 2 h signal strengths measured by CMS. The colours in the first column correspond
to the ones in Figures 1 and 2, the ones in the fifth column to the amount of underlying data like
in Table 3.

[98107] and the branching ratio [108, 109] of a specific decay, B. In order to compare
it with the experimental bounds, we assign Gaussian likelihoods with a central value of 0
to the ratio of the theoretical value and the observed upper limit of B. This method
agrees with the treatment of the upper limit of the Z signal strength mentioned above and
coincides with our approach in [31] under the assumption that the observed upper limit
does not deviate from the expected one. With no evidence for such a deviation in any of
the searches, this approximation seems to be justified. The experimental input from Run 1
and 2 can be found in Tables 5 and 6, respectively. These analyses comprise a large variety
of searches for heavy resonances decaying into fermionic or bosonic states: bb, , and
Z limits can be applied to both, the CP-even and CP-odd Higgs bosons; signatures with
two massive bosons can exclusively stem from H decays at tree-level in the 2HDM, while
a CP-odd resonance decaying to one h and one Z is interpreted as A; finally, the searches
for charged Higgs particles were performed looking for the final states tb or . If the
branching ratio into a specific final state like for instance ()(bb) is included in the
upper limit, we list it in the table. If the final state is not included in the B limits,
but its information is needed to distinguish it from other searches, we write it in square

6
 Label Channel Experiment Mass range L
[GeV] [fb1 ]
bb
C8b bb H/A bb CMS [50] [100;900] 19.7
A8 ATLAS [51] [90;1000] 19.5-20.3
gg H/A
C8 CMS [52] [90;1000] 19.7
A8b ATLAS [51] [90;1000] 19.5-20.3
bb H/A
C8b CMS [52] [90;1000] 19.7
A
8 gg H/A ATLAS [53] [65;600] 20.3
AZ
8 ATLAS [54] [200;1600] 20.3
pp H/A Z (``)
C8Z CMS [55] [200;1200] 19.7
AZZ
8 gg H ZZ ATLAS [56] [140;1000] 20.3
AZZ
8V V V H ZZ ATLAS [56] [140;1000] 20.3
AW8
W
gg H W W ATLAS [57] [300;1500] 20.3
WW
A8V V V H WW ATLAS [57] [300;1500] 20.3
C8V V pp H V V CMS [58] [145;1000] 24.8
C84b pp H hh (bb)(bb) CMS [59] [270;1100] 17.9
C822b pp H hh ()(bb) CMS [60] [260;1100] 19.7
Ahh
8 gg H hh ATLAS [61] [260;1000] 20.3
C82b2 pp H hh[ (bb)( )] CMS [62] [300;1000] 18.3
AbbZ
8 gg A hZ (bb)Z ATLAS [63] [220;1000] 20.3
A8 Z gg A hZ ( )Z ATLAS [63] [220;1000] 20.3
C82b2` gg A hZ (bb)(``) CMS [64] [225;600] 19.7
C82 2` gg A hZ ( )(``) CMS [65] [220;350] 19.7
A8 pp H ATLAS [66] [180;1000] 19.5
C8 pp H + + CMS [67] [180;600] 19.7
Atb
8 pp H tb ATLAS [68] [200;600] 20.3
C8tb pp H + tb CMS [67] [180;600] 19.7

Table 5. List of the available heavy Higgs searches from LHC Run 1 relevant for the 2HDM. In the
first column, we assign a label and colour to each search, which correspond to the ones in Figures
3 to 9. Details of production and decay modes are given in the second column. The third column
contains the corresponding reference. The mass ranges, for which the corresponding limits on B
are given, and the integrated luminosity the searches are based on, can be found in the fourth and
fifth column. The CMS Run 1 limits of the di-photon channel are included in their Run 2 bounds.
V V refers to either W W or ZZ. C8V V provides signal strength limits. Ahh 8 contains information
about the decays of hh to 4b, 2 2b, 22b and 22W .

brackets. The secondary decay products of one particle are combined in parentheses. In
the case in which two concurring searches are available that are partially based on the
same set of data, we use the limit which is derived from the larger amount of data. For
instance, the latest CMS results of the searches for an H decaying via ZZ into two leptons
and two neutrinos are available only for mH > 600 GeV. Lighter mH scenarios will be
constrained using an older publication based on an integrated luminosity of 2.3 fb1 . Also
for the upper limits on H hh (bb)(bb) by CMS and on H + tb by ATLAS we apply

7
 Label Channel Experiment Mass range [TeV] L [fb1 ]
bb
C13 pp H/A bb CMS [69] [0.55;1.2] 2.69
A13 ATLAS [70] [0.2;2.25] 36.1
gg H/A
C13 CMS [71] [0.09;3.2] 12.9
A13b

ATLAS [70] [0.2;2.25] 36.1
bb H/A
C13b CMS [71] [0.09;3.2] 12.9
A
13 pp H/A ATLAS [72] [0.2;2.7] 36.7

C13 gg H/A CMS [73] [0.5;4] 35.9
AZ
13 gg H/A Z[ (``)] ATLAS [74] [0.25;2.4] 36.1
Z
C13 gg H/A Z CMS [75] [0.35;4] 35.9
A2`2L
13 gg H ZZ[ (``)(``, )] ATLAS [76] [0.2;1.2] 36.1
A2`2L
13V V V H ZZ[ (``)(``, )] ATLAS [76] [0.2;1.2] 36.1
2`2
C13 pp H ZZ[ (``)()] CMS [77] [0.6;2.5] 35.9
2`2
C13g gg H ZZ[ (``)()] CMS [78] [0.2;0.6] 2.3
2`2
C13V V V H ZZ[ (``)()] CMS [78] [0.2;0.6] 2.3
4`
C13V (V V + V H) H ZZ (``)(``) CMS [79] [0.13;2.53] 12.9
2`2q
C13 pp H ZZ[ (``)(qq)] CMS [80] [0.5;2] 12.9
2L2q
A13 gg H ZZ[ (``, )(qq)] ATLAS [81] [0.3;3] 36.1
A2L2q
13V V V H ZZ[ (``, )(qq)] ATLAS [81] [0.3;3] 36.1
2(`)
A13 gg H W W [ (e)()] ATLAS [82] [0.25;3] 13.2
2(`)
A13V V V H W W [ (e)()] ATLAS [82] [0.25;3] 13.2
2(`)
C13 (gg+V V ) H W W (`)(`) CMS [83] [0.2;1] 2.3
A`2q
13 gg H W W [ (`)(qq)] ATLAS [84] [0.3;3] 36.1
A`2q
13V V V H W W [ (`)(qq)] ATLAS [84] [0.3;3] 36.1
A4q
13 pp H V V [ (qq)(qq)] ATLAS [74] [1.2;3] 36.7
A4b
13 ATLAS [85] [0.3;3] 13.3
4b pp H hh (bb)(bb)
C13 CMS [86] [0.26;0.76] 2.3
4b
C13g gg H hh (bb)(bb) CMS [87] [0.76;3] 35.9
22b
A13 pp H hh[ ()(bb)] ATLAS [88] [0.275;0.4] 3.2
22b
C13 pp H hh ()(bb) CMS [89] [0.25;0.9] 35.9
2b2
C13 pp H hh (bb)( ) CMS [90] [0.25;0.9] 35.9
2b2V
C13 pp H hh (bb)(V V ``) CMS [91] [0.26;0.9] 36
22W
A13 gg H hh[ ()(W W )] ATLAS [92] [0.25;0.5] 13.3
AbbZ
13 gg A hZ (bb)Z ATLAS [93] [0.2;2] 36.1
AbbZ
13b bb A hZ (bb)Z ATLAS [93] [0.2;2] 36.1
A13 ATLAS [94] [0.2;2] 14.7
pp H
C13 CMS [95] [0.18;3] 12.9
Atb
13 ATLAS [96] [0.3;1] 13.2
pp H + tb
Atb
13 ATLAS [97] [0.2;0.3][1;2] 13.2

Table 6. List of the available heavy Higgs searches from LHC Run 2 relevant for the 2HDM. For
an explanation, see the description below Table 5. In the last column, we additionally highlight an
underlying integrated luminosity of around 3, 13 or 36 fb1 in red, yellow or green, respectively.

8
different searches depending on the masses. For gg X , CMS combined their 8
and 13 TeV data; the limits are given for the 13 TeV production cross section. A detailed
discussion of how the different searches constrain the 2HDM can be found in Section 5,
where we show the results ordered by the decay products.
Apart from the discussed tree-level Higgs observables the 2HDM scalars can also con-
tribute to the quantum corrections of other observables, the most important ones being
the electroweak precision observables, the b s branching ratio and the mass difference
in the Bs meson system. While the implementation into HEPfit was already explained in
[31], we updated the experimental values [110112]. Also for the treatment of theoretical
constraints we refer to [31], with two exceptions: We do not apply any constraints arising
from the renormalization group evolution and define our model at the electroweak scale.
And for the next-to-leading order unitarity bounds we chose the most conservative ap-
proach that appeared reasonable to us, namely requiring that the real and imaginary parts
of the S-matrix eigenvalues should be between 0.5 and 0.5 and between 0 and 1, respec-
tively. Moreover we impose perturbativity by discarding scenarios for which the one-loop
contribution to these eigenvalues exceeds the tree-level term in magnitude.
As numerical set-up we use the open-source package HEPfit [113], interfaced with the
release candidate of the Bayesian Analysis Toolkit (BAT) [114]. The former calculates all
mentioned 2HDM observables and feeds them into the parallelized BAT, which applies the
Bayesian fit with Markov chain Monte Carlo simulations.

4 h signal strengths

In this section we show the impact of the h signal strengths on the 2HDM parameters. The
fits were done with the most up-to-date experimental inputs; for a comparison with the
status before EPS-HEP 2017, see [115]. The differences in the Z2 symmetry assignment
to the fermions result in a type dependent treatment of their couplings to the light Higgs
boson. The signal strength of the process with a given initial state i producing an h which
decays to the final state f can be written as
rf
fi = ri P , (4.1)
rf 0 BSM (h f 0 )
f0

where rx is the ratio of the 2HDM and the SM partial width of an h decaying into x and
BSM (h x) is the corresponding SM branching ratio. From this equation one can see that
every signal strength depends on the 2HDM h couplings of all decay products.
In Figure 1 we show the individual impact of the signal strengths with a specific final
state on the vs. tan plane as well as their combination in all four types of Z2
symmetry. We have tried to adopt the colouring scheme from Figure 14 of the Run 1
combination [3]. All contours delimit the regions allowed with a probability of 95.4%. The
upper limit on Z
pp is included in the combination, but not shown separately as its effect
is minimal.
In type I all fermions have the same relative coupling to h: rtt = rbb = r = r =
cos( )/ tan + sin( ). This can only deviate significantly from 1 if tan is

9
Figure 1. The impact of the h signal strength measurements is illustrated in the vs. tan
plane in all four 2HDM types. We show the 95.4% posterior probability contours for individual fits
to data from h decays to , bb, , , W W and ZZ in red, cyan, purple, orange, blue and green,
respectively. The resulting 95.4% regions of the combined fits to all signal strengths are the grey
areas.

smaller than 1 and is not close to the alignment limit /2. In these regions, the
di-photon signal strengths are the most constraining ones, see the upper left panel of Figure
1. For large tan values, the ZZ and W W signal strengths become the most important
constraints. In the combined fit to all signal strengths, the largest possible deviation of
from /2 is 0.26 at 95.4% if we marginalize over all other parameters.
The upper right panel of Figure 1 shows fit results with the same inputs for type
II. Here, the relative down-type fermion and lepton couplings to h are different from the
top coupling, and thus the fermionic signal strengths yield more powerful constraints.
But also the signal strengths with a bosonic final state become stronger because of the
modifications of the loop coupling rgg and the fermionic couplings in the denominator of
Eq. (4.1). Especially the W W and ZZ signal strengths constrain to be very close
to /2; the largest deviation from the alignment limit in the one-dimensional fit to all
signal strengths is 0.043 at 95.4%. The so-called wrong-sign solution for the fermionic

10
Figure 2. The 95.4% probability contours for different combinations of signal strengths are shown
in the plane of the relative one-loop couplings of the h to gluons and photons. While the colours
of the shaded contours with solid borders correspond to the ones in Figure 1, the Run 1 and Run
2 combinations are bounded by the dark green and brown dashed contours, respectively.

couplings, which is represented by the lower branches of the individual W W , ZZ, bb

and signal strength fits for tan > 3 and < 1.5, can (cannot) be excluded in the
combined fit to all signal strengths with a probability of 95.4% (99.7%). These scenarios
have also been shown to be incompatible with the assumption that the 2HDM of type II
is stable under a renormalization group evolution up to O(1) TeV [31, 116].
In type X, the h couplings of the down-type quarks agree with the ones of the top
quark, but the leptonic couplings are like in type II. Consequently, the contour of the bb
decays in the lower left panel of Figure 1 has a similar shape as the one of type I, while
the and decays behave more like in type II for large tan . For tan > 2 the latter
two are the dominant signal strengths. For very large tan , the wrong-sign solution of the
fermion couplings is allowed at 95.4%. However, no larger deviations of from /2
than 0.080 are allowed at the 95.4% level if we combine all signal strength information and
marginalize over all other parameters.

11
 Finally, the type Y fit can be found in the lower right panel of Figure 1. Like in type II,
has to be very close to the alignment limit with the bosonic signal strengths being the
strongest constraints. But like in type X, the wrong-sign coupling of the fermions cannot
be completely excluded at 95.4% in the fit combining all signal strengths, although it is
only possible for very large tan . In this types combined fit and marginalizing over the
other parameters, cannot be further away from /2 than 0.049 with a probability of
95.4%.
As compared to the status before EPS-HEP 2017 [115], the W W , and bb signal
strengths have become more constraining; the latter changed drastically, because now we
have information about the inclusive search for h bb, which contains information about
the ggF produced h. In type II, a small spot of the wrong sign branch around tan = 3
and = 1 was allowed at the 95.4% before summer 2017 and has disappeared now.
The two angles and define all tree-level couplings of fermions and bosons to the
light Higgs h, but the loop couplings to gluons and photons are more complicated. In order
to analyse their allowed ranges, we show the rgg vs. r plane in Figure 2. Apart from the
individual fits to the different final states and their combination like in Figure 1, we also
add the contours from a fit to only Run 1 and only Run 2 data, respectively.
In all types, the combined fit to all signal strengths is dominated by the bosonic
decays. While the mainly delimit r , rgg is constrained also by the W W and ZZ
measurements. The maximal deviation of r (rgg ) from its SM value is roughly 30%
(20%). The wrong-sign solution for the fermion couplings can be seen in type II, X and Y:
The regions for rgg > 1 in Figure 2 contain the lower branches of Figure 1. For type II, it
has been shown that the wrong-sign couplings feature increased rgg and reduced r [12].
In the lower right panel of Figure 2, this second solution is visible between 1.1 and 1.2
for rgg as spikes in the W W , ZZ and contours for large r as well as in the combined
signal strength fit in both r directions. Comparing all Run 1 signal strengths with all
Run 2 signal strengths, one can see that generally the Run 2 data is more constraining in
all types. However, the Run 1 signal strengths prefer a smaller rgg and thus determine the
upper limit of the gluon coupling ratio in the combined fit to all signal strengths. This can
especially be seen in type X.

5 Heavy Higgs searches

In the following, we will scrutinize the impact of the searches for heavy Higgs particles
in all four types of a 2HDM with a softly broken Z2 symmetry, ordered by their decay
products. First we will address the fermionic decays to tt, bb and and the loop induced
decays with and Z in the final state. The searches for signals in these channels apply
to both, H and A bosons. After that, we will turn towards the H specific decays into two
massive vector bosons or two h bosons and the A specific channel with an h and a Z in
the final state. Finally, the decays of a charged Higgs to and tb will be discussed. The
narrow width approximation will be applied throughout this section; we will comment in
its validity at the end of the next section.

12
 The grey shaded regions in the plots in this section depict the prediction of the 13
TeV B for the corresponding channel without applying any theoretical or experimental
constraints on the model; in other words they correspond to our priors. The black dashed
lines delimit the available ranges of the B for the corresponding channel when only the
theoretical constraints defined in the Sec. 3 have been used in the fit. The areas within
the various coloured solid lines depict the 95.4% posterior ranges of B after imposing
the experimental constraints from the LHC for a particular measurement. The legend
of each plot refers to the channels described in Tables 5 and 6. The horizontal coloured
lines on the top of the panels mark the mass ranges analyzed at the LHC for each of the
searches denoted in the legend. In the following plots the posterior prediction of B after
considering a particular direct search replicates the prior behaviour unless it deviates from
rest of the posteriors for the same channel. In other words, the direct searches only have a
visible impact on the 2HDM parameters if their contour is lower than the other coloured
contours. Also the lower lines of the searches mostly represent the priors and do not pose
any significant lower limit on B.

5.1 H and A decays to tt

For H and A masses heavier than two top quarks the decay to tt is the dominant in the
2HDM, at least for moderate values of tan . Unfortunately, a possible signal strongly
interferes with the tree-level background process gg tt. The only available experimental
limits of an analysis which takes into account this interference is [117] for the 2HDM of
type II. Its limits, however, constrain only a small region for tan 1 and mH/A 500
GeV, which has been excluded by indirect constraints. Therefore, we do not take into
account this direct measurement.

5.2 H and A decays to bb

The direct search for H bb decay does not put any constraint on B in all four 2HDM
types considered in the analysis. In Figure 3 one can see that theoretical constraints provide
a suppression of B by roughly an order of magnitude compared to the fit without any
constraint in type I and X. Similar to the previous case, pseudoscalar decaying to bb searches
do not provide any stronger constraint on B than the fit with theory constraint alone
in all four types. For this search, theory constraints alone restrict B by at least one
order in magnitude with respect to the fit without any constraints in the parameter space
analyzed for all the types. This suppression is more dominant in type II and Y compared
to the other two cases. In type II and Y, in the regime mA . 600 GeV the A bb search
from Run 1 suppresses B compared to the fit without any constraints but it remains
sub-dominant or at most of similar strength to the fit with theory constraint alone.

5.3 H and A decays to

In the upper panel of Figure 4 we show that the H searches suppress the B limit
by at least one order of magnitude compared to the fit with theory constraints alone in
the regime where the heavy Higgs mass is below 250 GeV. In this regime the strongest
constraint comes from Run 1 data and the suppression of B is more pronounced in the

13
Figure 3. 95% contours of the single searches for H/A bb in the B vs. mH/A planes for the
four 2HDM types (top: H, bottom: A). For details, see text.

14
types I and Y. Theory constraints alone restrict B by roughly an order of magnitude
compared to the fit without any constraints for type I and Y, whereas in type II and X,
theory constraints raise the lower limit of B for mH > 500 GeV. This can be understood
as a sensitivity of the fit to fine-tuned scenarios with extreme tan values, which are
disfavoured in a fit to the theoretical bounds.
From the fit without any constraints in the lower panel of Figure 4 we see that the
predicted ranges of B for the pseudoscalar decaying to are quite narrow for type II
and X compared to the other two 2HDM types. The theory constraints yield a suppression
by at least one order of magnitude for B compared to the fit without any constraints
in the types I and Y whereas for type II and X the theory constraints push up the lower
limit on B by one order of magnitude compared to the fit without any constraints for
mA > 350 GeV. The direct search limits for A suppress B by roughly one to two
orders of magnitude compared to the fit with theory constraints alone for all the types
except for type Y as long as mA . 300 GeV. In type Y, the experimental upper limits on
B are stronger than the theoretical ones for pseudoscalar masses between 200 and 400
GeV, in type II even up to 1 TeV. Scenarios with very light A and B values between
105 . B . 10 pb seem also to be excluded at 95% by the prior. However, since we
have no experimental data on this region, this prior dependence is not an issue here.

5.4 H and A decays to

The theory constraints on H/A suppress B by one to three orders of magnitude
compared to the fit without any constraints in all four 2HDM types, see Figure 5. Direct
searches for a heavy CP-even Higgs decaying to two photons constrain B by roughly
one order of magnitude compared the fit with theory constraints for mH . 250 GeV in
all types. The searches in the di-photon decay channel of a pseudoscalar Higgs yield a
suppression of B by one to three orders of magnitude compared to the fit with theory
constraints for mA . 600 GeV for all four types considered. In the types II and Y, we
observe again that certain intermediate B regions for low mA are disfavoured by the
prior.

5.5 H and A decays to Z

We see from the top panel in Figure 6 that for H Z, theory constraints suppress B
by one to four orders of magnitude compared to the fit without any constraints for all
types. In all four types, direct search for this channel does not provide any constraint on
B except for a very small window below mH ' 250 GeV, but it remains sub-dominant
compared to the fit with theory bounds.
Similar to the heavy CP-even Higgs case, theory constraints yield a suppression of B
by one to two orders of magnitude for the decay A Z compared to the fit without any
constraints in all four types. Direct searches for this channel provide a suppression of B
by an order of magnitude compared to the fit with theory constraints in the mass window
250 . mA . 350 GeV in all four types. Again, some parts of the B region for light A
masses are disfavoured by the prior in the types II and Y.

15
Figure 4. 95% contours of the single searches for H/A in the B vs. mH/A planes for the
four 2HDM types (top: H, bottom: A). For details, see text.

16
Figure 5. 95% contours of the single searches for H/A in the B vs. mH/A planes for the
four 2HDM types (top: H, bottom: A). For details, see text.

17
Figure 6. 95% contours of the single searches for H/A Z in the B vs. mH/A planes for the
four 2HDM types (top: H, bottom: A). For details, see text.

18
5.6 H decays to ZZ or W W
The heavy Higgs decays to massive gauge bosons can be divided into searches for ZZ and
W W , but the coupling of H to two vector bosons V V = ZZ, W W is universal and type
independent. However, the production of the H differs between the types. We show the
H ZZ channels in the upper panel of Figure 7 and the searches for H W W as well
as the combined searches for H V V in its lower panel. The B are constrained by the
theoretical bounds in the decoupling limit, where mH > 600 GeV. The direct LHC searches
for this channel yield a strong suppression of B by one to three orders of magnitude
compared to the fit with theory constraint in the mass regime 150 . mH . 800 GeV
(150 . mH . 750) for the ZZ (W W ) channel. For the ZZ searches the mH . 250 GeV
region is constrained by Run 1 data whereas Run 2 data determine the dominant limits for
the rest of the region. For the W W searches, Run 1 data dictate the limit until mH ' 600
GeV and the high mass range is dominated by Run 2 data. Additionally the C8V V search
for H W W severely constrains B in the 200 . mH . 250 GeV region for type II and
Y.

5.7 H decays to hh
In the upper panel of Figure 8 we show that for the H decaying to two h bosons, theory
constraints already yield a strong suppression of B compared to the fit without any
constraints in all four types. Direct searches suppress B by at most one order of magnitude
compared to the fit with the theory constraints in the mass range 200 . mH . 600 (700)
GeV in type I and X (type II and Y). The main constraints stem from Run 1 data and
the Run 2 searches for hh resonances decaying to two photons and two bottom quarks.
Although different searches at Run 1 and Run 2 continue to constrain the B for this
channel up to mH ' 1200 GeV, they remain sub-dominant to the limit from the fit with
theory constraints.

5.8 A decays to hZ
The searches for A decaying into hZ are shown in the B vs. mA planes in the lower panel
of Figure 8; more precisely, they are projected onto the B of the decay to (bb)Z. As
compared to the fit without any constraints, theory constraints effectively are important
in the decoupling limit in all types as well as for mA . 300 GeV in type I and X. Direct
searches for this channel yield a strong suppression (one to three orders of magnitude) of
B compared to the fit with theory constraints for all four types as long as mA . 800 GeV.
In all types, the strongest bounds come from searches in the A hZ (bb)Z channel; in
type X we additionally observe that the A hZ ( )Z limits yield the most important
constraints for mA between 200 GeV and 300 GeV.

5.9 H + decays
At the present, charged Higgs decaying into does not provide any constraint on B in
the 2HDMs under consideration, see the upper panel of Figure 9. The theory constraints
yield a suppression of B by roughly one order of magnitude as compared to the fit

19
Figure 7. 95% contours of the single searches for H ZZ (top) and the remaining H V V
(bottom) in the B vs. mH planes for the four 2HDM types. For details, see text.

20
Figure 8. 95% contours of the single searches for H hh (top) and A hZ (bottom) in the B
vs. mH/A planes for the four 2HDM types. For details, see text.

21
without any constraint for all types but type X as well as for all types if the charged Higgs
decays into tb (lower panel of Figure 9). In the latter case, the direct searches from Run
1 (Run 2) provide even stronger limits on B between 180 GeV and 600 GeV (1 TeV),
with only small differences between the 2HDM types.

5.10 All heavy Higgs searches

In Figure 10 we show the available parameter space for 2HDM masses and angles from
the fit where the heavy Higgs searches are taken into account. The region inside the
various coloured patches are disfavoured by the corresponding search category denoted in
the legend. The central areas inside the solid grey line mark the 95.4% allowed regions
when all heavy Higgs searches are considered in the fit. In the panels in the first row the
black dashed lines mark the limit from the fit to theory constraints only. The combination
of all H/A/H + searches is represented by the orange/blue/green dashed contours. The
channels described in the previous sub-sections which are not constraining or very weakly
constraining in this mass vs. angles plane are not shown in the figure.
From the first row of Figure 10 we can see that the region around = /2 remains
unconstrained in all four types of 2HDMs when all the heavy Higgs searches are taken
in account. From moderate to high masses, the di-Higgs channels dominate the excluded
regions in all four types whereas H , H and H V V are the most important
constraints below the hh threshold. In type I and X, the final exclusion region is mainly
dominated by the heavy Higgs to two light Higgs channel, while in type II and Y this is
only true if > /2; for < /2 the final constraint on the mass is weak. The main
constraint in this region is due to H searches, which only constrain mH up to 400
GeV for up to 1.2. Although the di-photon searches alone disfavour mH . 600 GeV
for regions near ' /2 this region is allowed when considering all the heavy Higgs
searches in the fit. The H V V decays only are susceptible to the tan . 1 region and
for mH & 1000 GeV in all four types; this again is an effect of fine-tuning. In the displayed
mH vs. tan ranges, all the H searches become ineffective for mH & 1000 GeV in type II
and Y, whereas in type I and X the tan . 1 regions remain inaccessible even if the heavy
Higgs mass is as large as 1500 GeV. This is an effect of the combination of all heavy Higgs
searches, in which scenarios with small tan are sampled less by the fitter. This feature is
not worrisome, because these regions are also suppressed by the flavour observabels as we
will show in the next section.
In the pseudoscalar mass vs. tan planes we see that the di-photon channel constrains
low tan and mA up to 600 GeV for all four types. The A hZ channel can exclude tan
values up to 10 and mA almost as heavy as 1000 GeV in all four types. The exclusion even
applies for the large tan regions in the types II and Y. The next most important channel
for the pseudoscalar searches in type II is the A channel, which efficiently excludes
high tan regions for mA as large as 1000 GeV. In type X this channel is susceptible to
tan . 20 and mA . 400 GeV. The exclusion from this channel is weaker for type I and Y.
The contour for all pseudoscalar searches is mainly dominated by the hZ channel in type I
and Y, and a combination of hZ and in type II and X. Combining all the pseudoscalar
Higgs searches the present data constrain regions where mA . 1000 GeV. For type X, the

22
Figure 9. 95% contours of the single searches for H + (top) and H + tb (bottom) in the
B vs. mH + planes for the four 2HDM types. For details, see text.

23
Figure 10. In the 2HDM masses vs. angles planes we display the regions excluded by all heavy
Higgs searches with a probability of 95.4% by the central area inside the grey solid line. We compare
them with the areas excluded by searches in various final states represented by the coloured patches.
The areas inside the coloured dashed lines correspond to the exclusion at 95.4% when all H searches
(orange), all A searches (dark blue) and all H + searches (dark green) are considered. In the first
row the limits from theory constraints are shown by black dashed lines.

24
mA < 400 GeV region remains available only if tan & 10. In the same mass regions a
very narrow range of intermediate tan remains accessible for type II when all constrains
are taken into account. The upper bound on the pseudoscalar mass is similar to the heavy
Higgs case when all constraints are taken into account.
As described in the previous sub-section, the main channel which constrains the
charged Higgs mass is H + tb. The exclusion region of this channel is shown in red
in the last row of Figure 10. From the figures we see that the present searches for the
charged Higgs mass can only constrain the regions with tan . 1 and mH + . 1000 GeV
in all four types. The inclusion of all the searches in the fit yields stronger constraints in
the mH + vs. tan plane than the fit to all charged Higgs searches only. In type II and Y
this is due to the H hh 4b searches, which are particularly sensitive to large tan
and disfavour certain regions featuring tan & 15. For type I and X it is more difficult
to pinpoint one particular channel for the seeming exclusion of tan values up to 2 for
charged Higgs masses above 1 TeV, but in the next section we see that these bounds can
be relaxed once we take into account also other constraints.

6 Combination of all constraints

After discussing the individual effects of the h signal strengths and the searches for H,
A and H + on the 2HDM, we want to confront these constraints with the other bounds
on the parameters. In Figure 11 we copy the information about all heavy Higgs searches
from Figure 10 in the mass vs. angle planes, and add the bounds from signal strengths and
theoretical constraints to the mH vs. planes and the impact of the flavour observables
to the planes with tan . Finally, the global fit to all constraints is represented by the grey
regions.
The signal strength bounds do not depend on the masses of the heavy Higgs particles;
their limits on differ for each type, see Section 4. The theory conditions force
the 2HDMs into the alignment limit for mH > 600 GeV, decoupling the heavy Higgs
particle from physics around the electroweak scale. The type dependence of this effect is
negligible as it only enters via sub-leading Yukawa terms in the beta function parts of the
NLO unitarity conditions. Besides the obvious consequences for certain constellations of
mH and , the combination of the signal strengths with the theory constraints also
disfavours large values of tan . This is because extreme values for the latter result in a
destabilization of the unitarity conditions [31]. Specific combinations of the 2HDM angles
can still fulfil the theoretical constraints, but these solutions are highly fine-tuned and thus
have a low posterior probability. The flavour constraints have been discussed many times
in the literature; the summary is that in all types the Bs mass difference sets lower limits on
tan (at least for masses within the reach of the LHC), while the branching ratio of b s
processes enforces mH + & 580 GeV at 95% C.L. in type II and Y [118]. In combination
with theory and electroweak precision bounds, these limits on the charged Higgs mass can
be translated to lower limits on the neutral masses. (The individual impact of ST U will be
explained below as it is not visible in these two-dimensional projections of the parameter
space.)

25
 The combination of all constraints is more intricate than the naive superposition of
all individual bounds. First of all, we should mention that it also depends on the prior
we choose for the masses: while the direct experimental observables depend on the masses
of the heavy Higgs bosons, the theoretical bounds and the loop-induced effects are only
sensitive to the mass squares. Since we want to combine both, we need to decide whether
we want to use a flat prior for the masses or the mass squares. A detailed discussion can
be found in Appendix A. The contours we show here are a superposition of a fit with flat
mass priors and a fit with flat mass square priors in order to be as conservative as possible.
In the mH vs. planes of type I and X, the combination of signal strengths and theory
only leave a very small strip around the alignment limit of = /2. Heavy Higgs
searches additionally exclude mH < 360 GeV in type X. The reason for this are not only
the H searches as observed in Figure 10, as they are similar to the bounds in type I,
but mainly it is an interplay of the strong bounds of the A searches for low tan and the
above-mentioned exclusion of large tan due to signal strengths and theory, which together
disallow mA < 380 GeV. This bound translates to a limit on mH using the unitarity and
electroweak precision constraints, since these bounds delimit the mass splittings, see below.
Also in the type II and Y planes we see a lower limit of mH > 550 GeV, which in this case
derives from the lower bound on mH + from the b s measurements and the fact that
the mass difference mH mH + cannot be too large. The absolute maximal deviation of
from /2 is 0.079 in type I and 0.029 in the types II, X and Y. (This corresponds to
1 sin( ) < 3 103 and < 4 104 , respectively.) In the mH vs. tan planes one can
see that the fine-tuning for large tan scenarios disfavours these regions and pushes the
allowed contours towards smaller values of tan . Only in type I and for mH < 300 GeV,
the heavy Higgs searches have a visible impact on this plane, excluding tan . 2.5. More
or less the same holds for mA vs. tan , where in type I tan < 3 is excluded by direct
A searches if mA < 300 GeV. We already mentioned above that in type X, the interplay
between fine-tuning and A searches sets a lower limit of 380 GeV on mA . Having a look at
the mH + vs. tan planes, one can see the lower bounds on the charged Higgs mass in type
II and Y from b s, which we quantify to be 600 GeV in our fit. Also here, we observe
that large tan values are disfavoured and the posterior regions are shifted towards small
tan .
The fit only to electroweak precision data does not exclude any region in the two-
dimensional mass vs. angle projections. What it does constrain, are mass differences
between H, A and H + . That is why in Figure 12 we show the difference between the
pseudoscalar and charged Higgs mass, once depending on the H mass (left column) and
once against the mH mH + difference. In the mA mH + vs. mH planes the dominant
constraints come from the theory bounds, at least in the decoupling limit mH > 600 GeV.
Solely in type I, the ST U pseudo-observables are stronger if mH < 250 GeV and yield a
lower bound on mA mH + . But if we look at the mA mH + vs. mH mH + planes, we
observe that here the oblique parameters are the strongest constraint on mA mH + if the
charged Higgs mass is larger than mH . Combining all constraints and marginalizing over
all other parameters, we obtain the following ranges for the mass differences allowed with
a probability of 95%:

26
Figure 11. In the 2HDM mass vs. angles planes we display the regions allowed by all constraints
with a probability of 95.4% in light grey and compare them with the areas excluded by various sets
of bounds: The 95.4% contours of heavy Higgs searches (dark grey), flavour observables (yellow)
and h signal strengths (pink) as well as the 99.7% limits from theory constraints (purple).

27
 mH mA [GeV] mH mH + [GeV] mA mH + [GeV]
Type I [-202;90] [-165;75] [-104;183]
Type II [-127;69] [-120;62] [-92;110]
Type X [-168;79] [-142;70] [-98;155]
Type Y [-130;70] [-120;60] [-90;110]

We can thus exclude the decays H H + H , H AA, H H + W in all types as well

as H AZ in the types II, X and Y with a probability of 95.4%.
For all heavy Higgs search limits, we implicitly assumed the narrow width approxi-
mation. In a simultaneous fit to all constraints except for these limits we find that with
a probability of 95% the decay widths of H, A and H + never exceed 5.5% of the mass
of the particle in the types II and Y. For masses below 1 TeV the maximal decay widths
are less than 3.5% in these two types. In type I and X and for = H, H + the fits yield
/m < 3.5% (< 5%) if m < 1 TeV (< 1.5 TeV). Only the ratio A /mA can reach 7%
for mA 550 GeV, but in the decoupling limit mA > 600 GeV similar bounds apply as for
H and H + . We would like to stress here that all these 95% limits are maximally allowed
values and that in a typical 2HDM scenario the widths are significantly smaller. There-
fore we conclude that the narrow width approximation is a reasonable choice for 2HDM
scenarios.
Finally, addressing the last variable of our chosen parametrization, we also observe
limits on the soft Z2 breaking parameter m212 . The upper limits strongly depend on the
maximally allowed physical Higgs masses and are around (1.05 TeV)2 in all types. Due
to the lower mass limits on the physical Higgs particles in the types II, X and Y, we also
observe that m212 is limited from below, the respective minimal values being (300 GeV)2 ,
(170 GeV)2 and (260 GeV)2 . Only in type I an unbroken Z2 symmetry is still compatible
with all constraints.

7 Conclusions

In all four 2HDM types with a softly broken Z2 symmetry we have presented global fits to
the most recent data.
Focussing on the latest measurements from LHC, we have showed explicitly how the
individual signal strengths affect the leading order h couplings at tree-level and at one-
loop level. Combining all information about the signal strengths, we find that the quantity
| /2| cannot exceed 0.26, 0.043, 0.080 and 0.049 in the types I, II, X and Y. The
one-loop couplings of the h to gluons and photons cannot differ by more than 20% and
30%, respectively, relative to their SM values.
In order to systematically discuss the searches for H, A and H + , we have categorized
them according to their decay products and have compared the exclusion strength of the
single available ATLAS and CMS analyses on the production cross section times branching
ratio, depending on the masses. We have then combined all decay categories and have

28
Figure 12. In the mA mH + vs. mH (left panels) and mA mH + vs. mH mH + (right panels)
planes we show the allowed regions by various sets of constraints: The heavy Higgs searches, the
oblique parameters and the flavour observables determine the 95.4% allowed contours in dark grey,
light blue and yellow, respectively. For the theoretical constraints, the 99.7% regions are given by
the purple shaded areas. We superimpose the 95.4% probability combination from the global fit to
all observables in light grey.

29
showed their impact on the 2HDM masses and mixing angles. For mH below 1 TeV we
observe strong bounds on extreme values for the angles, that is if is very different
from the alignment limit /2 or if tan is smaller than 1 or larger than 10. The exact
limits depend on the model type and mH . Also the LHC searches for pseudoscalars severely
constrain the 2HDM parameters: For mA < 1 TeV, the lower limits on tan reach values
of around 10 in the types I and X. In the types II and Y, these limits are weaker, but there
are also mass dependent upper limits. The bounds from charged Higgs searches are less
constraining in comparison; nevertheless, they also start to be stronger than the indirect
constraints in the regions with low mH + and low tan .
Finally, we have confronted the LHC h signal strengths and heavy Higgs searches
with all other relevant indirect constraints from theory and experiment. In detail, we
have showed how stability and unitarity constraints and B physics observables set mass
dependent limits on the 2HDM angles and on the differences between the heavy Higgs
masses, while electroweak precision data only affect the latter. We have compared all
different sets of constraints and have showed the results in the mass vs. angle planes as
well as in the mA mH + vs. mH (mH + ) planes for all four types of Z2 symmetric 2HDMs
together with the simultaneous fit to all constraints. In this global fit we find the following
95% probability limits on the 2HDM parameters marginalizing over all other parameters:
| /2| cannot be larger than 0.08 in type I and 0.03 in the other types. In type
II and Y, mH > 710 GeV, mA > 750 GeV, mH + > 750 GeV and m212 > 260 GeV2 , while
we observe lower mass limits of mH > 420 GeV, mA > 500 GeV, mH + > 460 GeV and
m212 > 170 GeV2 for type X. For the latter, it is the first time that a statistically significant
lower limit on the massive parameters has been observed a global fit in the analyzed mass
ranges. Also, if we discard particularly fine-tuned scenarios, only the following ranges for
tan are allowed for masses below 1.6 TeV: [0.93; 8.5] in type I, [0.93; 4.8] in type II, [0.93;
5.5] in type X and [0.91; 4.9] in type Y. However, the upper limits are no strict bounds and
have to be taken with a grain of salt. Moreover, we can put type dependent upper limits
of order of 100 GeV on the differences between mH , mA and mH + , and thus kinematically
exclude all decays of H or A into another heavy Higgs particle except for H AZ in type
I. As a consequence, the decay widths of H and H + cannot exceed 5.5% of their mass in
all types, at least as long as we consider masses below 1.5 TeV. While in the types II and
Y we see a similar limit for the A decay width, it can amount to up to 7% of mA in the
types I and X.

Acknowledgments

We thank Enrico Franco, Ayan Paul, Maurizio Pierini and Luca Silvestrini for useful dis-
cussions. The fits were run on the Roma Tre Cluster. We are grateful for the availability of
these resources and especially want to thank Antonio Budano for the support. This work
was supported by the European Research Council under the European Unions Seventh
Framework Programme (FP/2007-2013) / ERC Grant Agreement n. 279972, by CEFIPRA
(Grant no. 5404-2) and by the Spanish Government and ERDF funds from the European
Commission (Grants No. FPA2014-53631-C2-1-P and SEV-2014-0398).

30
A Prior dependence of the massive parameters

In this appendix we discuss the prior dependence of our analysis when all the constraints
have been taken into consideration. In Figure 13 we compare the allowed parameter space
in the mass versus angles planes and the mass difference versus mass (difference) planes
from the fit with flat mass priors (blue solid and dashed curves) with the fit with flat mass
square priors (red solid and dashed curves). At a first glance one can see that the fit with
flat mass square priors prefers high mass regions and raises the lower limit on the masses
by O(100) GeV compared to the fit with flat mass priors.
It is well known that Bayesian statistics do not provide a unique rule to determine the
prior distribution and in general the posterior distribution is a prior dependent quantity.
A thumb rule would be to choose a flat prior for the parameter on which the observables
depend linearly. For example, if an observable quadratically depends on a particle mass,
one would choose a flat mass square prior. Unfortunately, in the 2HDM the theoretical
and indirect experimental constraints depend on the mass squares, whereas the direct
experimental observables dependent on the masses. Not having sufficiently constraining
data makes assigning the mass priors in the fit with all constraints is a delicate task. This
would not be problematic if the observables were measured with a high precision; in fact, we
can see that in the types II and Y the difference between the two priors are considerably
smaller as there are strong lower mass limits. In order to be as conservative and prior
independent as possible, we decided to combine the 95.4% regions for both priors: The
light grey contours for the fits with all constraints in Figures 11,12 and 13 are obtained
by superimposing a fit with flat mass priors and a fit with flat mass square priors. The
corresponding numerical results mentioned in Section 6 are based on the more conservative
fit; for instance, the limits for masses and mass differences were extracted from the fit with
flat mass priors, while the upper limits on the decay widths were larger in the fits using
flat mass square priors.

References

[1] ATLAS collaboration, G. Aad et al., Observation of a new particle in the search for the
Standard Model Higgs boson with the ATLAS detector at the LHC, Phys.Lett. B716 (2012)
129, [1207.7214].
[2] CMS collaboration, S. Chatrchyan et al., Observation of a new boson at a mass of 125 GeV
with the CMS experiment at the LHC, Phys. Lett. B716 (2012) 3061, [1207.7235].
[3] ATLAS, CMS collaboration, G. Aad et al., Measurements of the Higgs boson production
and decay rates and constraints on its couplings from a combined ATLAS and CMS analysis

of the LHC pp collision data at s = 7 and 8 TeV, 1606.02266.
[4] T. Lee, A Theory of Spontaneous T Violation, Phys.Rev. D8 (1973) 12261239.
[5] J. F. Gunion and H. E. Haber, The CP conserving two Higgs doublet model: The Approach
to the decoupling limit, Phys. Rev. D67 (2003) 075019, [hep-ph/0207010].
[6] G. Branco, P. Ferreira, L. Lavoura, M. Rebelo, M. Sher et al., Theory and phenomenology
of two-Higgs-doublet models, Phys.Rept. 516 (2012) 1102, [1106.0034].

31
Figure 13. The mass prior dependence of the fit with all constraints is shown in the mH/A/H +
vs. tan planes, in the mH vs. planes (corresponding to Figure 11) and in the mA mH +
vs. mH and mA mH + vs. mH mH + planes (corresponding to Figure 12), from top to bottom.

32
 [7] C.-Y. Chen and S. Dawson, Exploring Two Higgs Doublet Models Through Higgs
Production, Phys.Rev. D87 (2013) 055016, [1301.0309].
[8] C.-W. Chiang and K. Yagyu, Implications of Higgs boson search data on the two-Higgs
doublet models with a softly broken Z2 symmetry, JHEP 1307 (2013) 160, [1303.0168].
[9] B. Grinstein and P. Uttayarat, Carving Out Parameter Space in Type-II Two Higgs
Doublets Model, JHEP 1306 (2013) 094, [1304.0028].
[10] A. Barroso, P. Ferreira, R. Santos, M. Sher and J. P. Silva, 2HDM at the LHC - the story
so far, 1304.5225.
[11] B. Coleppa, F. Kling and S. Su, Constraining Type II 2HDM in Light of LHC Higgs
Searches, JHEP 1401 (2014) 161, [1305.0002].
[12] O. Eberhardt, U. Nierste and M. Wiebusch, Status of the two-Higgs-doublet model of type
II, JHEP 1307 (2013) 118, [1305.1649].
[13] G. Belanger, B. Dumont, U. Ellwanger, J. Gunion and S. Kraml, Global fit to Higgs signal
strengths and couplings and implications for extended Higgs sectors, Phys.Rev. D88 (2013)
075008, [1306.2941].
[14] S. Chang, S. K. Kang, J.-P. Lee, K. Y. Lee, S. C. Park et al., Two Higgs doublet models for

the LHC Higgs boson data at s = 7 and 8 TeV, JHEP 1409 (2014) 101, [1310.3374].
[15] K. Cheung, J. S. Lee and P.-Y. Tseng, Higgcision in the Two-Higgs Doublet Models, JHEP
1401 (2014) 085, [1310.3937].
[16] A. Celis, V. Ilisie and A. Pich, Towards a general analysis of LHC data within
two-Higgs-doublet models, JHEP 1312 (2013) 095, [1310.7941].
[17] L. Wang and X.-F. Han, Status of the aligned two-Higgs-doublet model confronted with the
Higgs data, JHEP 1404 (2014) 128, [1312.4759].
[18] J. Baglio, O. Eberhardt, U. Nierste and M. Wiebusch, Benchmarks for Higgs Pair
Production and Heavy Higgs boson Searches in the Two-Higgs-Doublet Model of Type II,
Phys. Rev. D90 (2014) 015008, [1403.1264].
[19] S. Inoue, M. J. Ramsey-Musolf and Y. Zhang, CP-violating phenomenology of flavor
conserving two Higgs doublet models, Phys.Rev. D89 (2014) 115023, [1403.4257].
[20] B. Dumont, J. F. Gunion, Y. Jiang and S. Kraml, Constraints on and future prospects for
Two-Higgs-Doublet Models in light of the LHC Higgs signal, Phys. Rev. D90 (2014) 035021,
[1405.3584].
[21] S. Kanemura, K. Tsumura, K. Yagyu and H. Yokoya, Fingerprinting nonminimal Higgs
sectors, Phys.Rev. D90 (2014) 075001, [1406.3294].
[22] P. Ferreira, R. Guedes, J. F. Gunion, H. E. Haber, M. O. P. Sampaio et al., The
CP-conserving 2HDM after the 8 TeV run, 1407.4396.
[23] A. Broggio, E. J. Chun, M. Passera, K. M. Patel and S. K. Vempati, Limiting
two-Higgs-doublet models, JHEP 1411 (2014) 058, [1409.3199].
[24] B. Dumont, J. F. Gunion, Y. Jiang and S. Kraml, Addendum to Constraints on and future
prospects for Two-Higgs-Doublet Models in light of the LHC Higgs signal, 1409.4088.
[25] J. Bernon, J. F. Gunion, Y. Jiang and S. Kraml, Light Higgs bosons in Two-Higgs-Doublet
Models, 1412.3385.

33
[26] C.-Y. Chen, S. Dawson and Y. Zhang, Complementarity of LHC and EDMs for Exploring
Higgs CP Violation, 1503.01114.
[27] D. Chowdhury and O. Eberhardt, Global fits of the two-loop renormalized
Two-Higgs-Doublet model with soft Z2 breaking, JHEP 11 (2015) 052, [1503.08216].
[28] N. Craig, F. DEramo, P. Draper, S. Thomas and H. Zhang, The Hunt for the Rest of the
Higgs Bosons, JHEP 06 (2015) 137, [1504.04630].
[29] J. Bernon, J. F. Gunion, H. E. Haber, Y. Jiang and S. Kraml, Scrutinizing the alignment
limit in two-Higgs-doublet models: mh =125 GeV, Phys. Rev. D92 (2015) 075004,
[1507.00933].
[30] J. Bernon, J. F. Gunion, H. E. Haber, Y. Jiang and S. Kraml, Scrutinizing the alignment
limit in two-Higgs-doublet models. II. mH =125 GeV, Phys. Rev. D93 (2016) 035027,
[1511.03682].
[31] V. Cacchio, D. Chowdhury, O. Eberhardt and C. W. Murphy, Next-to-leading order
unitarity fits in Two-Higgs-Doublet models with soft Z2 breaking, JHEP 11 (2016) 026,
[1609.01290].
[32] H. Belusca-Mato, A. Falkowski, D. Fontes, J. C. Romao and J. P. Silva, Higgs EFT for
2HDM and beyond, Eur. Phys. J. C77 (2017) 176, [1611.01112].
[33] J. de Blas, M. Ciuchini, E. Franco, S. Mishima, M. Pierini, L. Reina et al., Electroweak
precision observables and Higgs-boson signal strengths in the Standard Model and beyond:
present and future, 1608.01509.
[34] O. Deschamps, S. Descotes-Genon, S. Monteil, V. Niess, S. TJampens and V. Tisserand,
The Two Higgs Doublet of Type II facing flavour physics data, Phys. Rev. D82 (2010)
073012, [0907.5135].
[35] J. F. Gunion, H. E. Haber, G. L. Kane and S. Dawson, The Higgs Hunters Guide, Front.
Phys. 80 (2000) 1448.
[36] ATLAS collaboration, T. A. collaboration, Measurements of Higgs boson properties in the
diphoton decay channel with 36.1 fb1 pp collision data at the center-of-mass energy of 13
TeV with the ATLAS detector, .
[37] ATLAS collaboration, T. A. collaboration, Combined measurements of Higgs boson

production and decay in the H ZZ 4` and H channels using s = 13 TeV pp
collision data collected with the ATLAS experiment, .
[38] ATLAS collaboration, T. A. collaboration, Measurement of the Higgs boson coupling

properties in the H ZZ 4` decay channel at s = 13 TeV with the ATLAS detector, .
[39] ATLAS Collaboration collaboration, Measurements of the Higgs boson production cross
section via Vector Boson Fusion and associated W H production in the W W `` decay

mode with the ATLAS detector at s = 13 TeV, Tech. Rep. ATLAS-CONF-2016-112,
CERN, Geneva, Nov, 2016.
[40] ATLAS collaboration, M. Aaboud et al., Evidence for the H bb decay with the ATLAS
detector, 1708.03299.
[41] ATLAS Collaboration collaboration, Search for the Standard Model Higgs boson

produced in association with top quarks and decaying into bb in pp collisions at s = 13
TeV with the ATLAS detector, Tech. Rep. ATLAS-CONF-2016-080, CERN, Geneva, Aug,
2016.

34
[42] ATLAS collaboration, M. Aaboud et al., Search for the dimuon decay of the Higgs boson in

pp collisions at s = 13 TeV with the ATLAS detector, 1705.04582.
[43] ATLAS collaboration, P. Thompson, Search for non-standard, rare or invisible decays of
the Higgs boson with the ATLAS detector, tech. rep., CERN, Geneva, 2017.
[44] CMS collaboration, C. Collaboration, Measurements of properties of the Higgs boson in the
diphoton decay channel with the full 2016 data set, .
[45] CMS collaboration, A. M. Sirunyan et al., Measurements of properties of the Higgs boson
decaying into the four-lepton final state in pp collisions at sqrt(s) = 13 TeV, 1706.09936.
[46] CMS Collaboration collaboration, Higgs to WW measurements with 15.2 fb1 of 13
TeV proton-proton collisions, Tech. Rep. CMS-PAS-HIG-16-021, CERN, Geneva, 2017.
[47] CMS collaboration, A. M. Sirunyan et al., Inclusive search for a highly boosted Higgs boson
decaying to a bottom quark-antiquark pair, 1709.05543.
[48] CMS collaboration, A. M. Sirunyan et al., Observation of the Higgs boson decay to a pair of
tau leptons, 1708.00373.
[49] CMS collaboration, C. Collaboration, Search for the associated production of a Higgs boson

with a top quark pair in final states with a lepton at s = 13 TeV, .
[50] CMS collaboration, V. Khachatryan et al., Search for Neutral MSSM Higgs Bosons
Decaying into A Pair of Bottom Quarks, JHEP 11 (2015) 071, [1506.08329].
[51] ATLAS collaboration, G. Aad et al., Search for neutral Higgs bosons of the minimal

supersymmetric standard model in pp collisions at s = 8 TeV with the ATLAS detector,
JHEP 11 (2014) 056, [1409.6064].
[52] CMS collaboration, Search for additional neutral Higgs bosons decaying to a pair of tau

leptons in pp collisions at s = 7 and 8 TeV. CMS-PAS-HIG-14-029, 2015.
[53] ATLAS collaboration, G. Aad et al., Search for Scalar Diphoton Resonances in the Mass

Range 65 600 GeV with the ATLAS Detector in pp Collision Data at s = 8 T eV , Phys.
Rev. Lett. 113 (2014) 171801, [1407.6583].
[54] ATLAS collaboration, G. Aad et al., Search for new resonances in W and Z final states

in pp collisions at s = 8 TeV with the ATLAS detector, Phys. Lett. B738 (2014) 428447,
[1407.8150].
[55] CMS collaboration, Search for scalar resonances in the 2001200 GeV mass range decaying

into a Z and a photon in pp collisions at s = 8 TeV. CMS-PAS-HIG-16-014, 2016.
[56] ATLAS collaboration, G. Aad et al., Search for an additional, heavy Higgs boson in the

H ZZ decay channel at s = 8 TeV in pp collision data with the ATLAS detector,
Eur. Phys. J. C76 (2016) 45, [1507.05930].
[57] ATLAS collaboration, G. Aad et al., Search for a high-mass Higgs boson decaying to a W

boson pair in pp collisions at s = 8 TeV with the ATLAS detector, JHEP 01 (2016) 032,
[1509.00389].
[58] CMS collaboration, V. Khachatryan et al., Search for a Higgs Boson in the Mass Range
from 145 to 1000 GeV Decaying to a Pair of W or Z Bosons, JHEP 10 (2015) 144,
[1504.00936].
[59] CMS collaboration, V. Khachatryan et al., Search for resonant pair production of Higgs

35
 bosons decaying to two bottom quarkantiquark pairs in protonproton collisions at 8 TeV,
Phys. Lett. B749 (2015) 560582, [1503.04114].
[60] CMS collaboration, V. Khachatryan et al., Search for two Higgs bosons in final states
containing two photons and two bottom quarks, Submitted to: Phys. Rev. D (2016) ,
[1603.06896].
[61] ATLAS collaboration, G. Aad et al., Searches for Higgs boson pair production in the
hh bb , W W , bb, bbbb channels with the ATLAS detector, Phys. Rev. D92 (2015)
092004, [1509.04670].
[62] CMS collaboration, A. M. Sirunyan et al., Searchp for Higgs boson pair production in the
bb final state in proton-proton collisions at (s) = 8 TeV, Phys. Rev. D96 (2017)
072004, [1707.00350].
[63] ATLAS collaboration, G. Aad et al., Search for a CP-odd Higgs boson decaying to Zh in pp

collisions at s = 8 TeV with the ATLAS detector, Phys. Lett. B744 (2015) 163183,
[1502.04478].
[64] CMS collaboration, V. Khachatryan et al., Search for a pseudoscalar boson decaying into a
Z boson and the 125 GeV Higgs boson in `+ `bb final states, Phys. Lett. B748 (2015)
221243, [1504.04710].
[65] CMS collaboration, V. Khachatryan et al., Searches for a heavy scalar boson H decaying to
a pair of 125 GeV Higgs bosons hh or for a heavy pseudoscalar boson A decaying to Zh, in
the final states with h , Phys. Lett. B755 (2016) 217244, [1510.01181].
[66] ATLAS collaboration, G. Aad et al., Search for charged Higgs bosons decaying via

H in fully hadronic final states using pp collision data at s = 8 TeV with the
ATLAS detector, JHEP 03 (2015) 088, [1412.6663].
[67] CMS collaboration, V. Khachatryan et al., Search for a charged Higgs boson in pp

collisions at s = 8 TeV, JHEP 11 (2015) 018, [1508.07774].
[68] ATLAS collaboration, G. Aad et al., Search for charged Higgs bosons in the H tb decay

channel in pp collisions at s = 8 TeV using the ATLAS detector, JHEP 03 (2016) 127,
[1512.03704].
[69] CMS collaboration, C. Collaboration, Search for a narrow heavy decaying to bottom quark
pairs in the 13 TeV data sample, .
[70] ATLAS collaboration, M. Aaboud et al., Search for additional heavy neutral Higgs and

gauge bosons in the ditau final state produced in 36 fb1 of pp collisions at s = 13 TeV
with the ATLAS detector, 1709.07242.
[71] CMS collaboration, C. Collaboration, Search for a neutral MSSM Higgs boson decaying

into with 12.9 fb1 of data at s = 13 TeV, .
[72] ATLAS collaboration, M. Aaboud et al., Search for new phenomena in high-mass diphoton

final states using 37 fb1 of protonproton collisions collected at s = 13 TeV with the
ATLAS detector, 1707.04147.
[73] CMS collaboration, C. Collaboration, Search for resonant production of high mass photon

pairs using 12.9 fb1 of proton-proton collisions at s = 13 TeV and combined
interpretation of searches at 8 and 13 TeV, .
[74] ATLAS collaboration, L. Carminati, Search for high mass bosonic resonances with the
ATLAS detector. 2017.

36
[75] CMS collaboration, C. Collaboration, Search for high-mass Z resonances in proton-proton

collisions at s = 13 TeV, .
[76] ATLAS collaboration, T. A. collaboration, Search for heavy ZZ resonances in the

`+ ` `+ ` and `+ ` final states using protonproton collisions at s = 13 TeV with the
ATLAS detector, .
[77] CMS collaboration, C. Collaboration, Search for diboson resonances in the 2l2 final state, .
[78] CMS collaboration, C. Collaboration, Search for a heavy scalar boson decaying into a pair
of Z bosons in the 2`2 final state, .
[79] CMS collaboration, C. Collaboration, Measurements of properties of the Higgs boson and
search for an additional resonance in the four-lepton final state at sqrt(s) = 13 TeV, .
[80] CMS collaboration, C. Collaboration, Search for new diboson resonances in the dilepton +

jets final state at s = 13 TeV with 2016 data, .
[81] ATLAS collaboration, M. Aaboud et al., Searches for heavy ZZ and ZW resonances in the

``qq and qq final states in pp collisions at s = 13 TeV with the ATLAS detector,
1708.09638.
[82] ATLAS collaboration, T. A. collaboration, Search for a high-mass Higgs boson decaying to

a pair of W bosons in pp collisions at s=13 TeV with the ATLAS detector, .
[83] CMS collaboration, C. Collaboration, Search for high mass Higgs to WW with fully
leptonic decays using 2015 data, .
[84] ATLAS collaboration, M. Aaboud et al., Search for W W/W Z resonance production in

`qq final states in pp collisions at s = 13 TeV with the ATLAS detector, 1710.07235.
[85] ATLAS collaboration, T. A. collaboration, Search for pair production of Higgs bosons in the

bbbb final state using protonproton collisions at s = 13 TeV with the ATLAS detector, .
[86] CMS collaboration, C. Collaboration, Search for resonant pair production of Higgs bosons
decaying to two bottom quark-antiquark pairs in proton-proton collisions at 13 TeV, .
[87] CMS collaboration, A. M. Sirunyan et al., Search for a massive resonance decaying to a

pair of Higgs bosons in the four b quark final state in proton-proton collisions at s = 13
TeV, 1710.04960.
[88] T. A. collaboration, Search for Higgs boson pair production in the bb final state using pp

collision data at s = 13 TeV with the ATLAS detector, .
[89] CMS collaboration, C. Collaboration, Search for Higgs boson pair production in the final
state containing two photons and two bottom quarks in proton-proton collisions at

s = 13 TeV, .
[90] CMS collaboration, A. M. Sirunyan et al., Search for Higgs boson pair production in events
with two bottom quarks and two tau leptons in proton-proton collisions at sqrt(s) = 13 TeV,
1707.02909.
[91] CMS collaboration, C. Collaboration, Search for resonant and non-resonant Higgs boson

pair production in the bbll final state at s = 13 TeV, .
[92] ATLAS collaboration, T. A. collaboration, Search for Higgs boson pair production in the

final state of W W ( ljj) using 13.3 fb1 of pp collision data recorded at s = 13
TeV with the ATLAS detector, .

37
 [93] ATLAS collaboration, T. A. collaboration, Search for heavy resonances decaying to a W or
Z boson and a Higgs boson in final states with leptons and b-jets in 36.1 fb1 of pp collision

data at s = 13 TeV with the ATLAS detector, .
[94] ATLAS collaboration, T. A. collaboration, Search for charged Higgs bosons in the +jets

final state using 14.7 fb1 of pp collision data recorded at s = 13 TeV with the ATLAS
experiment, .
[95] CMS collaboration, C. Collaboration, Search for charged Higgs bosons with the H

decay channel in the fully hadronic final state at s = 13 TeV, .
[96] ATLAS collaboration, T. A. collaboration, Search for charged Higgs bosons in the H tb

decay channel in pp collisions at s = 13 TeV using the ATLAS detector, .
[97] ATLAS collaboration, T. A. collaboration, Search for new phenomena in tt final states with

additional heavy-flavour jets in pp collisions at s = 13 TeV with the ATLAS detector, .
[98] LHC Higgs Cross Section Working Group collaboration.
https://twiki.cern.ch/twiki/bin/view/LHCPhysics/LHCHXSWG, 2015.
[99] M. Spira, HIGLU: A program for the calculation of the total Higgs production cross-section
at hadron colliders via gluon fusion including QCD corrections, hep-ph/9510347.
[100] R. V. Harlander, S. Liebler and H. Mantler, SusHi: A program for the calculation of Higgs
production in gluon fusion and bottom-quark annihilation in the Standard Model and the
MSSM, Comput. Phys. Commun. 184 (2013) 16051617, [1212.3249].
[101] J. Alwall, R. Frederix, S. Frixione, V. Hirschi, F. Maltoni et al., The automated
computation of tree-level and next-to-leading order differential cross sections, and their
matching to parton shower simulations, JHEP 1407 (2014) 079, [1405.0301].
[102] C. Degrande, R. Frederix, V. Hirschi, M. Ubiali, M. Wiesemann and M. Zaro, Accurate
predictions for charged Higgs production: closing the mH mt window, 1607.05291.
[103] C. Degrande, M. Ubiali, M. Wiesemann and M. Zaro, Heavy charged Higgs boson
production at the LHC, JHEP 10 (2015) 145, [1507.02549].
[104] M. Flechl, R. Klees, M. Kramer, M. Spira and M. Ubiali, Improved cross-section predictions
for heavy charged Higgs boson production at the LHC, Phys. Rev. D91 (2015) 075015,
[1409.5615].
[105] LHC Higgs Cross Section Working Group collaboration, D. de Florian et al.,
Handbook of LHC Higgs Cross Sections: 4. Deciphering the Nature of the Higgs Sector,
1610.07922.
[106] S. Dittmaier, M. Kramer, M. Spira and M. Walser, Charged-Higgs-boson production at the
LHC: NLO supersymmetric QCD corrections, Phys. Rev. D83 (2011) 055005, [0906.2648].
[107] E. L. Berger, T. Han, J. Jiang and T. Plehn, Associated production of a top quark and a
charged Higgs boson, Phys. Rev. D71 (2005) 115012, [hep-ph/0312286].
[108] A. Djouadi, J. Kalinowski and M. Spira, HDECAY: A Program for Higgs boson decays in
the standard model and its supersymmetric extension, Comput. Phys. Commun. 108 (1998)
5674, [hep-ph/9704448].
[109] Particle Data Group collaboration, K. A. Olive et al., Review of Particle Physics, Chin.
Phys. C38 (2014) 090001.

38
[110] The global electroweak fit: present status, constraints on new physics, and prospects at
future colliders. Talk given at EPS-HEP 2017.
[111] HFLAV collaboration. http://www.slac.stanford.edu/xorg/hflav/.
[112] Y. Amhis et al., Averages of b-hadron, c-hadron, and -lepton properties as of summer
2016, 1612.07233.
[113] HEPfit collaboration. HEPfit: a Code for the Combination of Indirect and Direct
Constraints on High Energy Physics Models, in preparation.
http://hepfit.roma1.infn.it.
[114] A. Caldwell, D. Kollar and K. Kroninger, BAT: The Bayesian Analysis Toolkit, Comput.
Phys. Commun. 180 (2009) 21972209, [0808.2552].
[115] Two-Higgs-doublet model fits with HEPfit. Talk given at EPS-HEP 2017.
[116] P. Basler, P. M. Ferreira, M. Muhlleitner and R. Santos, High scale impact in alignment and
decoupling in two-Higgs doublet models, 1710.10410.
[117] ATLAS collaboration, M. Aaboud et al., Search for heavy Higgs bosons A/H decaying to a

top quark pair in pp collisions at s = 8 TeV with the ATLAS detector, 1707.06025.
[118] M. Misiak and M. Steinhauser, Weak radiative decays of the B meson and bounds on MH
in the Two-Higgs-Doublet Model, Eur. Phys. J. C77 (2017) 201, [1702.04571].

39