String Theory

Niklas Beisert
ITP,

PSI Summer School

More than Higgs – Effective Theories for Particle Physics
Lyceum Alpinum, Zuoz
21 August 2014
 Introduction
You’ve heard a lot about particle physics in the past few days:
• particles of the standard model;
• particles that have long been predicted,
and that have recently been observed;
• particles that have long been predicted,
and that have recently not been observed;
• further conjectured particles,
that have not been observed yet,
and probably never will be observed,
because they probably do not exist.

This talk:

String Theory
Zuoz 2014, Niklas Beisert 2
 Effective Theory
Why talk about String Theory? Here?!
• string theory is a model of particle physics . . .
• . . . at low energies (compared to Planck scale).
• it predicts new particles and features, many of them, . . .
• . . . at least at excessively high energies.
• so we can only dream about them. After this talk, please!
Many results in string theory use effective theory concepts.
• In fact, if you believe string theory describes nature,
all that you’ve heard about here
is nothing but an effective theory of string theory.

. . . but still a long way to go.

Zuoz 2014, Niklas Beisert 3

 Overview
Therefore let me introduce string theory
and some of its basic concepts:
• What is String Theory?
• Problem of Quantum Gravity.
• String Theory Basics.
• String Theory and Gravity.
• D-branes.

Material from a one-semester lecture course . . .

Zuoz 2014, Niklas Beisert 4

 I. What is String Theory?

Zuoz 2014, Niklas Beisert 5

 Einstein with a violin?

Zuoz 2014, Niklas Beisert 6

 Nothing to do with violins!

Zuoz 2014, Niklas Beisert 7

 What is String Theory?
String Theory: [Wikipedia]
• In physics, string theory is a theoretical framework in which the
point-like particles of particle physics are replaced by
one-dimensional objects called strings.
• String theory aims to explain all types of observed elementary
particles using quantum states of these strings.
• In addition to the particles postulated by the standard model of
particle physics, string theory naturally incorporates gravity, and so
is a candidate for a theory of everything, a self-contained
mathematical model that describes all fundamental forces and
forms of matter.
• Besides this hypothesized role in particle physics, string theory is
now widely used as a theoretical tool in physics, and has shed light
on many aspects of quantum field theory and quantum gravity.

Zuoz 2014, Niklas Beisert 8

 What else is String Theory?
• incredibly exciting (for some),
• rich in hidden, exceptional and beautiful structures,
• popular:
I Brian Greene,
The Elegant Universe
(book, TV)
I Woody Allen,
Whatever Works
(movie)
I ...

Zuoz 2014, Niklas Beisert 9

 Cartoon: Witten’s Dog
from Futurama, Episode 11: Mars University

Zuoz 2014, Niklas Beisert 10

 String Theory vs. Loop Quantum Gravity
Except from “The Big Bang Theory”
Episode 2.2 – “The Codpiece Topology”

Zuoz 2014, Niklas Beisert 11

 So What is String Theory?

Cartoon on String Theory

xkcd webcomic: 171

Zuoz 2014, Niklas Beisert 12

 Strings
String theory is a physical model whose fundamental objects are
strings. Strings have the following properties:
• spatially 1D extended

• with tension

• without inner structure

• open or closed

Zuoz 2014, Niklas Beisert 13

 Spacetime Diagrams
When painting the path of a string one gets 2D surfaces in spacetime:

compare: string open rotating closed

particle motion string string string

Zuoz 2014, Niklas Beisert 14

 String Interactions
fundamental interactions between strings:

open closed mixed compare:

strings strings strings point-particle
Benefits of string theory (precitive power):
There are merely two constants of nature, namely
• string tension (within a string),
• string coupling (coupling between strings).

Zuoz 2014, Niklas Beisert 15

 Cartoon: Witten’s Dog
from Futurama, Episode 11: Mars University

Zuoz 2014, Niklas Beisert 16

 Why String Theorie?

Cartoon on String Theory

xkcd webcomic: 171

Zuoz 2014, Niklas Beisert 17

 String Spectrum
How does this relate to particle physics?
Limit of large string tension:

many types
of particles!

linear spectrum m2n = m20 + n · ∆m2 .

spectrum of hadronic excitations similar!
String theory was a promising candidate.
Later better explanation: quark model, QCD, SM.
Instead string theory incorporates a consistent quantum gravity!
Zuoz 2014, Niklas Beisert 18
 II. Problem of Quantum Gravity

Zuoz 2014, Niklas Beisert 19

 Einstein Gravity
Einstein’s general relativity described by Hilbert action
Z
1 p
SGR = d4 x − det g R[g].
16πG
• fundamental field is metric gµν measuring distances.
• R = g µν Rµν [g] = g µν Rρ µρν [g] is curvature scalar.
• coupling to matter via diffeo.-covariant derivative: ∂µ → Dµ [g].
Gravity has a non-linear action translating to infinitely many vertices

√
SGR = + G +G + G3/2 + G2 + ....

Not a fundamental problem; moreover G is small (gravity is weak).

Zuoz 2014, Niklas Beisert 20

 Non-Renormalisability
Additional terms in gravity action conceivable
Z
S ∼ d4 x R + ∗R2 + ∗R3 + ∗D2 R2 + . . . .
 

Translates to additional interaction vertices

√
S∼ G + (G + c4 ) + (G3/2 + ∗) + ....

Additional constants required to absorb divergences in loops, e.g.

(G + c4 ) + G2 + G3 + ....

Problem: Infinitely many constants needed to absorb all divergences!

Zuoz 2014, Niklas Beisert 21
 String Theory and Quantum Gravity
String theory promises to solve the problem of quantum gravity:
• string excitation spectrum contains massless spin-2 particles;
• they behave as Einstein gravitons to first approximation;
• string theory is finite, no divergences (of the above kind);
• string theory has just two fundamental coupling constants.
Is all well now!? Beware of the vacuum. . .

Other Features.
• String spectrum contains many other particles (gauge bosons, . . . );
but still unclear how to extract the standard model.
• GUT-like groups SU(5), SO(10), . . . , E8 appear.
• String theory is supersymmetric; need to break at low energies.
• Exciting geometrical / mathematical concepts arise.

Zuoz 2014, Niklas Beisert 22

 III. String Theory Basics

Zuoz 2014, Niklas Beisert 23

 Particle Worldline Action
Start with an analogous system: Relativistic Particle
Embedding of worldline

X µ (τ ) : R → RD−1,1 .

Proper time action

dX µ
Z Z Z q
S ∼ m ds := m dτ |Ẋ| := m dτ −Ẋ µ Ẋµ , Ẋ µ = .
dτ
Equations of motion
m µ
Ṗ µ = 0 with P µ = Ẋ .
|Ẋ|
Note: momentum P µ and equation of motion covariant under
reparametrisation τ → τ 0 (τ ).
Zuoz 2014, Niklas Beisert 24
 New Particle Worldline Action
Can use an alternative action with additional field e(τ ) : R → R
Z
S ∼ dτ − 21 e−1 Ẋ µ Ẋµ + 12 em2 .

Equations of motions
Ẋ µ
Ṗ µ = 0 with P µ = and |Ẋ| = em.
e
First equation linear in Ẋ µ and equivalent to above (with second).
Second equation is algebraic in e!
• e does not carry additional (quantum) degrees of freedom;
• recover the original action upon substituting e = |Ẋ|/m.
  Z
|Ẋ| 2
Z
m µ
S ∼ dτ − Ẋ Ẋµ + m = dτ m|Ẋ|.
2|Ẋ| 2m

Zuoz 2014, Niklas Beisert 25

 Nambu–Goto Action
Repeat steps for a string embedding
X µ (τ, σ) : R1,1 → RD−1,1
Proper area action (Nambu–Goto)
Z p
S ∼ T d2 ξ − det γ with γαβ := ∂α X µ ∂β Xµ .

Remarks:
• γαβ [X]
√ is induced metric on worldsheet;
• d2 σ − det γ is worldsheet “area” element;
• T is string tension.
Equation of motion extremises worldsheet area
p
∂α − det γ γ αβ ∂β X µ = 0.


Equation is invariant under worldsheet reparametrisation ξ α 7→ ξ 0α (ξ).

Zuoz 2014, Niklas Beisert 26
 Polyakov Action
Nambu–Goto equations of motion non-linear; difficult!!
Introduce worldsheet metric gαβ
Z
T p
S∼ d2 ξ − det g g αβ ∂α X µ ∂β Xµ .
2
Again, two equations of motion
p
∂α − det g g αβ ∂β X µ = 0, γαβ [X] − 12 gαβ g γδ γγδ [X] = 0.


Remarks:
• first equation linear in X;
• second equation algebraic in g;
worldsheet metric proportional to induced metric: gαβ ∼ γαβ [X];
• equations equivalent to original ones.
Zuoz 2014, Niklas Beisert 27
 Symmetries and Conformal Gauge
Action and equations of motion invariant under:
• target space Poincaré transformations X µ (ξ) → M µ ν X ν (ξ) + B µ .
• worldsheet diffeomorphisms ξ α 7→ ξ 0α (ξ);
• worldsheet metric rescaling gαβ (ξ) → α(ξ)gαβ (ξ);
2 + 1 fields carry redundant information (gauge).

Conformal Gauge: Exploit redundancy to set gαβ = ηαβ (3 eq.).

Reduced action and equations of motion:
Z
T
S∼ d2 ξ η αβ ∂α X µ ∂β Xµ , η αβ ∂α ∂β X µ = 0.
2
Almost simple linear system, X µ are massless waves on worldsheet.
Do not forget other equation! Virasoro constraint (non-linear):
Tαβ [X] := γαβ [X] − 12 ηαβ η γδ γγδ [X] = 0.
Zuoz 2014, Niklas Beisert 28
 String Modes
General solution for X by massless plane waves
Z
dk  µ
µ
a (k) eikσ+i|k|τ + aµ (k)∗ e−ikσ−i|k|τ .

X (ξ) =
2|k|
Need to impose boundary conditions. E.g. closed string:
X µ (τ, σ + 2π) = X µ (τ, σ).
Periodicity forces k to be integer: string Fourier modes
∞  µ µ
pµ τ αL,n αR,n
X 
µ µ −in(τ +σ) −in(τ −σ)
X (ξ) = x + + Im √ e +√ e .
2πT n=1 πT n πT n
Solutions parametrised by:
• centre of gravity position and momentum xµ , pµ ;
µ
• string mode excitation amplitudes αL/R,n , n = 1, 2, . . ..
Zuoz 2014, Niklas Beisert 29
 Quantisation
We want to quantise the system
• a free particle (xµ , pµ )
µ µ,∗
• a set of harmonic oscillators (αR/L,n , αR/L,n )

Canonical quantisation:
 µ ν  µ ν,∗
x , p = iη µν , = mδm,n η µν .

αR/L,m , αR/L,n

States are tensor products of

• momentum eigenstate |P i;
• excitation eigenstate |N i = (α† )N |0i; for each string mode.

String ground state |0; P i = |P i ⊗ |0i ⊗ |0i ⊗ . . .

• Almost point particle α ≈ 0: X µ ≈ xµ +√pµ τ /2πT .
• HO ground state localisation: ∆X ∼ 1/ T , typical string length.

Zuoz 2014, Niklas Beisert 30

 String Spectrum
Furthermore need to consider constraints Tαβ = 0:
• removes negative norm states;
• balances left/right total excitation numbers
X∞
µ,† ν
NL − aL = NR − aR , NL/R = ηµν αL/R,n αL/R,n ;
n=1
aL/R are normal ordering constants of the quantum theory;
• determines mass of string states
P 2 = −M 2 , M 2 = 8πT (N − a).
Furthermore, spin is bounded by S ≤ 2N . S
massive
string
Leading Regge trajectory state

graviton
S − 2a 2-form
M 2 = 4πT (S − 2a) = . 2a
α0 /2 α0
tachyon dilaton M2
Zuoz 2014, Niklas Beisert 31
 Anomalies and Consistency
To obtain a consistent spectrum without negative norm states:
• need D = 26 dimensions of spacetime;
• need aL = aR = 1;
Underlying reason is quantum anomaly of symmetries:
• cannot have worldsheet reparametrisation and Lorentz symmetry
• unless D = 26, aL = aR = 1.

Interesting:
• particle at level N = 1 with spin S = 2 is massless: graviton?!
Strange:
• particle at level N = 0 with S = 0 is tachyonic;
• need 22 extra dimensions (visible at least at string scale).
Why D = 26? 26 = 2 + 24; “ ∞ 1 1
P
n=1 2 n = 2 ζ(−1) = −1/24”.

Zuoz 2014, Niklas Beisert 32

 Superstrings
Can we do something about the problems?
• compactify additional dimensions (many options . . . );
• find true ground state of quantum system (in practice?!);
• . . . but will never have half-integer spin / fermionic states.

Need additional fermionic fields on worldsheet. Long story:

• various bosonic and fermionic particles;
• tachyon is absent (gladly);
• critical dimension is reduced to D = 10 (somewhat better);
• supersymmetry is inevitable (at least at string scale).

Stick to bosonic strings here:

• conceptually simpler system;
• qualitatively the same physics;
• can largely ignore the tachyonic mode.
Zuoz 2014, Niklas Beisert 33
 IV. Strings and Gravity

Zuoz 2014, Niklas Beisert 34

 Virasoro–Shapiro Amplitude
Want to compare massless spin-2 state with graviton.
Can compute scattering of string states:
• create external states on string worldsheet;
• integrate over insertions.
Scattering of 4 tachyons (simpler than gravitons)
Z
A4 ∼ gs δ (P ) d2 z |z|p2 ·p4 /2πT |1 − z|p3 ·p4 /2πT
2 D

Γ(−1 − s/8πT ) Γ(−1 − t/8πT ) Γ(−1 − u/8πT )

∼ gs2 δ D (P ) .
Γ(+2 + s/8πT ) Γ(+2 + t/8πT ) Γ(+2 + u/8πT )
• very soft in the UV;
• manifestly crossing symmetric;
• poles of Γ correspond to virtual particles exchanged in s, t and u
channels: string modes!
Zuoz 2014, Niklas Beisert 35
 Graviton Scattering
Corresponding scattering amplitude for spin-2 excitations:
• conceptually similar;
• correct polarisation dependence;
• agrees with leading Einstein gravity at leading large T ;
• Planck length equals string length;
• correction terms at higher orders: α0 ∼ 1/T corrections;
• higher orders in gs : genus corrections

∞
gs2 1 + gs4 0 + gs6 + ...
0

integration over moduli of higher-genus surfaces;

• corrections equivalent to gravity correction terms D2k Rn ;
effective low-energy theory summarising string modes;
Zuoz 2014, Niklas Beisert 36
 Strings and Gravity
String theory contains gravity!
• gravitons behave like in special relativity at low energy;
• there are stringy corrections to gravity at higher orders;

So far:
• started with strings in flat Minkowski space;
• quantisation led to the emergence of gravitons;
• gravitons are the quanta of general relativity;
• general relativity is a theory of dynamical geometry.

What if we started from a curved background spacetime?

Zuoz 2014, Niklas Beisert 37

 Strings in Curved Backgrounds
Put strings into a curved background. Introduce metric field

ηµν → Gµν (x).

Simple replacement for Nambu–Goto action (induced metric γ)

Z p
S ∼ T d2 ξ − det γ with γαβ := Gµν (X) ∂α X µ ∂β X ν .

Polyakov action becomes a non-linear sigma model

Z
T p
S∼ d2 ξ − det g g αβ Gµν (X) ∂α X µ ∂β X ν .
2
Equations of motion now non-linear; cannot be solved in general.

Zuoz 2014, Niklas Beisert 38

 Gravitons and Geometric Deformations
Can show equivalence of stringy gravitons and geometric deformation.
Compare for graviton with momentum p and polarisation µν :
• massless spin-2 excitation at level 1:
µ †
|; pi = µν (αL,1 ν
) (αR,1 )† |0; pi;

• translation to string fields: (αnµ )† → ∂ n X µ ; |0; pi → eip·X ;

• graviton excited by worldsheet operator
Z p
V [, p] = d2 ξ − det g µν g αβ ∂α X µ ∂β X ν eip·X ;

• graviton is plane wave deformation

Gµν (x) = ηµν + µν eip·x + . . . .

Variation of Polyakov action is excitation operator δS = V [p; ].
Zuoz 2014, Niklas Beisert 39
 Divergences of Worldsheet Theory
Curved background changes quantum theory. UV divergences?

...

Metric field G(x) serves as collection of coupling constants for QFT

Z
T p
S∼ d2 ξ − det g g αβ Gµν (X) ∂α X µ ∂β X ν .
2
Divergences at loop level to be absorbed into G (and similar fields).
Running coupling constants
µ∂ 1
Gµν = βµν = Rµν [G].
∂µ 2πT

Zuoz 2014, Niklas Beisert 40

 Anomalies
Beta-function introduces a new scale on worldsheet.
Scale would violate reparametrisation invariance!

βµν = 0 ⇐⇒ Rµν [G] = 0.

Einstein equation! Strings like to propagate in general relativity.

In fact, for general dimension D:

1
βµν = 13 (D − 26) Gµν + Rµν [G] − 21 Gµν R[G] .


2πT
In other words D − 26 serves as a cosmological constant.
For D 6= 26 space curved at the order of the Planck scale.

Zuoz 2014, Niklas Beisert 41

 Low-Energy Effective Action
Gravitons are low-energy excitations (massless).
Can summarise their dynamics by effective action
√
Z  
D 2 1
S ∼ d x − det G − (D − 26) + R + ... .
3 2πT

Remarks:
• Hilbert action with cosmological term;
• corrections from higher orders in 1/T and gs ;
• resulting equations of motion are equivalent to βµν = 0:
string excitations will not cause an anomaly.

Zuoz 2014, Niklas Beisert 42

 Summary
• One of the massless excitations is the graviton;
• string theory clearly contains quantum gravity; finite!
• general relativity plus stringy corrections;
• string scale is the Planck scale;
• may start with flat or curved background.

Background dependence in string theory?

• Quantum theory depends on background;
• similar example: QED; clearly depends on background;
e.g. vacuum vs. constant magnetic field; motion of electrons;
• also ordinary quantum gravity depends on background;
• other backgrounds are highly excited, coherent quantum states;
not accessible in perturbation theory;
• asymptotical behaviour of background may make a difference.
Zuoz 2014, Niklas Beisert 43
 V. D-Branes

Zuoz 2014, Niklas Beisert 44

 Open Strings
For closed strings we simply assume periodicity X µ (σ + 2π) = X µ (σ).
For open strings, we vary the action
Z
S ∼ d2 ξ η αβ ∂α X µ ∂β Xµ

and pay attention to boundary terms (integration by parts) at σ = 0, π

Z
σ=π
δS ∼ dτ δX µ Xµ0 σ=0 + . . . .


Remarks: 0 π
0
• For general variations one finds Neumann condition Xµ = 0.
• Virasoro constraints Tαβ = 0 furthermore imply Ẋ 2 = 0.
• Ends of string must move at the speed of light.
Zuoz 2014, Niklas Beisert 45
 Open String Spectrum
General solution with open boundary conditions X 0 = 0 at σ = 0, π:
• similar to closed strings
• boundary conditions couple left- and right-moving modes
0 π
µ µ 2π
αL,n = αR,n . l
0 π

Mass formula for open strings:

M 2 = 2πT (N − a).

Features:
• vacuum |0; pio is tachyonic;
ignore; can be avoided in superstrings;
• massless vector state at level 1: (α1µ )† |0; pio : photon!

Zuoz 2014, Niklas Beisert 46

 D-Branes
Alternative option: constrain variation at boundary, δX µ = 0:
• Dirichlet boundary condition: fix ends X µ = const.;
• mix boundary conditions: p + 1 Neumann, D − p − 1 Dirichlet;
• free motion in p + 1 directions; fixed in orthogonal directions;
• string ends confined to (p + 1)-dimensional surface: Dp-brane.
Why? e
r an
• why should there be such objects? p-b
• where should they be located? 0 π D
• breaks Poincaré symmetry of background!
)
( p, 1
Why not?!
• add more features to background; D−p−1
• go further: allow strings to end on curved hyperplane.
• string dualities; two different models may yield same physics;
here: T-duality introduces/changes D-branes.
Zuoz 2014, Niklas Beisert 47
 Spectrum with D-Branes
What changes about the spectrum?
• no transversal momentum; motion confined to D-brane;
µ
• identification αLµ = ±αR depends on µ;
• same mass formula.
level-1 massless states (α1µ )† |0; pi split up:
• longitudinal µ: photon along D-brane;
• transversal µ: D − p − 1 scalar particles on D-brane;

What do massless open string modes mean?

Zuoz 2014, Niklas Beisert 48

 Brane Dynamics
Recall massless closed string modes:
• spin-2 particles are gravitons;
• have the same effect as deforming geometry.
Now consider longitudinal massless open string modes:
• spin-1 particles are photons;
• have the same effect as coupling of gauge field to string ends
Z
dτ Ẋ µ Aµ (X);

• gauge field naturally confined to D-brane.

How about transversal massless open string modes?
• have the same effect as deforming the D-brane itself;
• D-branes become dynamical upon string quantisation!

Zuoz 2014, Niklas Beisert 49

 Effective Theories
Can do same steps as for closed strings.

Avoiding UV divergences and anomalies:

• for gravitons: Einstein equations;
• for photons: Maxwell equations;
• for deformations: D-brane equations of motion.

Corresponding low-energy effective actions:

• gravitons: Hilbert action;
• photons: action of electromagnetism;
• deformations: Dirac action for brane motion (induced volume);
• combinations with gravity and higher orders: Dirac–Born–Infeld
Z p
S ∼ dp+1 x − det(Gab + 2πκ2 Fab ) .

Zuoz 2014, Niklas Beisert 50

 Multiple Branes and Gauge Groups
Can now combine branes to construct interesting physics.

Simplest case: N coincident branes

• strings will start on brane j = 1, . . . , N ;
• strings will end on brane k = 1, . . . , N ;
• N 2 massless photon-like modes.
Yang–Mills theory with U(N ) gauge group.

Splitting a stack of branes: N → K + (N − K)

• 2N K strings stretch between stacks;
minimum length leads to mass shift. K N −K
corresponding photons become massive spin-1 particles.
• K 2 plus (N − K)2 strings end on same stack;
corresponding photons remain massless;
Spontaneous breaking of gauge symmetry to U(K) × U(N − K).
Zuoz 2014, Niklas Beisert 51
 String Phenomenology
Finally, combine various features of backgrounds:
• non-compact dimensions and compact manifolds;
• different types of branes;
• stacks of branes, intersections of branes, parallel branes;
in order to obtain the desired
• spectrum of light particles (w.r.t. Planck scale);
• set of global symmetries and supersymmetries;
• set of gauge symmetries;
• symmetry breaking patterns;

Pay attention to:

• stability;
• anomalies;
• number of continuous moduli for the configuration.
Zuoz 2014, Niklas Beisert 52
 VI. Conclusions

Zuoz 2014, Niklas Beisert 53

 Conclusions
Introduced and sketched some basic concepts of string theory:
• Formulation(s) of classical string theory.
• Quantisation and the string spectrum.
• Scattering in string theory.
• How gravity emerges from string theory.
• Low-energy effective theory.
• Open strings and emergent extended objects.
• Construction of suitable particle physics.

Many further results and insights from string theory:

• AdS/CFT correspondence
• mathematics and geometry
• ...

Zuoz 2014, Niklas Beisert 54