Spontaneous Parametric Down Conversion of Photons Through

β-Barium Borate

A Senior Project

By

Luke Horowitz

Advisor, Dr. Glen D. Gillen

Department of Physics, California Polytechnic University SLO

May 24, 2016

1
 Approval Page

Title: Parametric Down Conversion of Photons Through β - Barium Borate

Author: Luke Horowitz

Date Submitted: May 24, 2016

Senior Project Advisor: Dr. Glen D. Gillen

Signature

Date

2
Contents

1 Introduction 6

2 Theory 6
2.0.1 Linear and Nonlinear Optics Primer . . . . . . . . . . . . . . . . . . . . . . . . 6
2.0.2 Non-Linear Optics and SPDC . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.0.3 Phase Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 Experiment 12
3.1 List of Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2.1 Optics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2.2 BBO Crystal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.3 Electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.3 Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

4 Conclusions 17

5 Future Work 18

List of Tables

1 Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Type I vs Type II Phase Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

List of Figures

1 a) General DFG, b) SPDC with crystal geometry, c) Energy level diagram for SPDC, d)
Momentum conservation for SPDC . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3
2 Phase matching angle θ for SPDC where the ordinary polarization is normal to the page,
and the extraordinary polarization is in the plane of the page as illustrated. . . . . . . . 11
3 Optical Path Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4 Mount for BBO crystal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5 Block diagram of electronics showing the how coincidence counts are obtained from the
APD’s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

4
Acronym Meaning
BBO β-Barium Borate Crystal (β-BaB2 O4 )
SFG Sum Frequency Generation
DFG Difference Frequency Generation
OR Optical Rectification
c.c. Complex Conjugates
SHG Second Harmonic Generation
SPDC Spontaneous Parametric Down Conversion
MI Michelson Interferometer
APD Avalanche Photodiode
SCA Single Channel Analyzer
HeNe Helium-Neon Laser

Table 1: Acronyms

5
1 Introduction

In most undergraduate physics programs, there are usually few opportunities for students to do meaning-
ful laboratory work, especially not in unconventional subjects like quantum mechanics or applied optics.
For students who are interested in these subjects, a “hands-on” experiment can be an invaluable learning
tool to help them understand difficult concepts. Galvez, et al., [1] present an apparatus for detecting
photons that undergo quantum entanglement that can be constructed within a reasonable budget of an
undergraduate physics lab. This apparatus can be used to perform a number of different experiments and
would be an excellent addition to an advanced physics laboratory curriculum. Building this apparatus
is relevant because it can be used to teach both advanced theoretical physics like non-linear quantum
optics as well as advanced experimental techniques in optics and electronics. I was motivated to do this
project because of what I learned in Dr. Glen D. Gillen’s Advanced Optics course at Cal Poly SLO. I
wanted to get more experience with experimental optics and was interested in exploring the non-linear
relationships that connect optics and quantum mechanics.

2 Theory
2.0.1 Linear and Nonlinear Optics Primer

Laser light is an oscillating and traveling electromagnetic wave with specific spatial and temporal param-
eters k and ω, respectively. The temporal parameter is the angular frequency, ω, and the spatial parameter
is the wavenumber, k, and they are related via the equation ω = vk, where v is the speed of light in the
medium. The relevant information about the wave is contained in the electric part of the electromagnetic
wave, which, in this case, is mathematically convenient to express as a complex plane wave. This has
the form [2]
Ẽ (t) = Ẽ0 e−iωt + c.c., (1)

where Ẽ0 = E0 ei(k·r+δ) is the complex wave amplitude, δ is the phase shift, and E0 is the actual electric
N V
field amplitude, usually given in units of C or m. Because waves obey the principle of superposition,

6
two incident beams of different frequencies would combine to give [2]

Ẽnet = Ẽ1 e−iω1 t + Ẽ2 e−iω2 t + c.c., (2)

where Ẽ1 = E1 ei(k1 ·r+δ1 ) and Ẽ2 = E2 ei(k2 ·r+δ2 ) . These equations describe fields of just the incident
light. For electric fields traveling through a medium, it is necessary to introduce the polarization of the
medium, which to the first order is given by the relationship [2]

P̃ (t) = χ(1) Ẽ (t) , (3)

where χ(1) is known as the linear susceptibility. More generally, the polarization is actually a nonlinear
process described by a power series expansion of the field strength. This is written mathematically as [2]

P̃ (t) = χ(1) Ẽ (t) + χ(2) Ẽ2 (t) + χ(3) Ẽ3 (t) + · · · (4)

≡ P̃(1) (t) + P̃(2) (t) + P̃(3) (t) + · · · . (5)

For the case where there are two frequencies incident on a medium characterized by a second-order
nonlinear susceptibility, the electric field is given by Eq. 2, only the second-order polarization is relevant,
and there are actually five components of P̃ (2) . This is due to the second-order non-linear polarization
which is dependent upon the square of Ẽnet . Each component of P̃ (2) refers to a different frequency
which can be created in the medium. These components and their acronyms are given by [2]

P (2ω1 ) = χ(2) E12 (SHG) (6)

P (2ω2 ) = χ(2) E22 (SHG) (7)

P (ω1 + ω2 ) = 2χ(2) E1 E2 (SFG) (8)

P (ω1 − ω2 ) = 2χ(2) E1 E2∗ (DFG) (9)

P (0) = 2χ(2) (E1 E1∗ + E2 E2∗ ) (OR) (10)

where E ∗ is the complex conjugate of E. Equations 6 and 7 say that it is possible to create an electro-
magnetic wave with twice the frequency of waves one and two respectively. This process is known as
second harmonic generation, or SHG. Similarly, Eq. 8 says that it is possible to create an electromagnetic
wave with a frequency equal to the sum of the input frequencies and Eq. 9 says that it is possible to create

7
an electromagnetic wave with a frequency equal to the difference in input frequencies. These processes
are known as sum frequency generation, SFG, and difference frequency generation, DFG, respectively.
Finally, Eq. 10 says that it is possible to create a static polarization in the material with zero frequency.
This is known as optical rectification, or OR. The total polarization is given by the sum of each of the
components above.

2.0.2 Non-Linear Optics and SPDC

More relevant to this experiment, the process of spontaneous parametric down-conversion, SPDC, is
similar to the process of DFG, but in reverse. This is shown schematically in Figure 1. One beam with
ω1
frequency ω1 is incident on the crystal, and two beams with ω2 = ω3 = 2 are emitted. The photons
obey both

ω1 = ω2 + ω3 & k1 = k2 + k 3 ; (11)

these are simply statements of conservation of energy and momentum respectively, as Ephoton = ~ω and
pphoton = ~k.

Figure 1: a) General DFG, b) SPDC with crystal geometry, c) Energy level diagram for SPDC, d) Mo-
mentum conservation for SPDC

8
See Boyd’s Nonlinear Optics text, [2], for a complete derivation of the following discussion. Assuming
a lossless medium, the net electric field inside the crystal for two electromagnetic waves is given by

Ẽnet = E1 e−iω1 t + E2 e−iω2 t + c.c., (12)

where Ej = Aj (z) eikj z and Aj is a slowly varying electric field amplitude. The wavenumber and
refractive index for each frequency are given respectively by
n j ωj h i1
2
kj = & nj = (1) (ωj ) , (13)
c
where (1) (ωj ) is the dielectric permittivity tensor. The total intensity of the two beams is I = I1 + I2 ,
where
Pj nj c 2
Ij = = A . (14)
πw0 2 2π j
Pj is the power due to each beam, and w0 is the focal spot size. To simplify the math, Boyd introduces
new normalized field amplitudes u1 and u2 such that

u21 (ξ) + u22 (ξ) = 1, (15)

where
 12
n21 n2 c3

z 1
ξ= and l= . (16)
l 2πI 8πω1 deff
ξ is called the normalized distance parameter, l is the characteristic distance over which the fields ex-
χ(2)
change energy, and deff = 2 . Using the driven non-linear wave equation, it can be shown with some
lengthy algebra that the normalized field amplitudes are given by

u1 (ξ) = sechξ (17)

u2 (ξ) = tanhξ. (18)

When the beam is at optimal focusing, the normalized distance parameter is

!1
1024π 5 d2eff LP 2
ξ= , (19)
n1 n2 cλ31
and the conversion efficiency for this process is given by
u22 Ll

 
2 L
η= 2 = tanh . (20)
u1 (0) l
For typical values, this efficiency turns out to be around 2%.

9
2.0.3 Phase Matching

Not only is the efficiency of this process somewhat low, but it requires a very precise matching of the
wavevectors for the input and output beams. For SPDC, the wavevector mismatch is defined as

∆k ≡ k1 − 2k2 , (21)

where k1 is the the wavevector of the input beam and k2 is the wavevector of the output beam. The
intensity of the output beam depends on ∆k as

sin2 ∆kL

 
2 2 ∆kL
I2 = I2 (max) = I2 (max) sinc . (22)
∆kL 2 2


2

In order to maximize the intensity of the down-converted beam, it is necessary to minimize ∆k. It is
possible do to this a few different ways, but the simplest is called angle tuning. This method involves
precise angular orientation of the crystal with respect to the propagation direction of the incident light.
Reference [2] describes the case of a uniaxial crystal, which pertains to this experiment. Uniaxial crystals
are described by a particular direction known as the optic axis (or c axis). Light polarized perpendicular
to the plane containing the propagation vector k and the optic axis is called the ordinary polarization.
Such light experiences the ordinary refractive index no . Light polarized in the plane containing k and
the optic axis is called the extraordinary polarization and experiences a refractive index ne (θ) which
depends on the angle θ between the optic axis and k according to the formula

1 sin2 θ cos2 θ
= + . (23)
ne (θ)2 n̄2e n2o

The parameter n̄e is the principle value of the extraordinary refractive index. Note that ne = n̄e when
θ = 90◦ and ne = no when θ = 0◦ . The extraordinary refractive index can be tuned by changing θ to
ensure that ∆k = 0. The angle θ is depicted in Figure 2. Depending on the polarization of the output
beams, there are two types of phase matching. Type I phase matching is the case in which the two lower-
frequency beams (in this case ω2 = ω3 ) have the same polarization, and type II is the case where the
polarizations are orthogonal. Differences between the two types are shown in Table 2. In the quantum
photon model, individual incident photons are being split into two photons within the crystal, each with
half the energy and thus half the frequency. These photons experience quantum entanglement, meaning

10
Figure 2: Phase matching angle θ for SPDC where the ordinary polarization is normal to the page, and
the extraordinary polarization is in the plane of the page as illustrated.

Positive uniaxial (ne > no ) Negative uniaxial (ne > no )

Type I no3 ω3 = ne1 ω1 + ne2 ω2 ne3 ω3 = no1 ω1 + no2 ω2
Type II no3 ω3 = no1 ω1 + ne2 ω2 ne3 ω3 = ne1 ω1 + no2 ω2

Table 2: Type I vs Type II Phase Matching

11
the state of each photon is dependent of the other photon’s state, and their intrinsic quantized spins are
always opposite. For this experiment, type I phase matching occurs in a negative uniaxial crystal, so the
phase matching condition is
ne (2ω, θ) = no (ω) . (24)

Using this condition, Eq. 23, and some trigonometry, the angle for ideal phase matching is found to be
1 1
no (ω)2
− n (2ω)2
2 o
sin θ = 1 1 . (25)
n̄e (2ω)2
− n (2ω)2
o

As the pairs of photons exit the crystal, they may do so with some relative angle between their paths.
In general, this angle is given by the equation [1]

n1 k1 = 2n2 k2 cosθc . (26)

k1
Because k2 = 2 , this simplifies to
n1
cosθc = , (27)
n2
which means that when perfect phase matching is achieved, n1 = n2 and θc = 0◦ .

3 Experiment
3.1 List of Equipment

Item Use
405 nm Diode Laser Laser Source
Thorlabs Diode Laser Controller (LDC202) Laser control
632 nm HeNe Laser Alignment
Ortec Fast Coincidence Module (414A) Detecting pairs of down-converted photons
Ortec Dual Counter/Timer (994) Counting Coincident pairs
Ortec Single Channel Analyzer (406A) x 2 Analog to Digital signal conversion
Michelson Interferometer Overlay an interference pattern on the output beam
Newport Universal Motion Controller (ESP300) changing arm length in Michelson
Kikusui 20 MHz Storage Oscilloscope (COS5020-ST) Signal Analysis
Perkin-Elmer Avalanche Photodiodes (SPCM-AQRH-13) x 2 Low-intensity photon detection

Table 3: Equipment

12
3.2 Experimental Setup

Figure 3: Optical Path Block Diagram

3.2.1 Optics

The optical setup for this experiment begins with two lasers: a 632 nm HeNe laser for alignment and a
405 nm diode laser which acts as ω1 . To ensure that the beams travel exactly the same path, a dual iris
system is used. Each beam travels through the same two pinholes early on in the beam path, ensuring that

13
they are collinear. Thus the system can be aligned with the “easy-to-see” HeNe laser and still work for
its less visible counterpart. The beam then goes through two polarizers to give the beam a known polar-
ization and intensity control. After the polarizers, the beam is sent through the BBO crystal, which is in
a mount with two rotational degrees of freedom. Section 3.2.2 describes the crystal and its mount. After
the crystal, the relevant beam has a wavelength of 810 nm. For this reason, the mirrors after the BBO
crystal have a dielectric coating for near infrared light, while the mirrors before the crystal are coated for
light with a wavelength closer to the ultraviolet. The beam is then sent into a Michelson interferometer
in order to create an arm-length dependent interference pattern in the output signal. The arm length in
the interferometer is motion controlled, so very precise distance changes can be measured. The output
is then sent through either a converging lens to expand the beam for alignment of the interferometer, or
to a pair of Avalanche Photodiodes for detection. This is achieved through the use of a flip-down mirror.
There is still a high intensity of 405 nm light in the beam path, so narrow bandpass filters are placed in
front of the APD’s. These filters only let light with a wavelength of 810 ± 5 nm pass through, and thus
only the down-converted photons will be counted.

3.2.2 BBO Crystal

The BBO crystal is mounted such that the beam is at normal incidence, and the mount setup has two
rotational degrees of freedom: the crystal can be rotated along the axis formed by the normal of the
crystal face, as well as along the axis formed by the post supporting it. This is to match the polarization
of the beam with the crystal axis and to make slight adjustments to the angle of incidence if necessary.
The crystal is manufactured by PHOTOP, and consists of two BBO crystals that measure 7 x 7 x 0.1 mm,
are cut at 29 degrees with respect to the optic axis of the crystal, P-coated for optimal transmission at 810
and 405 nm and mounted together with 90 degrees rotated between them. This allows for the beam to
come in at normal incidence as mentioned before, which greatly simplifies the optical setup required. The
down converted photons are thus collinear with the pump beam, θc = 0◦ . Finally, the crystal is a negative
uniaxial crystal designed for Type 1 phase matching, which means that the ordinary and extraordinary
indices of refraction and the various frequencies are related with ne > no via

ne3 ω3 = no1 ω1 + no2 ω2 . (28)

14
Figure 4: Mount for BBO crystal

15
3.2.3 Electronics

Figure 5: Block diagram of electronics showing the how coincidence counts are obtained from the
APD’s.

The first and most important components of the electronics for this experiment are the avalanche
photodiodes. APD’s are extremely sensitive light detectors, capable of detecting single photons. It is
very important to keep light levels very low when the APD’s are powered up with their detectors ex-
posed, because too much light can render the APD broken and worthless. The APD’s are thus always
protected by a wooden box that is bolted to the table to cover them. An APD works by outputting an
analog electrical pulse for every photon detected. The APD’s used for this experiment output TTL pulses
35 ns wide that have a peak of at least 2.5-V in a 50-Ω load. There are 50 ns of dead time between
pulses. The APD’s require a 5-V supply with a current of around 0.5 A (1.9 A max). The pulses from
the APD’s are then sent into single channel analyzers to convert the pulses into a digital signal composed
of step-functions. These cleaner signals are easier to count and are sent into a coincidence module to

16
register only coincident events. The output of the coincidence module is then sent to a counter to keep
track of coincidence counts and count rates. If all goes right, the counter will show the number of pairs of
down-converted photons. In addition to the previously mentioned electronics, an oscilloscope and func-
tion generator were used for testing, and a motion controller was used to change the arm length of the MI.

3.3 Procedures

In order to take meaningful quantitative data, the optical system must be perfectly aligned for phase
matching to occur. Once that happens, a number of different experiments can be performed with the
down-converted photons. For an example, the arm-length of the Michelson Interferometer can be varied
in order to create a pattern of constructive and destructive interference at the detector. This pattern
can be seen by observing the number of coincidence counts at different arm lengths. See [1] for other
possible experiments. Unfortunately there was not enough time to properly build, align, test, and get
experimental data. A lot of time was spent designing and fabricating a box to fit in the tight space around
the APD’s, and the optical setup was so sensitive that it had to be re-aligned nearly every time it was
used. Because of the shortage of time and difficulty in alignment, phase matching was never achieved
and SPDC was never actually detected. Given more time however, phase matching could certainly be
attained and down-converted photons could be detected by the APD’s.

4 Conclusions

In conclusion, the experiment was set up but not yet perfectly aligned. Once it is aligned and phase
matching occurs, a number of experiments can be done with relative ease. Despite not finishing the
project and getting any data, I have learned a lot about theoretical and experimental non-linear quantum
optics, electronics, and signal processing. Additionally, this project has given me invaluable experience
in the flow of and dedication required to do research in experimental physics. The process is filled with
surprises, successes, and failures; a researcher never knows exactly what is going to happen next.

17
5 Future Work

After I graduate, I hope to continue working on refining this project as a Masters student in Electrical
Engineering. There are a number of additional experiments that can be done with the entangled photons
exiting the crystal, and for this reason we are working to make this experiment a part of the advanced
Quantum Laboratory curriculum. This would be done by automating the data collection process using
Labview, in order to more easily take data and make this an approachable experiment for most under-
graduate physics majors.

18
References

[1] E. J. Galvez et al, Am. J. Phys 73, Feb (2004).

[2] R. W. Boyd, Nonlinear Optics, Second Edition, (Academic Press, San Diego, CA, 2003.)

[3] F. L. Pedrotti, L. S. Pedrotti, and L. M. Pedrotti, Introduction to Optics, Third Edition, (Person
Prentice Hall, Upper Saddler River, NJ, 2007.)

19