Physics of Microwave Ovens
Physics of Microwave Ovens
Physics of Microwave Ovens
www.iop.org/journals/physed
Abstract
This is the first of two articles about the physics of microwave ovens. This
article deals with the generation of microwaves in the oven and includes the
operation of the magnetrons, waveguides and standing waves in resonant
cavities. It then considers the absorption of microwaves by foods, discussing
the dielectric relaxation of water, penetration depths of electromagnetic
waves in matter and, in considering the possible chemical changes during
the microwave heating, multi-photon ionization or dissociation.
The wavelengths λx , λy and λz are determined by the linear dimensions Lx , Ly and Lz of the
chamber:
λx λy λz
Lx = l Ly = m Lz = n (3)
2 2 2
where l, m, n are natural numbers. The solutions for given (l, m, n) are denoted as modes of the
resonator. Obviously equation (2) allows more than one possibility to satisfy the same given
value of λ. In commercial microwave ovens, the dimensions Lx , Ly and Lz scatter appreciably;
e.g., a survey of about 20 different ovens gave values of Lx = 28–35 cm, Ly = 27–33 cm and
Lz = 17–21 cm. In the following an example with Lx = 29 cm, Ly = 29 cm and Lz = 19 cm
will be discussed. Equations (2) and (3) allow a number of resonator frequencies which lie close
to the emitted magnetron frequency. Table 1 demonstrates that there are six possible solutions in
the wavelength range between 12 cm and 12.5 cm.
The quality factor Q [1, 7] is a measure of the energy losses and, hence, also the frequency
width ω of the modes. Q gives the ratio of the energy stored in the resonator and the energy
loss per cycle:
Q = ωE/(dE/dt). (4)
For Q 1 this may be written as
Q ≈ ω/ω. (5)
The energy losses are due to four factors. First, microwaves may exit through the housing of the
microwave oven. Safety regulations ensure that this contribution is negligibly small. Second,
losses occur due to absorption in the walls, third due to absorption in the food within in the
cooking chamber (the desired mechanism), and fourth, there is the chance that microwaves are
coupled back into the magnetron. The latter mechanism may play a role if the oven is used empty,
and it should be avoided in order to ensure a long lifetime of the magnetron. The overall quality
factor can be calculated from
1/Q = 1/Q1 + 1/Q2 + . . . . (6)
If the oven is used empty, wall losses are most important. They can be estimated from the
penetration depth δ [1, 7]:
Qempty ≈ V /Sδ (7)
where V is the volume and S the inner surface area of the chamber. Typically, with air-filled
volumes of about 29 × 29 × 19 cm3 and δ ≈ 1 µm one finds Q-values of the order of 104 . For
frequencies of 2.45 GHz, this leads to frequency widths of only 0.25 MHz, i.e. the resonance
modes are very narrow compared with the magnetron frequency width. Figure 3 depicts
schematically a frequency spectrum of the modes of a cooking chamber. If some water containing
food is placed in the chamber, i.e. a dielectric is inserted, the additional losses—being much
larger than the wall losses—lead to a shift of the resonances to lower frequencies as well as an
appreciable broadening of the modes. Due to the shifts and broadening, described by a quality
factor that may well be of the order of 102 , more than one resonator mode may be excited. The
larger the oven and the more losses are present, the more modes may be excited simultaneously.
These multimode cavities have a more homogeneous field distribution, which is desirable for
many applications, including cooking.
3. How varies
(a)
1000
Q
(b)
100
Figure 4. Visualization of the horizontal mode
structure in a microwave oven using infrared thermal
imaging. A glass plate with a thin water film was
placed at a height of 8 cm and heated for 15 s with a
Frequency microwave power of 800 W without using the turntable
(for more details on the experiment, see [10]).
Figure 3. Schematic diagram showing how Q varies
with frequency for two different modes of resonance
in a microwave oven (not to scale): (a) with oven the wavelength of the microwaves is roughly the
empty, (b) with food. The shaded band shows the same as the linear dimensions of the chamber.
frequency of the magnetron. If drawn to scale the
widths of the modes in (b) would be about 100 times An ideal microwave oven cooks all food
those in (a). evenly but the nodes and antinodes of the standing
waves can cause the food to burn in some places but
Table 1. All modes of a microwave oven of to remain cool in others. The homogeneity of the
dimensions 29 cm × 29 cm × 19 cm, which lead to a field distribution may be estimated theoretically
wavelength in the interval 12.0 cm < λ0 < 12.5 cm. from the number of modes that may be excited
Four modes are quite close to the magnetron within a narrow frequency range close to the
wavelength of 12.25 cm.
magnetron frequency (see below) or studied
λ0 (cm) l m n experimentally [8].
12.103 1 1 3 Figure 4 shows the intensity distribution
12.274 2 4 1 within an oven of 29 × 29 × 19 cm3 at a
12.274 4 2 1 height of about 8 cm. A horizontal glass plate
12.277 2 3 2 covered with a thin film of water was placed in
12.277 3 2 2 a microwave (without its rotating turntable) on
12.375 0 1 3
full power (800 W) for about 15 s. The false
colour image was obtained with a thermal infrared
the waveguide with a dielectric shifts the cut- camera [9]. With only a small amount of water
off frequency and wavelength. Because only present, the image shows the microwave intensity
one inner dimension is greater than the cut-off distribution in a nearly empty chamber. There is
length the microwaves leaving the waveguide a pronounced horizontal mode structure, which
are polarized. As the waves are reflected in would lead to uneven heating of food. This is the
the cooking chamber the degree of polarization reason for having a rotating turntable: the rotation
decreases. will move the food in and out of the hot spots.
Some ovens have a mode stirrer, i.e. a rotating
reflector at the top to get a more homogeneous
The cooking chamber: why do microwave ovens field distribution, and there are investigations into
use a rotating turntable? how geometrical changes of the wall structure
Once the microwaves have been coupled into the may improve the situation. More details on
cooking chamber, they are effectively reflected by experiments on horizontal and vertical modes and
the metallic walls. The waves resonate in the also on changes of the mode structure upon filling
cavity and form standing waves. The analysis of of the cooking chamber with food will be discussed
these standing waves is simplified by the fact that in the following article [10].
0 °C
80
20 °C
40 °C
60 60 °C
80 °C
ε1 and ε2
100 °C
40 Domestic
microwave 0 °C
frequency=2.45 GHz 20 °C
20 40 °C
60 °C
80 °C
100 °C
0
100 10 1 0.1 0.01
Wavelength in cm
Figure 5. Real and imaginary parts of the dielectric constant ε(ω) for microwaves and various temperatures
between 0 °C and 100 °C (after [12]).
10
Figure 6 also explains why microwave ovens
Real part of
5
use a frequency of about 2.45 GHz rather than
1
20 to 1000 GHz, as may have been guessed from
VIS figure 5. With increasing frequency, α increases
106 rapidly, i.e. the penetration depth δ = 1/α
decreases rapidly. The food in microwave ovens
has typical dimensions of the order of cm, and
Absorption coefficient α/cm—1
104
hence the penetration depth should be in this range.
With a frequency of 20 GHz, the penetration depth
102 would be much smaller, i.e. the energy would
be absorbed in a thin surface layer of the food
100 (toasting the food) while the interior would remain
cold. The lower frequency chosen results in
10—2 absorption of the microwaves everywhere in the
λ = 12 cm food. Therefore the surface will only get a brown
crust if additional grilling facilities are available.
10—4 λ = 1 cm
gravy x
soup x water
70 mashed x x carrots
potato
x
x peas
60
50 x raw beef
x cooked cod ε1
ε2
x
ε 1, ε 2
30
gravy
mashed x
x potato
20 raw x x carrots
raw pork x xbeef peas x soup
x cooked x
cooked cod water x
10 beef
0
60 65 70 75 80 85 90 95 100
Water content as a percentage
penetration depth is much larger and less energy of electrical appliances. The radiation level close
is absorbed. to the surface of most ovens amounts to about
It is still much quicker to defrost food in 1% of the allowed limit of 5 mW cm−2 . As you
a microwave than to leave it in air at room get further from the oven the intensity rapidly
temperature. The oven is periodically turned on decreases, such that the usual radiation dose is well
for short periods and then off for longer time below 1/1000 of the maximum permitted value.
intervals. If by chance, some ice has melted at Since a small amount of radiation does escape
some spot, the liquid water will be heated in the from the oven, it is also possible for microwaves to
‘on-phase’. During the ‘off-phase’, the contact of enter the closed oven. This may be demonstrated
this heated water with the surrounding ice melts by putting a cell phone into the oven (turned off!)
more ice such that in the next ‘on-phase’, more and calling it [10].
water will be heated and so on. Most ovens have an ‘off’ button which
simultaneously shuts off the magnetron and opens
the door. Can any lingering radiation escape when
Can microwaves get out of the oven?
this happens? How long will the microwaves
There are very strict regulations governing how remain in the cooking chamber after the magnetron
much radiation is allowed to be emitted from is turned off? Using equation (4) one may easily
microwave ovens: they could be a health risk estimate the decay times of the fields, which
and they could interfere with other electronic depend on the quality factor of the chamber:
apparatus. Because a microwave oven is a Faraday
cage little radiation is expected to escape. The E(t) = E(0) exp(−ωt/Q). (9)
most crucial part is the door, which is equipped For an empty oven (Q = 10 ) the field has decayed
4
with additional so-called λ/4 radiation traps [1]. to less than 1/1000 after a time of about 4.5 µs;
In every country there are official institutions, in if there is food in the oven, reducing Q to about
Germany for example it is the Bundesamt für 102 , the time changes to only 45 ns. Obviously,
Strahlenschutz (BfS) [15], which check the safety nobody can open an oven door that fast.
Is it possible that microwaves can change the [2] Smith B L and Carpentier M-H 1993 The
food chemically? Microwave Engineering Handbook vols 1–3
(London: Chapman and Hall)
People often assume that microwaved food is [3] Bloomfield L A 1997 How Things Work: The
unhealthy. Could it be chemically different to Physics of Everyday Life (New York: Wiley)
conventionally heated food? For this to happen, [4] Hecht E 1998 Optics 3rd edn (New York:
the microwaves would need to create chemical Addison Wesley)
[5] Feynman R 1974 Feynman Lectures on Physics,
radicals involving energies of the order of an
Vol II, Part 2: Mainly electromagnetism and
electron-volt. Microwave photons have energies matter (New York: Addison
of the order of 10−5 eV. Simple estimates easily Wesley/Oldenbourg)
show [8] that the number of microwave photons [6] Haala J 2000 Analyse von
within a commercial oven is orders of magnitude Mikrowellenheizprozessen mittels
selbstkonsistenter finiter Integrationsverfahren
too small to establish multiphoton dissociation or
Dissertation Universität Karlsruhe
ionization. One may therefore conclude that the [7] Jackson J D 1975 Classical Electrodynamics
food cannot be altered chemically while heated in 2nd edn (New York: Wiley)
a microwave oven. [8] Karstädt D, Möllmann K-P and Vollmer M 2004
Physik in unserer Zeit issues 1 and 2 plus
additional material on the journals homepage
Are microwave ovens only used for heating food? www.wiley-vch.de/home/phiuz (in German)
Microwave ovens are used in many other [9] Karstädt D, Möllmann K-P, Pinno F and Vollmer
industrial fields, e.g. pasteurization of vegetables, M 2001 Phys. Teacher 39 371–6
[10] Parker K and Vollmer M 2004 Phys. Educ. 39
drying of paper or textiles, thermal treatment 82–90
of pharmaceutical products and vulcanization of [11] Föll H, Universität Kiel: www.tf.uni-
rubber and elastomers (see [1, 2]). kiel.de/matwis/amat/elmat en/makeindex.html
[12] Chaplin M, South Bank University, London:
www.lsbu.ac.uk/water/microwave.html
Acknowledgments [13] Bengtsson N E and Ohlsson Th 1974 Proc. IEEE
Stimulating discussions and help with some of the 62 44–55
experiments by Klaus-Peter Möllmann and Detlef [14] Hasted J B 1973 Aqueous Dielectrics (London:
Chapman and Hall)
Karstädt are gratefully acknowledged. [15] Information from the Bundesamt für
Strahlenschutz: 12.8.98, Infoblatt 02/98
Received 24 October 2003 www.bfs.de
PII: S0031-9120(04)71706-2
DOI: 10.1088/0031-9120/39/1/006
Michael Vollmer is a Professor of
experimental physics at Fachhochschule
Brandenburg. He has a Diploma, PhD
and Habilitation from the University of
References Heidelberg. His current activities include
[1] Thuery J 1992 Microwaves, Industrial, Scientific applied research in infrared imaging,
and Medical Applications (Boston, MA: experiments for physics education and
Artech House) teacher training seminars.