Biomedical Engineering Reference
In-Depth Information
3.2.11 conclusion on Phase Matching in isotropic Media
In conclusion, we summarize here the main characteristics of THG microscopy in isotropic media:
1. There is no THG from a homogeneous medium due to poor phase matching caused by the Gouy
phase shift of the focused excitation beam. However, signal is obtained for inclusions that have a
size close to the effective coherence length. Since this effective coherence length depends on the
axial extent of the Gouy phase shift, changing the excitation NA changes the relative THG signals
obtained from objects of different sizes.
2. Because of the asymmetry in both phase and intensity between the axial and lateral axes, the THG
signal also depends on the orientation of structures. Axial structures are generally more visible
than structures orthogonal to the optical axis.
3. THG is a forward-directed process, because THG coherent signal buildup needs phase matching
over a significant distance. B-THG usually gives a negligible contribution to the THG signal in
isotropic media. Epidetection is however possible for some geometries using backscattered light
[19,26,29].
3.2.12 Sources of contrast
Whereas the second-order susceptibility (χ
(2)
) is nonzero only in the case of anisotropic media, there
is no such requirement for the third-order susceptibility (χ
(3)
), and most isotropic materials (such as
glass and water) have a nonzero χ
(3)
. However, as previously mentioned, the parameter governing
signal level in THG microscopy is not the absolute value of the χ
(3)
, but rather the spatial variations
of χ
(3)
around interfaces. The peak THG signal observed at an interface between two media scales as
κ |
− <
I
, where κ depends on geometry as discussed in the previous sections,
I
ω
is the excita-
tion intensity, and α
1
− α
2
depends on the optical properties of the two media. Under high excitation NA
phase matching is dominated by the Gouy shift and α ≈ χ
(3)
, whereas in the case of moderate focusing,
α ≈ χ
(3)
/(
n
3ω
(
n
3ω
−
n
ω
)) [30].
A few articles have reported measurements of χ
(3)
of solvents [31,32] and biological liquids [30,33].
The general principle of these measurements is to measure the THG intensity at the interface between
a known material (glass coverslip) and a liquid to be characterized. In particular, these studies have
shown that lipids and water have very different χ
(3)
(−3ω;ω,ω,ω) at 1.2 μm (Figure 3.11) [30].
α
α
|
2
3
1
2
ω
λ = 1.18 µm
χ
(3)
(×10
−22
m
2
V
−2
)
|χ
(3)
− χ
(3)
water
|
2
Water
—
1.68 ± 0.08
0
NaCl 1M
Ions
1.79 ± 0.09
1.2 × 10
−2
Glucose 1M
Sugar
1.83 ± 0.08
2.2 × 10
−2
Glycine 1M
Amino acid
1.69 ± 0.13
1.0 × 10
−4
Triglycine 1M
Polypeptide
1.69 ± 0.12
1.0 × 10
−4
BSA 1mM
Protein
1.75 ± 0.13
4.9 × 10
−3
Triglycerides
Lipids
2.58 ± 0.5
0.81
Oil
Lipids
2.71 ± 0.5
1.06
BK7
Glass
2.79
1.2
FIgurE 3.11
Third-order nonlinear susceptibility χ
(3)
of biological liquids. Lipids have a large non-resonant
nonlinear susceptibility. Lipid/water interfaces provide strong χ
(3)
contrast, making lipid structures visible in THG
images of cells and tissues. The excitation wavelength is λ = 1180 nm. (Adapted from Débarre D and Beaurepaire E
2007.
Biophys. J.
92:603-612.)
