Letters

https://doi.org/10.1038/s41567-022-01639-3

Size-controlled quantum dots reveal the impact

of intraband transitions on high-order harmonic
generation in solids
Kotaro Nakagawa1, Hideki Hirori 1 ✉, Shunsuke A. Sato 2,3, Hirokazu Tahara 1, Fumiya Sekiguchi 1,
Go Yumoto 1, Masaki Saruyama 1, Ryota Sato1, Toshiharu Teranishi 1 and Yoshihiko Kanemitsu 1 ✉

Since the discovery of high-order harmonic generation (HHG) (HHG) in solids because in addition to interband transition, non-
in solids1–3, much effort has been devoted to understand its linear intraband transition is considered to be responsible for the
generation mechanism and both inter- and intraband transi- elementary excitation process behind this phenomenon. However,
tions are known to be essential1–10. However, intraband transi- the relation between nonlinear carrier generation and HHG has not
tions are affected by the electronic structure of a solid, and yet been experimentally studied, and thus, it has remained elusive
how they contribute to nonlinear carrier generation and HHG how intraband transitions correlate with the electronic structure of
remains an open question. Here we use mid-infrared laser a solid and eventually with nonlinear carrier generation and HHG.
pulses to study HHG in CdSe and CdS quantum dots, where To clarify the impact of intraband transitions on extreme nonlin-
quantum confinement can be used to control the intraband ear optical phenomena, it is necessary to study it from materials in
transitions. We find that both HHG intensity per excited vol- which the electronic band structure and thus intraband transition
ume and generated carrier density increase when the average can be freely controlled. Because the nature of the electronic bands
quantum dot size is increased from about 2 to 3 nm. We show of quantum dots (QDs) can be continuously tuned from atom-like
that the reduction in sub-bandgap energy in larger quantum discrete states to a solid-state band continuum simply by chang-
dots enhances intraband transitions, and this—in turn— ing their size, without changing the constituting elements20–23, they
increases the rate of photocarrier injection by coupling with are ideal materials to examine the role of intraband transitions in
interband transitions, resulting in enhanced HHG. carrier generation and its relation to HHG.
Transitions between the valence band (VB) and conduction Here we studied HHG in CdSe and CdS QD films. Figure 1a
band (CB) are responsible for photon absorption in a semiconduc- (dashed lines, left) illustrates a schematic of the energy levels of a
tor11. However, a strong nonlinear optical response is not necessarily QD; Eg is the QD bandgap energy and Δsub is the first sub-bandgap
governed only by such interband transitions, and thus, we need to energy. Both parameters depend on the QD diameter. Figure 1a
understand the interplay between inter- and intraband transitions. (middle) shows the absorption spectra of CdSe QDs with different
In the case that the electric field E(t) of a laser accelerates an elec- diameters (Supplementary Section I). Owing to the strong quantum
tron in the band of a solid (which constitutes an intraband transi- confinement in these small QDs, Eg is larger than in the bulk by up
tion), the temporal change in the electron wavenumber k(t) can be to several hundred millielectronvolts (the bandgap energies of bulk
described by ℏdk(t)/dt = eE(t), where e is the electron charge and ℏ CdSe and CdS are 1.75 and 2.58 eV, respectively), and the electronic
is the reduced Planck constant. Depending on the degree of change states become discrete24–27. Figure 1a (right) shows the excitation
in k, two regimes can be considered: conventional and extreme non- and detection geometry used to measure the HHG emission spec-
linear optical phenomena. In the conventional regime (that is, under tra of the CdSe QD film (Methods) and the typical transmission
resonant or near-resonant conditions), the optical field induces only electron microscopy (TEM) images. Figure 1b shows the high-order
small changes in the wavenumber, leading to a dominant contri- harmonic (HH) intensities per excited volume as a function of pho-
bution of the interband transitions12–14. In the extreme nonlinear ton energy ℏω, namely, IHHG(ℏω), for CdSe and CdS QD films under
regime where the excitation photon energy ℏω0 is much smaller excitation with linearly polarized mid-infrared (MIR) light. The
than bandgap energy Eg, efficient carrier acceleration is possible spectra extend from the visible to ultraviolet region and the peaks
without damaging the sample by excessive carrier generation. Thus, correspond to the 7th–13th orders. We find that IHHG(ℏω) tends to
it is possible to use strong long-wavelength laser fields that induce increase as the average QD diameter d increases, and it increases
large changes in the wavenumber. In this case, intraband transitions abruptly in the range from 2 to 3 nm.
play a major role in the extreme nonlinear dynamics of the system. Figure 2 provides data on CdSe (red circles) and CdS (blue
Recently, a rather unexpected role of intraband transitions was squares): it shows the integrated HH peak intensity of each order
found: carrier injection into the CB of GaAs is enhanced by cou- (Ih, where h is the harmonic order) as a function of d to clarify the
pling of inter- and intraband transitions15. Also, it has been shown QD size dependence of the HH intensity. For CdSe QDs, Ih sub-
that nonlinear-carrier-generation processes in the extreme non- stantially increases with d in the range d ≈ 1.8–3.8 nm. For example,
linear regime play an important role in the modification of opti- I7 increases by a factor of about 100 from d = 2.1 to 3.8 nm. Note
cal and electric properties of solids in the ultrafast timescales16–19. that the Eg value of CdSe QDs with d in the range of 2.1–3.8 nm
These phenomena are related to high-order harmonic generation changes from 2.6 to 2.1 eV, but the absorption spectra (Fig. 1a)

1
Institute for Chemical Research, Kyoto University, Uji, Japan. 2Center for Computational Sciences, University of Tsukuba, Tsukuba, Japan. 3Max Planck Institute
for the Structure and Dynamics of Matter, Hamburg, Germany. ✉e-mail: hirori@scl.kyoto-u.ac.jp; kanemitu@scl.kyoto-u.ac.jp

874 NatUre Physics | VOL 18 | August 2022 | 874–878 | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature Physics Letters
a
Order of multiphoton absorption Spectrometer
5 6 7 8 CCD camera
1.5 MIR
CdSe 14.0 nm

Energy
pulse

OD (normalized)
Eg 6.4 nm HH
Δsub
(QD) 1.0 3.8 nm QD films
Eg 2.8 nm
(Bulk) 2.4 nm CdSe 3.8 nm 14 nm
0.5 2.1 nm
1.8 nm

0
d 1.5 2.0 2.5 3.0 10 nm 10 nm
Photon energy (eV)
b
Harmonic order
7 9 11 13 15
CdSe d = 2.1 nm 2.8 nm 3.8 nm 14.0 nm
7th
100
9th
10–1 11th 13th

10–2 15th
IHHG (arb. units)

10–3
CdS d = 1.6 nm 2.1 nm 2.4 nm 3.5 nm
100

10–1

10–2

10–3

10–4
2 3 4 5 2 3 4 5 2 3 4 5 2 3 4 5
Photon energy (eV)

Fig. 1 | HHG in CdSe and CdS QD films. a, Energy-level diagram (left), absorption spectra of CdSe QDs with different sizes (middle) and experimental
setup including TEM images of CdSe QD films (right). The ODs are normalized by the lowest-energy exciton peaks. The TEM images of CdSe QD film
(3.8 and 14.0 nm) are shown. b, HH spectra of CdSe and CdS QD films with different average QD diameters. The plotted intensities are the measured
spectral intensity values divided by the excited volumes (that is, the number of excited QDs times the average volume of a single QD), which were
determined by dividing the absorbance at the band edge of the QD film by the absorption cross section per unit volume.

indicate that the required number of photons in the multiphoton measurements (Supplementary Sections IV and V). The QDs with
absorption process for the VB–CB transition remains almost the d ≤ 2.8 nm have very low excited carrier densities compared with
same (that is, the band edge lies in the region near h = 7 or 2.48 eV). those in larger QDs. Hence, the QD size dependence of nd has the
Although the bandgap energy for d = 2.4 nm lies closer to the 7th same tendency as that of HH intensity (Fig. 2). This implies that
resonant multiphoton absorption than that for 2.8 nm, the HH with respect to nonlinear responses such as HHG, the primary
intensities of the smaller QDs are much smaller. Also, an almost effect of a larger crystal is more efficient nonlinear carrier genera-
constant behaviour in the range d ≈ 3.8–14.0 nm (Fig. 2a, inset) is tion, and this results in larger HH intensities in the case of solids.
observed despite the difference in the order of multiphoton absorp- If laser light is applied to a semiconductor, carriers are acceler-
tion. In addition, the excitation intensity dependence shows that the ated by the electric field and their change in behaviour in real space
HHG mechanism in CdSe QDs under these excitation conditions is corresponds to an intraband transition in the Brillouin zone or
non-perturbative (Supplementary Sections II and III). These results k space. In the case of acceleration of excited carriers in a bulk crys-
show that the size dependence of Ih cannot be simply explained by tal, they pass through a continuum of states in momentum space.
the difference in the multiphoton absorption process considering However, for a sample with discrete electronic states (due to quan-
only interband transitions. tum confinement), intraband transitions are less likely to occur due
To obtain the relation between the actual carrier density gener- to energy gaps between the discrete states. Therefore, we discuss
ated by the MIR pump pulse and HHG, we measured the transient the suppression of HHG from the viewpoint of intraband transi-
absorption (TA) change in the CdSe QD films in experiments with tions: to elaborate the impact of quantum confinement on intra-
an MIR pump and a white-light probe. The schematic of the setup band transitions for HHG, we investigated the optically induced
is shown in Fig. 3a (Methods). Figure 3a also shows a typical TA electron dynamics in a simple one-dimensional dimer chain
spectrum (expressed in terms of the differential optical density model (Supplementary Section VI). Here we considered a system
ΔOD divided by optical density OD). The four graphs shown in Hamiltonian where the different sites of the chain correspond to
Fig. 3b show the TA dynamics integrated over the energy region two species of atoms, such as Cd and Se, and the coupling between
near Eg for different QD diameters. A large absorption change can the sites is represented by the nearest-neighbour hopping param-
be observed for larger QDs (d = 3.8 and 6.4 nm), whereas smaller eter. We decomposed the perturbation part of the Hamiltonian into
QDs (d = 2.4 and 2.8 nm) exhibit notably smaller changes. Figure 3c components corresponding to inter- and intraband transitions and
shows the average number of carriers per excited volume, nd, esti- calculated the HHG power spectrum I(ℏω) of a single QD.
mated from our TA data. The carrier density was determined by Figure 4a shows I(ℏω) divided by the QD volume, IHHG(ℏω), for
comparing the results of the above MIR-pump–white-light-probe different QD diameters (chain lengths). As shown in Fig. 4b, we deter-
TA measurement with those of visible-pump–white-light-probe TA mined the QD size dependence of I7 and nd by considering different

NatUre Physics | VOL 18 | August 2022 | 874–878 | www.nature.com/naturephysics 875

Content courtesy of Springer Nature, terms of use apply. Rights reserved
Letters Nature Physics

I7
components in the Hamiltonian are set to zero (Fig. 4b, blue curve,
0
and Supplementary Section VI), the intensities are substantially
10
reduced. This result indicates that the observed nonlinear responses
are determined by the size dependence of the intraband transitions.
Note that Eg increases as d becomes smaller (∝1/d2) and energy gap
CdSe
Δsub between the quantum states with quantum numbers n = 1 and
10–1
n = 2 is roughly proportional to n2/m*d2 (Extended Data Fig. 1a),
where m* is the reduced mass. To see which parameter governs the
CdS size dependence, we also evaluated the size dependence of I7 under
the assumption of a size-independent Eg. The obtained size depen-
10–2
dence (Extended Data Fig. 1b) resembles the red solid curve shown
in Fig. 4b, and this indicates that the actual increase in Eg for smaller
BG CdSe
QDs does not govern the observed size dependence. To understand
2 4 6 10 the effect of discretization of QDs on the HH intensity, we stud-
10–3 ied the influence of m* on the size dependence: Fig. 4c shows that
I9 I7 of QDs with small diameters (d ≈ 1 nm) becomes smaller as the
100 mass becomes smaller. This result is consistent with the experi-
mental results shown in Fig. 2, since the m* value of CdSe (0.1 m0)
is smaller than that of CdS (0.2 m0) (ref. 23). Moreover, the depen-
10–1 dence of I7 on Δsub, which can be changed by varying m*, is shown in
Intensity Ih (arb. units)

Extended Data Fig. 1c. The obtained curves of I7 as a function of Δsub

are similar regardless of the value of m*. This result shows that the
10–2 discrete electronic states due to confinement in small QDs suppress
intraband transitions.
The good agreement between the experimental QD size depen-
dence and our calculations shows that intraband transitions cause
I11
efficient nonlinear carrier generation and HHG in larger QDs.
100
Nonlinear carrier generation is a coherent optical process and
thus involves a superposition of multiple transitions from vari-
ous VB–CB excitation paths (Supplementary Section VI). It is due
10–1 to nonlinear coupling between inter- and intraband transitions:
in addition to the contribution from pure interband-transition
terms, the contribution from the coupling terms increases when
10–2 the intraband transitions are enhanced (Fig. 4d, schematic, and
Extended Data Fig. 2). In the case of coherent excitation of small
QDs, carrier acceleration (intraband transition) is drastically sup-
I13 pressed by discrete electronic states. Meanwhile, larger QDs (bulk
10 0 form) provide more efficient carrier acceleration and thus cause
an additional coupling between inter- and intraband transitions,
which results in efficient carrier generation and consequently
10–1
stronger HHG.
Note that although our calculations can explain the overall
behaviour of the experimental results (Figs. 2 and 3), there is a
slight quantitative discrepancy between the observed trends
10–2
(Figs. 2 and 3). To clarify this, we considered the yield ratio,
which is the HH intensity per ionization event, namely, I7/nd. In
2 3 4
Extended Data Fig. 3, we can confirm an increase in the yield ratio
Diameter (nm)
of the 7th harmonic as the QD diameter decreases from about
Fig. 2 | QD size dependence of HHG. Integrated peak intensity per excited
7 to 3 nm. Therefore, with respect to the number of electrons in
volume Ih as a function of d for different orders. The vertical and horizontal
the CB, the 7th harmonic is generated more efficiently in smaller
error bars represent the standard deviation of integrated peak intensity
QDs. In smaller QDs, which are tightly confined systems, the
as well as the diameter. The solid curves are a guide to the eye. We
number of excited carriers is reduced by the amount by which
normalized each curve to the value at d = 3.8 nm for CdSe (d = 3.5 nm for
intraband transitions are quenched as the sub-bandgap energy is
CdS). The excitation power densities were 0.45 and 0.75 TW cm–2 for CdSe
increased. On the other hand, the overlap of excited electrons and
and CdS, respectively. The dotted lines indicate the background level. The
holes increases in smaller QDs, and thus, the probability of
larger-gap semiconductor CdS has a higher breakdown threshold and can
recombination increases21,28. Therefore, the yield ratio of smaller
withstand longer exposures. Thus, the exposure time for CdS QDs was four
QDs can be larger.
times longer than that for CdSe QDs (1,000 and 250 s, respectively), which
By using size-controlled QDs, we show that quantum confine-
allows the background (BG) level of CdS to be lower than that of CdSe.
ment can indeed be used to control intraband transitions and
therefore influence the carrier density and HHG. This means that
the size of a structure constitutes a parameter that can be used to
calculation conditions. The red solid and dashed curves are results control extreme nonlinear optical phenomena. Our findings on
obtained when both inter- and intraband transitions are considered the control of nonlinear optical phenomena by nanosizing can be
and the QD energy levels depend on d (full model). These curves used in designing sophisticated petahertz optoelectronic devices
show that I7 and nd abruptly increase when the QD size becomes that will be implemented at the nanoscale29,30. The impact of intra-
larger than 3 nm. On the other hand, when the intraband-transition band transitions on carrier excitation shown in our work has

876 NatUre Physics | VOL 18 | August 2022 | 874–878 | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature Physics Letters
a d = 6.4 nm c
0
8
6
CCD camera ΔOD/OD 4

Carrier density, nd (cm–3)

Delay time (ps)
200
MIR 0.20
2
pump
0.15
Spectrometer 1018

0.10 8
CdSe QD films 400 6
White-light probe BG
0.05 4

0
2
1.80 1.90 2.00 2.10 3 4 5 6 7
Photon energy (eV) Diameter (nm)

b
0.30
d = 2.4 nm 2.8 nm 3.8 nm 6.4 nm
0.25

0.20
ΔOD/OD

0.15 (×5)
0.10
(×30)
0.05
(×30)
0

0 200 400 0 200 400 0 200 400 0 200 400

Delay time (ps)

Fig. 3 | TA measurements. a, Schematic of the experiment and typical TA spectrum of CdSe QDs obtained using an MIR pump pulse (3 μm, 0.36 TW cm–2).
b, Time evolution of ΔOD/OD near the bandgap energy for d = 2.4, 2.8, 3.8 and 6.4 nm. Here ΔOD/OD was obtained by measuring the sample
transmissivities with and without pump pulse excitation. c, Average number of carriers per excited volume, nd, estimated from the exciton amplitudes in
TA signals. The vertical and horizontal error bars represent the standard deviation calculated from the exciton amplitude as well as the diameter. The solid
curve is a guide to the eye. Since the TA signal of the sample with d = 2.4 nm was smaller than the background level, the point at d = 2.4 nm plots the upper
limit of nd estimated from the background level.

a b
Harmonic order
7 9 11 13 15 17 19

Calculation 100 100

100

Carrier density, nd (arb. units)

10–4 Full model 10–4
IHHG (arb. units)

10–8
I7 (arb. units)

(intraband + interband)
10 nm
10–8 10–8
–16
10

2.1 nm 10–12 Without intraband 10–12

0.9 nm
10–24
2 3 4 5 6 7 0 2 4 6 8 10 12
Photon energy (eV) Diameter (nm)

c d
QD Bulk
100
Energy

Energy

10–4
I7 (arb. units)

10–8
0.20m0
Intraband
10–12 0.10m0

Eg Coupling
10–16
Interband Interband
10–20 m* = 0.05m0

1 10
Diameter (nm) Density of states Density of states

Fig. 4 | Calculation results. a, HH spectra per excited volume for different QD diameters (chain lengths). For this calculation, we assumed ℏω0 = 0.35 eV
and E = 11 MV cm–1, which resembles the experimental conditions. b, Diameter dependence of I7 and nd obtained by using the full model (red circles and
squares) and diameter dependence of I7 obtained by using a model excluding the intraband-transition term (blue triangles). c, Dependence of I7 on the
diameter for different reduced masses. d, Schematic describing nonlinear carrier generation via an additional path including excitation processes due to
the coupling of inter- and intraband transitions.

NatUre Physics | VOL 18 | August 2022 | 874–878 | www.nature.com/naturephysics 877

Content courtesy of Springer Nature, terms of use apply. Rights reserved
Letters Nature Physics

important implications for light-driven control of material prop- 14. Mücke, O. D., Tritschler, T., Wegener, M., Morgner, U. & Kärtner, F. X. Role
erties and also for micromachining31,32, which can be realized, for of the carrier-envelope offset phase of few-cycle pulses in nonperturbative
resonant nonlinear optics. Phys. Rev. Lett. 89, 127401 (2002).
example, by using two light sources with different wavelengths and 15. Schlaepfer, F. et al. Attosecond optical-field-enhanced carrier injection into
a tunable phase offset or different polarization states to manipulate the GaAs conduction band. Nat. Phys. 14, 560–564 (2018).
intraband transitions33. 16. Schiffrin, A. et al. Optical-field-induced current in dielectrics. Nature 493,
70–74 (2013).
17. Schultze, M. et al. Controlling dielectrics with the electric field of light.
Online content Nature 493, 75–78 (2013).
Any methods, additional references, Nature Research report- 18. Sommer, A. et al. Attosecond nonlinear polarization and light–matter energy
ing summaries, source data, extended data, supplementary infor- transfer in solids. Nature 534, 86–90 (2016).
mation, acknowledgements, peer review information; details of 19. Higuchi, T., Heide, C., Ullmann, K., Weber, H. B. & Hommelhoff, P.
author contributions and competing interests; and statements of Light-field-driven currents in graphene. Nature 550, 224–228 (2017).
20. Wang, F. et al. Exciton polarizability in semiconductor nanocrystals.
data and code availability are available at https://doi.org/10.1038/ Nat. Mater. 5, 861–864 (2006).
s41567-022-01639-3. 21. McDonald, C. R., Amin, K. S., Aalmalki, S. & Brabec, T. Enhancing high
harmonic output in solids through quantum confinement. Phys. Rev. Lett.
Received: 21 December 2021; Accepted: 13 May 2022; 119, 183902 (2017).
Published online: 23 June 2022 22. Efros, Al. L. & Efros, A. L. Interband absorption of light in a semiconductor
sphere. Sov. Phys. Semicond. 16, 772–775 (1982).
23. Brus, L. E. A simple model for the ionization potential, electron affinity, and
References aqueous redox potentials of small semiconductor crystallites. J. Chem. Phys.
1. Chin, A. H., Calderón, O. G. & Kono, J. Extreme midinfrared nonlinear
79, 5566–5571 (1983).
optics in semiconductors. Phys. Rev. Lett. 86, 3292–3295 (2001).
24. Ekimov, A. I. et al. Absorption and intensity-dependent photoluminescence
2. Ghimire, S. et al. Observation of high-order harmonic generation in a bulk
measurements on CdSe quantum dots: assignment of the first electronic
crystal. Nat. Phys. 7, 138–141 (2011).
transitions. J. Opt. Soc. Am. B 10, 100–107 (1993).
3. Schubert, O. et al. Sub-cycle control of terahertz high-harmonic
25. Murray, C. B., Norris, D. J. & Bawendi, M. G. Synthesis and characterization
generation by dynamical Bloch oscillations. Nat. Photon. 8,
of nearly monodisperse CdE (E = sulfur, selenium, tellurium) semiconductor
119–123 (2014).
nanocrystallites. J. Am. Chem. Soc. 115, 8706–8715 (1993).
4. Golde, D., Meier, T. & Koch, S. W. High harmonics generated
26. Alivisatos, A. P. Semiconductor clusters, nanocrystals, and quantum dots.
in semiconductor nanostructures by the coupled dynamics
Science 271, 933–937 (1996).
of optical inter- and intraband excitations. Phys. Rev. B 77,
27. Pietryga, J. I. et al. Spectroscopic and device aspects of nanocrystal quantum
075330 (2008).
dots. Chem. Rev. 116, 10513–10622 (2016).
5. Kuehn, W. et al. Coherent ballistic motion of electrons in a periodic potential.
28. Liu, H. et al. High-harmonic generation from an atomically thin
Phys. Rev. Lett. 104, 146602 (2010).
semiconductor. Nat. Phys. 13, 262–265 (2017).
6. Hohenleutner, M. et al. Real-time observation of interfering
29. Sivis, M. et al. Tailored semiconductors for high-harmonic optoelectronics.
crystal electrons in high-harmonic generation. Nature 523,
Science 357, 303–306 (2017).
572–575 (2015).
30. Ludwig, M. et al. Sub-femtosecond electron transport in a nanoscale gap.
7. Luu, T. T. et al. Extreme ultraviolet high-harmonic spectroscopy of solids.
Nat. Phys. 16, 341–345 (2020).
Nature 521, 498–502 (2015).
31. Lenzner, M. et al. Femtosecond optical breakdown in dielectrics. Phys. Rev.
8. Vampa, G. et al. Linking high harmonics from gases and solids. Nature 522,
Lett. 80, 4076–4079 (1998).
462–464 (2015).
32. Jürgens, P. et al. Origin of strong-field-induced low-order harmonic
9. Garg, M. et al. Multi-petahertz electronic metrology. Nature 538,
generation in amorphous quartz. Nat. Phys. 16, 1035–1039 (2020).
359–363 (2016).
33. Sanari, Y., Otobe, T., Kanemitsu, Y. & Hirori, H. Modifying angular and
10. Aversa, C. & Sipe, J. E. Nonlinear optical susceptibilities of semiconductors:
polarization selection rules of high-order harmonics by controlling electron
results with a length-gauge analysis. Phys. Rev. B 52, 14636–14645 (1995).
trajectories in k-space. Nat. Commun. 11, 3069 (2020).
11. Haug, H. & Koch, S. W. Quantum Theory of the Optical and Electronic
Properties of Semiconductors (World Scientific, 2004).
12. Boyd, R. W. Nonlinear Optics 3rd edn (Academic Press, 2008). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
13. Cundiff, S. T. et al. Rabi flopping in semiconductors. Phys. Rev. Lett. 73, published maps and institutional affiliations.
1178–1181 (1994). © The Author(s), under exclusive licence to Springer Nature Limited 2022

878 NatUre Physics | VOL 18 | August 2022 | 874–878 | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature Physics Letters
Methods (λ0 = 517 nm) and temporal pulse width of 180 fs), we used frequency doubling of
QD synthesis. The CdSe and CdS QDs were prepared in our laboratory by the fundamental pulses in a beta barium borate crystal. Here we used QDs with
wet chemical synthesis, which provides precise control over the QD size d = 2.4, 2.8, 3.1, 3.8, 4.5 and 6.4 nm because the band-edge exciton energies were
(Supplementary Section I). 2.40, 2.31, 2.24, 2.19 and 1.91 eV, respectively, allowing a band-to-band excitation
by the second harmonic of the fundamental pulse (ħωvis = 2.40 eV). The probe
HHG spectrum measurements. For the measurement of the HH spectra, QD light transmitted through the sample was spectrally resolved by a spectrometer
thin films were excited by linearly polarized MIR pulses with ħω0 = 0.35 eV and detected using a CCD camera. To obtain a high signal-to-noise ratio, we
(λ0 = 3.5 μm). The MIR pulses were generated using a multistage optical synchronously chopped the pump beam at 500 Hz, thereby blocking every second
parametric amplifier system (OPerA with the NDFG1 option, Coherent) driven pump pulse. A computer collected the signals to obtain the TA signal, which was
by a Ti:sapphire laser with a repetition rate of 1 kHz. The QD thin-film samples calculated from the probe transmission with and without the influence of the
were prepared by the spin coating of colloidal QDs dispersed in hexane on a pump light.
sapphire substrate with a thickness of 0.5 mm. For all the QD samples, hexane
was evaporated after spin coating on the substrate, and it was confirmed that the Data availability
substrate and surface-protecting ligands for QD do not contribute to the HH Source data are provided with this paper. All other data that support the
spectra. The experiments were conducted with excitation intensities below the plots within this paper and other findings of this study are available from the
damage threshold of the QDs; the excitation peak intensities were 0.45 TW cm–2 corresponding authors upon reasonable request.
for CdSe QDs and 0.75 TW cm–2 for CdS QDs (the pulse width was 80 fs and spot
diameter was ~300 μm). We verified that linear absorption at the excitation spot Acknowledgements
did not change during the MIR pulse irradiation and that the observed change in Y.K. acknowledges support from the Japan Society for the Promotion of Science
IHHG for smaller QDs was not due to sample damage (Supplementary Fig. 3). All the (JSPS KAKENHI grant no. JP19H05465).
experiments were performed at room temperature. The HHs were detected using
a charge-coupled-device (CCD) camera (PIXIS, Teledyne Princeton Instruments)
attached to a spectrometer (SpectraPro, Teledyne Princeton Instruments). Author contributions
K.N. and H.H. carried out the experiments. K.N., H.H., S.A.S., H.T., F.S., G.Y. and Y.K.
TA spectroscopy. In the MIR-pump–visible-probe TA measurements, we used analysed the data. S.A.S. performed the simulations. M.S., R.S. and T.T. synthesized the
an MIR excitation pulse with ħω0 = 0.41 eV (λ0 = 3 μm and the temporal pulse QDs. H.H. and Y.K. conceived and supervised the project. All the authors discussed the
width was 100 fs), which was generated from a home-made optical parametric results and contributed to the writing of the paper.
amplifier system based on an Yb:KGW femtosecond laser system (PHAROS,
Light Conversion) with a centre wavelength of 1,033 nm, pulse width of 180 fs Competing interests
and 2 mJ per pulse at a repetition frequency of 1 kHz. To avoid the additional The authors declare no competing interests.
small thermal-decay contribution in the TA signal due to the ligands of QDs, it
was necessary to use an MIR excitation wavelength (3 μm) that is shorter than
the one used in the HH spectrum measurements (Fig. 1). We verified that the Additional information
different MIR excitation wavelengths did not cause a notable change in the size Extended data is available for this paper at https://doi.org/10.1038/s41567-022-01639-3.
dependence of Ih (Supplementary Fig. 4). For the probe pulses, white light was Supplementary information The online version contains supplementary material
generated by focusing a small fraction of the output beam of the Yb:KGW laser available at https://doi.org/10.1038/s41567-022-01639-3.
into a 10-mm-thick quartz cell containing water (Supplementary Section V). This
Correspondence and requests for materials should be addressed to
method provided a better signal-to-noise ratio compared with a Ti:sapphire laser
Hideki Hirori or Yoshihiko Kanemitsu.
owing to better laser stability. The generated continuum spectrum ranged from
the visible to near-infrared regime. The MIR pump and visible probe beams were Peer review information Nature Physics thanks the anonymous reviewers for their
focused onto the sample and the beam spot diameters were about 250 and 150 μm, contribution to the peer review of this work.
respectively. For the TA measurements with visible pump pulses (ħωvis = 2.40 eV Reprints and permissions information is available at www.nature.com/reprints.

NatUre Physics | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Letters Nature Physics

Extended Data Fig. 1 | Additional calculation results. a, Bandgap energy Eg (blue squares) and subband gap, Δsub (red circles), as a function of the QD
diameter (chain length). b, Diameter dependence of I7 obtained by assuming a size-independent Eg (green squares) and that obtained by the full model
with a size-dependent Eg (red circles). c, Dependence of I7 on Δsub for different reduced masses.

NatUre Physics | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature Physics Letters

Extended Data Fig. 2 | Schematics of multiple excitation paths. In addition to the contribution of the pure interband transition terms (left), the efficient
intraband transition in larger QDs (or bulk) opens multiple excitation paths due to the nonlinear coupling between the intra- and interband transitions
(right). These additional excitation channels due to the coupling promote nonlinear carrier injection and enhance HHG in larger QDs.

NatUre Physics | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Letters Nature Physics

Extended Data Fig. 3 | Yield ratio. Diameter dependence of the yield ratio of the 7th order for CdSe, I7/nd. The data is normalized to the value at
d = 6.4 nm. Vertical and horizontal error bars represent the standard deviation of yield ratio and that of diameter. The solid curve is a guide to the eye.

NatUre Physics | www.nature.com/naturephysics

Content courtesy of Springer Nature, terms of use apply. Rights reserved

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

onlineservice@springernature.com