Ultrafast universal quantum control of a quantum-dot charge qubit using Landau-Zener-Stuckelberg interference

Abstract

A basic requirement for quantum information processing is the ability to universally control the state of a single qubit on timescales much shorter than the coherence time. Although ultrafast optical control of a single spin has been achieved in quantum dots, scaling up such methods remains a challenge. Here we demonstrate complete control of the quantum-dot charge qubit on the picosecond scale, orders of magnitude faster than the previously measured electrically controlled charge- or spin-based qubits. We observe tunable qubit dynamics in a charge-stability diagram, in a time domain, and in a pulse amplitude space of the driven pulse. The observations are well described by Landau-Zener-Stückelberg interference. These results establish the feasibility of a full set of all-electrical single-qubit operations. Although our experiment is carried out in a solid-state architecture, the technique is independent of the physical encoding of the quantum information and has the potential for wider applications.

ARTICLE

Received 15 Jun 2012 | Accepted 20 Dec 2012 | Published 29 Jan 2013

Gang Cao1,*, Hai-Ou Li1,*, Tao Tu1, Li Wang1, Cheng Zhou1, Ming Xiao1, Guang-Can Guo1,

Hong-Wen Jiang2 & Guo-Ping Guo1

A basic requirement for quantum information processing is the ability to universally control the state of a single qubit on timescales much shorter than the coherence time. Although ultrafast optical control of a single spin has been achieved in quantum dots, scaling up such methods remains a challenge. Here we demonstrate complete control of the quantum-dot charge qubit on the picosecond scale, orders of magnitude faster than the previously measured electrically controlled charge- or spin-based qubits. We observe tunable qubit dynamics in a charge-stability diagram, in a time domain, and in a pulse amplitude space of the driven pulse. The observations are well described by LandauZenerStckelberg interference. These results establish the feasibility of a full set of all-electrical single-qubit operations. Although our experiment is carried out in a solid-state architecture, the technique is independent of the physical encoding of the quantum information and has the potential for wider applications.

DOI: 10.1038/ncomms2412 OPEN

Ultrafast universal quantum controlof a quantum-dot charge qubit using LandauZenerStckelberg interference

1 Key Laboratory of Quantum Information, Department of Optics and Optical Engineering, University of Science and Technology of China, Chinese Academy of Sciences, Hefei 230026, China. 2 Department of Physics and Astronomy, University of California at Los Angeles, California 90095, USA. * These authors contributed equally to this work. Correspondence and requests for materials should be addressed to G.-P.G. (email: mailto:[email protected]

Web End [email protected] )

NATURE COMMUNICATIONS | 4:1401 | DOI: 10.1038/ncomms2412 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2412

Universal single-qubit gates are key elements in a quantum computer, as they provide the fundamental building blocks for implementing complex operations13. In the

standard circuit model, arbitrary single-qubit rotations1, together with two-qubit controlled-NOT gates, provide a universal set of gates. In the alternative measurement-based models, such as the one-way quantum computer, the ability to carry out single-qubit operations from the source of a specic multi-particles state can generate every possible quantum state2 and offer practical algorithms. Additionally, with only single-qubit operations and teleportation, one can construct a universal quantum computer3. In the Bloch sphere model of qubit states, a universal single-qubit gate requires arbitrary rotations around at least two axes.

The charge4,5 or spin69 degrees of freedom of an electron in quantum dots are particularly attractive for the implementations of qubits. Owing to the fast charge or spin decoherence times in semiconductor quantum dots, which are typically less than a few nanoseconds49, control operating on the picosecond timescale may be necessary. Until now, ultrafast manipulations of a single qubit in quantum dots have been performed using pulsed laser elds1012. Alternatively, electrical pulses can be generated much more easily by simply exciting a local electrode. Logic gate operations and readouts can be carried out all-electrically, much like those in current mainstream semiconductor electronics. Additionally, this all-electric technique provides a simple pathway for greater spatial selectivity to locally address the individual qubits and remove obstacles to scalability. Therefore, universal electrical control on a picosecond scale is highly desirable to allow coherence to be maintained during the completion of a large number of operations.

Over the last few decades, the LandauZenerStckelberg (LZS) interference has served as a textbook model for quantum phenomena13, that occurs when a system sweeps through the anti-crossing of two energy levels. LZS has also gained particular interest for quantum control8,14 because it is less sensitive to certain types of noise and might enable the implementation of a universal gate with high delity1517. Here we experimentally demonstrate such a scheme for a single-charge qubit in a double quantum dot (DQD), using a single pulse. We may add that very recently LZS interference has been observed under a continuous microwave driving, in both an electrostatic-dened18 and a donor-based19 semiconductor DQD systems.

ResultsCharge qubit in a DQD. Figure 1a provides a scanning electron micrograph of the sample used in the experiments, in which the metal gate pattern electrostatically denes a DQD and a nearby quantum point contact (QPC) detector within a GaAs/AlGaAs heterostructure. All measurements were conducted in a dilution refrigerator equipped with high-frequency lines (details of sample structures and experimental techniques are given in the Methods section). A single-electron charge in the DQD is used to encode the charge qubit4,5,20. An excess valence electron in the left and right dots denes the basis states |LS and |RS, respectively. The schematic diagram in Fig. 1b illustrates that the energies of the qubit states can be continuously tuned by the level detuning e ER EL, in which EL and ER are the energy levels for the

electron in the left and right dots for an uncoupled DQD, respectively. In the presence of inter-dot tunnelling that couples two dots, a characteristic anti-crossing occurs between the twoqubit levels near the resonance (e 0). For our experiments, the

anti-crossing gap 2D is adjusted to 20.7 meV, which corresponds to a Rabi frequency of 5.0 GHz. We have experimentally determined the Rabi frequency from the coherent oscillations excited by a square, non-adiabatic pulse as shown in

Supplementary Fig. S1b. We denote the ground and excited states as |0S and |1S, which are roughly the charge eigenstates far from the resonance.

The characterization of the system is shown in the charge-stability diagram for the DQD (Fig. 1c), which integrates the few-electron regime so that B23 electrons occupy each dot (for the full diagram, see Supplementary Fig. S1a). Our experiments are performed for a variety of charge states. As they contain identical physics, for consistency, we present only the data collected from one charge conguration. The system can be conveniently described in a valence electron number conguration that consists of four relevant charge states: (0,0), (1,0), (0,1) and (1,1). A prominent boundary can be observed between (1,0) and (0,1) and marks the inter-dot transition line corresponding to the e 0 resonance.

Observations of LZS interference. Our scheme to control a single-charge qubit using a Gaussian-shaped short pulse is shown in Fig. 2a. The system is initially prepared in the |RS state at a positive detuning e0, which is far from the resonance. During the rising phase of the pulse, the sweeping pulse takes the system adiabatically to the anti-crossing point at t t1 at which a sig

nicant probability exists for a non-adiabatic transition to the excited state |1S. This probability is the LandauZener transition represented by the following formula:

PLZ exp 2pD2/u hin which u is the sweep velocity of the driven pulse through the anti-crossing point.

As the pulse takes the system further past e 0, two different

trajectories at different energies can coherently interfere. Upon returning to e 0 at t t2, the two trajectories, caused by the

coherent interference, have also accumulated a phase difference of magnitude

fi

1 h

Zt2

t1

E1t E0tdt:

A projective read-out is performed at the end of the pulse to measure the |LS state for a constructive interference and the |RS state for a destructive interference, known as the LZS interference. Thus, the LZS process consists of both the non-adiabatic level transition and the adiabatic phase accumulation.

The Bloch sphere model provides a convenient picture to understand the quantum control of a charge qubit. Using this model, the charge state is represented as a vector, in which the ground and excited states |0S and |1S are at the north and south poles, respectively. In this model, the dynamics of the qubit can be represented by applying the appropriate sequence of unitary operation matrices to the initial state. The matrices

Rxy exp iysx/2;

Rzf exp ifsz/2f;give rise to a rotation on the Bloch sphere around the x axis by an angle y and around the z axis by an angle f.

At a LandauZener transition, the initial state becomes a coherent superposition of |1S and |0S with a phase fLZ related to the Stokes phenomenon13. The relative amplitudes of |1S and |0S depend on PLZ. This behaviour corresponds to the transformation Rz( fLZ)Rx(yLZ)Rz( fLZ), seen as successive

x and z-rotations (see the Supplementary Discussion for details), and yLZ 2 sin 1 OPLZ. In the phase accumulation stage, the

qubit undergoes a single rotation about the z axis, referred to as the phase-shift gate operation Rz(fi). Thus, the combination of Rx

2 NATURE COMMUNICATIONS | 4:1401 | DOI: 10.1038/ncomms2412 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2412 ARTICLE

A1

0.24

(1.0)

0.12

0.245

0.25

0.255

0.28 0.275 0.27

VB3(V)

A4

L R

GQPC

(1.1)

= 0

V A3(V)

0.08

E

|1

|0

2

(0.0)

(0.1)

A2 A3 B1 B3 B2

B4

G(e2/h)

0.265 0.26

0.04

0

|L = (1.0) |R = (0.1)

0

= ER EL

Figure 1 | Experimental set-up to characterize, control and measure single-electron charges. (a) Scanning electron micrograph of the connement gates that dened the DQD and the QPC charge-sensing channel. The two dots have a lithographic dimension of B300 nm. (b) Energy diagram of the DQD charge qubit. The green and yellow solid lines represent the energy levels for the bonding and anti-bonding states, respectively. With a nite inter-dot coupling, the lines are anti-crossed near the resonance point. (c) Colour scale plot of the charge-sensing signal as a function of the gate voltages VA3 and

VB3. The notation (n, m) represents the effective left and right dot occupancy.

and Rz enables arbitrary one-qubit rotations R(y, f) on the Bloch sphere and the LZS pulse rotates the input state on the Bloch sphere to the output state

Coutj i Rz fLZRxyLZRz fLZRzfiRz fLZ

RxyLZRz fLZ Cin

j i Ry; f Cin

j i

as illustrated in Fig. 2a.

The coherent control of our charge qubit is evident in the stability diagram (Fig. 2b), which is considerably altered as compared with that without an applied pulse. We observe many additional lines parallel to the inter-dot transition line. These additional lines are a signature of the excitation of the LZS interference by the pulse. In our device, the right barrier to the bath is slightly more open than the left barrier, and one can observe that the charge-addition lines (0,0)(1,0) and (0,1) (1,1) are less visible. As a result, an interesting triangular-shaped area exists in which the electron has a high probability of escaping to the reservoir before the inter-dot transition is completed.

The additional lines in Fig. 2b can be easily understood. For the line labelled 0, the pulse takes the system just past the anti-crossing point. At the next line, the pulse can take the qubit further, passing e 0 and accumulating a total phase of 2p.

Therefore, the lines represent the constructive interference fringes between the successive LandauZener transitions that correspond to an accumulated total phase of 2pN.

To conrm our identication, we have derived an analytical expression for the locations, e0(N), of the constructive interference ngers for a triangular pulse21. The triangular pulse is a simple approximation of the actual Gaussian prole of our pulse and can readily yield an intuitive analytical expression (see the Supplementary Discussion), eN0 A

2p hAN/tr

p , in which A is the pulse amplitude with units of energy that can be converted from voltage using the lever-arm conversion factor and tr is the pulse-raising time. Figure 2c compares the experimental nger positions with a theoretical curve obtained using the above equation, and reasonable agreement can be achieved.

Complete and ultrafast quantum control of the charge states. The ability to fully control the charge qubit is also studied using the LZS interference patterns. Controlling the amplitude of the driven pulse while the time prole of the pulse is xed sets the speed u of the passage through the anti-crossing point, thus the Rx(yLZ) rotation angle yLZ and the Rz(fLZ) rotation angle fLZ. Tuning the pulse time interval in the phase accumulation stage

sets the Rz(fi) rotation angle fi. The parameters yLZ, fLZ and fi are sufcient to rotate the input qubit to any point on the Bloch sphere, or more generally, to implement a universal one-qubit operation R(y, f) using the LZS pulse prole22,23.

To demonstrate the ability of the LZS method to generate tunable unitary transformations, we use the amplitude of the driven source as a control parameter. The charge state probability P|LS as a function of the qubit detuning position e0 and the voltage amplitude A under a 150-ps short pulse is provided in

Fig. 3a. Given the detuning and driven pulse amplitude, such interference patterns exhibit fringes that rise again from the constructive interference between successive LandauZener transitions at f as a multiple of 2p. A characteristic of the LZS driving occurs when the z-rotations result in the total phase f 2pN; the total x-rotation angle y is generally maximized as

2yLZ, which increases monotonically with the driven amplitude (see Supplementary Discussion for details). We verify this result by extracting the total rotation angles (y, f) of the Bloch vector from Fig. 3b (a horizontal cut at e0 400 meV of Fig. 3a) as it

undergoes LZS interference. These rotation angles are also parametrically plotted in Fig. 3c,d, demonstrating the expected behaviour as a function of the pulse amplitude. In addition, we note that the t of f is better than that of y, indicating that the phase-shift gates have higher intrinsic resistance to certain decoherence15,16. These ndings suggest that the LZS pulse amplitude should be an important tuning parameter for optimizing the quantum control of two-level systems.

The LZS interferences are further studied in the time-domain. We use a low-pass lter to shape the time-varying, square-driven pulse into an approximately Gaussian prole for the investigations. Figure 4 shows the charge state occupation P|LS as a function of both the detuning energy position e0 and the pulse width tp. Up to 10 LZS interference fringes can be clearly observed. The brightest line in the gure can be understood as the detuning pulse precisely at the anti-crossing point, and the subsequent ner lines corresponds to full phase accumulations of 2p, 4p and so on. We also simulate the evolution of the charge qubit by numerically solving the master equations as described in the Supplementary Discussion. This simulation (as shown in Supplementary Fig. S2) is in reasonable agreement with the experimental data. In particular, fast time-evolutions are observed in the insert of Fig. 4. For example, at a detuning energy of e0

400 meV, only B10 ps is required to accumulate a phase of 2p that corresponds to a full cycle of Rz operations of the qubit.

NATURE COMMUNICATIONS | 4:1401 | DOI: 10.1038/ncomms2412 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2412

a

Path 1

Path 2

i

initialization LZ-transition

Phase accumulation

Read-out

R ( , )

|0

RZ( LZ)

RX( LZ)

Rz( LZ)

Rz( i) Rz( LZ) RX( LZ)

Rz( LZ)

|0

|0

z |

0z |

0

z

z

LZ

LZ

i

LZ

y x

y

x

y x

y x

LZ

|1 |1 |1 |1

b

0.235

c

600

(1,0)

(0,0)

0.245

0.25

0.255

0.15

0.1

0.05

0

0.24

[afii9830] = 0

(1,1)

400

V A3(V)

(N) ( eV)

A? 0

200

4

2

0.26

0.27

0.265

0

(0,1)

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13

0.26

V B3(V)

0.255

0.25

0.245

G(e2/h)

N

Figure 2 | Experimental demonstration of the Rx and Rz operations via the LZS interference. (a) Illustration of the LZS interference evolution in the four different time stages of the driven pulses, indicated by the solid dots. The diagrams in the below schematically illustrate the circuits for unitary transformations on one qubit and the resulting evolution of the qubit trajectories on the Bloch sphere. The bracket highlights the decomposition of the arbitrary single-qubit operation. (b) Under a short driven pulse, the charge-stability diagram in Fig. 1c has been largely altered and a sequence of additional lines has emerged. (c) The construction interference nger locations e0(N) (dots) along the line in (b) as a function of the interference order N. The error bars are determined from the least-squares t to the data. The solid line is obtained from a theoretical expression.

For gate dened GaAs qubits, the phase rotation time can be electrically manipulated at this time scale.

To further highlight the importance of the LZS interference for a general single-qubit gate, we consider the following example: Tuning the speed of passage to yield PLZ 1/2, the total rotation

is a Hadamard gate with an arbitrary phase22. As the Rabi frequency can be reliably tuned to 10 GHz, the total rotation can be completed in B50 ps.

Decoherence information in the amplitude spectroscopy. Decoherence of the qubit due to its environment can be readily

extracted from the amplitude spectroscopy24,25 (for details, see the Supplementary Discussion). It is useful to evaluate the Fourier transform of the occupation probability

PFTkA; ke Z Z exp ikAA ikeePA; ededA

in which kA and ke represent the reciprocal-space variables corresponding to the real-space variables A and e, respectively. The two-dimensional Fourier transform of the data in Fig. 3a is provided in Fig. 5. As ke is practically a timescale, decoherence

4 NATURE COMMUNICATIONS | 4:1401 | DOI: 10.1038/ncomms2412 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2412 ARTICLE

0 1

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0 400 500 600 700 800

|0

z

0.8

= 242

200

400

600

800

3 6 9 12 15 A(mV)

A (eV)

y

x

0(eV)

0.6

0.4

0.2

0

|1

P | L

P | L

|0

z

= 225

y

x

|1

A(eV)

300

30

240

180

(degree)

20

( )

120

10

60

0 400 500 600 700 800

0 400 500 600 700 800

A (eV)

Figure 3 | Arbitrary single-qubit rotations R(h, /) with a controlled LZS interference amplitude. (a) The probability of the qubit in the |LS state as a function of the energy position e0 and the driven pulse amplitude A. (b) Charge state probability P|LS along the line in Fig. 3a. The reconstructed evolution of the Bloch vector for two interference nodes is also labelled. (c,d) Total rotation angles y around the x axis and f around the z axis (dots) of the Bloch sphere were extracted from the data shown in Fig. 3b, as a function of the controlled pulse amplitude. The error bars are determined from the least-squares t to the data. The solid lines are theoretical predictions.

0 0.6

0.3 150 200 t p(ps)

250

1

0.8

2

InP FT

1

0

5

P | L

8

4

0

4

1

0

1 5 0 k (ns)

0.6

0.4

0.2

0

0(eV)

400

InP FT

800

kA (ns)

5

1

0

k (ns)

100 200 300 400 tp(ps)

Figure 4 | Rotation dynamics of the LZS interference. The occupation of the qubit in the |LS state as a function of the energy position e0 and pulse length tp. The pulse is shaped using a 2.7-GHz low-pass lter. The inset shows a line cut revealing clear, ultrafast oscillations as a function of time.

5

Figure 5 | Coherence information displayed in the amplitude spectroscopy of the LZS interference. Discrete Fourier transform of the experimental spectroscopy triangle in Fig. 3a. The three-dimentional plot of the Fourier intensity was used to extract the intrinsic dephasing time T2 and the ensemble averaging T2* from the exponential decay (the red line in the inset is a t, see the main text). The reciprocal-space variables kA and ke correspond to the real-space variables A and e, respectively.

leads to an attenuation in the ke direction as follows25:

PFTkA; ke / exp ke/T2 k2e/2T 22:

Therefore, both the intrinsic dephasing time T2 and the inhomogeneous broadening T2* can be extracted from the overall amplitude decay. One must emphasize that this amplitude spectroscopy has an advantage over the two types of dephasing times because T2 and T2* exhibit different ke dependences.

In our data, the Fourier intensity, three-dimensionally plotted in a log scale, is apparently dominated by a linear ke dependence.

Therefore, T2 is extracted without requiring spin-echo experiments. A typical trace form Fig. 5 yields an estimation of T2 40.6 ns, while T2* is difcult to extract as the quadratic

term is relatively small and is masked by noise. Nevertheless, the T2 decoherence time is much longer than the 10 ps required for a

NATURE COMMUNICATIONS | 4:1401 | DOI: 10.1038/ncomms2412 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

& 2013 Macmillan Publishers Limited. All rights reserved.

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms2412

2p phase rotation. The decoherence and relaxation times can greatly effect the single-qubit operation delity as shown in Supplementary Fig. S3.

DiscussionIn summary, we have used a shaped electrical pulse to create LZS interference in a semiconductor DQD charge qubit. The LZS interference-induced x-rotation operations along with the dynamic phase-gate operations can form a basis for rapid, universal, all-electric one-qubit gate operations in a few tens of picoseconds. These results represent progress towards the implementation of semiconductor quantum dot-based qubits. Our results are an important addition to the rapidly growing toolbox of quantum information processing because they are generally applicable to systems that avoid crossings, including both articial and natural two-level systems. This well-controlled, solid-state system could also be seen as an analogue quantum simulator of a real atom undergoing LZS interferometry26.

Methods

Devices. The DQD device is dened by electron beam lithography on a molecular beam epitaxially grown GaAs/AlGaAs heterostructure. The two-dimensional electron gas is located 95 nm below the surface. The two-dimensional electron gas has a density of 3.2 1011 cm 2 and a mobility of 1.5 105 cm2 V 1 s 1.

Figure 1a provides the scanning electron micrograph of the surface gates. Six gates A1, A2, A3, B1, B2 and B3shape the DQD. Gates B4 and B2 form a QPC charge-sensing channel for counting the DQD electron occupations via capacitive coupling.

Control and measurements. The experiments were performed in an Oxford Triton dilution refrigerator with a base temperature of 30 mK. An Agilent 81134A pulse generator with a time resolution of 1 ps was used to deliver fast pulse trains through semi-rigid coaxial transmission lines to the A3 side gate of the device. The conductance through the QPC, GQPC, depends on the change in local charge conguration and provides a sensitive metre for the number of electrons in the left and right dots. The charge state probability is determined by normalizing the charge sensor conductance to the adjacent plateaus in the charge-stability diagram. This measurement technique has been reportedly used in single-charge qubits and offers the experimental convenience of integrating initialization, manipulation and measurement in the same pulse7. In our experiment, a pulse repetition rate of30 MHz was chosen to ensure that the qubit is relaxed to the initial state and to carry out a sufcient number of projective measurements (B107 times) for an adequate signal-to-noise ratio. The ensemble averaging of these measurements, in terms of the average charge detector conductance, allows us to directly obtain the probability of the qubit states.

References

1. Buluta, I., Ashhab, S. & Nori, F. Natural and articial atoms for quantum computation. Rep. Progr. Phys. 74, 104401 (2011).

2. Briegel, H. J., Browne, D. E., Dr, W., Raussendorf, R. & Van den Nest, M. Measurement-based quantum computation. Nat. Phys. 5, 1926 (2009).

3. Gottesman, D. & Chuang, I. L. Demonstrating the viability of universal quantum computation using teleportation and single-qubit operations. Nature 402, 390393 (1999).

4. Petersson, K. D., Petta, J. R., Lu, H. & Gossard, A. C. Quantum coherence in a one-electron semiconductor charge qubit. Phys. Rev. Lett. 105, 246804 (2010).

5. Hayashi, T., Fujisawa, T., Cheong, H. D., Jeong, Y. H. & Hirayama, Y. Coherent manipulation of electronic states in a double quantum dot. Phys. Rev. Lett. 91, 226804 (2003).

6. Hanson, R., Kouwenhoven, L. P., Petta, J. R., Tarucha, S. & Vandersypen, L. M. K. Spins in few-electron quantum dots. Rev. Mod. Phys. 79, 12171265 (2007).

7. Nowack, K. C., Koppens, F. H. L., Nazarov, Y. V. & Vandersypen, L. M. K. Coherent control of a single electron spin with electric elds. Science 318, 14301433 (2007).

8. Petta, J. R., Lu, H. & Gossard, A. C. A coherent beam splitter for electronic spin states. Science 327, 669672 (2010).

9. Gaudreau, L. et al. Coherent control of three-spin states in a triple quantum dot. Nat. Phys. 8, 5458 (2012).

10. Berezovsky, J., Mikkelsen, M. H., Stoltz, N. G., Coldren, L. A. & Awschalom, D. D. Picosecond coherent optical manipulation of a single electron spin in a quantum dot. Science 320, 349352 (2008).

11. Press, D., Ladd, T. D., Zhang, B. & Yamamoto, Y. Complete quantum control of a single quantum dot spin using ultrafast optical pulses. Nature 456, 218221 (2008).

12. De Greve, K. et al. Ultrafast coherent control and suppressed nuclear feedback of a single quantum dot hole qubit. Nat. Phys. 7, 872878 (2011).

13. Shevchenko, S. N., Ashhab, S. & Nori, F. Landau-Zener-Stckerlberg interferometry. Phys. Rep. 492, 130 (2010).

14. Oliver, W. D. et al. Mach-Zehnder interferometry in a strongly driven superconducting qubit. Science 310, 16531657 (2005).

15. Hicke, C., Santos, L. F. & Dykman, M. I. Fault-tolerant Landau-Zener quantum gates. Phys. Rev. A 73, 012342 (2006).

16. Bason, M. G. et al. High-delity quantum driving. Nat. Phys. 8, 147152 (2012).

17. Nalbach, P., Knorzer, J. & Ludwig, S. Nonequilibrium Landau-Zener-Stckerlberg spectroscopy in a double quantum dot. Preprint at http://arXiv.org/abs/1209.2040

Web End =http:// http://arXiv.org/abs/1209.2040

Web End =arXiv.org/abs/1209.2040 (2012).

18. Stehlik, J. et al. Landau-Zener-Stckerlberg interferometry of a single electron charge qubit. Phys. Rev. B 86, 121303(R) (2012).

19. Dupont-Ferrier, E. et al. Coupling and coherent electrical control of two dopants in a silicon nanowire. Preprint at http://arXiv.org/abs/1207.1884

Web End =http://arXiv.org/abs/1207.1884 (2012).

20. Shinkai, G., Hayashi, T., Ota, T. & Fujisawa, T. Correlated coherent oscillations in coupled semiconductor charge qubits. Phys. Rev. Lett. 103, 056802 (2009).

21. Sun, G. et al. Landau-Zener-Stckelberg interference of microwave-dressed states of a superconducting phase qubit. Phys. Rev. B 83, 180507 (2011).

22. Saito, K. & Kayanuma, Y. Nonadiabatic electron manipulation in quantum dot arrays. Phys. Rev. B 70, 201304(R) (2004).

23. Ribeiro, H., Petta, J. R. & Burkard, G. Harnessing the GaAs quantum dot nuclear spin bath for quantum control. Phys. Rev. B 82, 115445 (2010).

24. Berns, D. M. et al. Amplitude spectroscopy of a solid-state articial atom. Nature 455, 5157 (2008).

25. Rudner, M. S. et al. Quantum phase tomography of a strongly driven qubit. Phys. Rev. Lett. 101, 190502 (2008).

26. Buluta, I. & Nori, F. Quantum simulators. Science 326, 108111 (2009).

Acknowledgements

This work was supported by the National Fundamental Research Programme (Grant No. 2011CBA00200), and National Natural Science Foundation (Grant Nos. 11222438, 10934006, 11274294, 11074243, 11174267 and 91121014).

Author contributions

G.C., H.O.L., C.Z., M.X. and H.W.J. fabricated the samples and performed the measurements. T.T., L.W. and G.C.G. designed the experiment, provided theoretical support and analysed the data. G.P.G. and H.W.J. supervised the project. All authors contributed to write the paper.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Cao, G. et al. Ultrafast universal quantum control of a quantum-dot charge qubit using LandauZenerStckelberg interference. Nat. Commun. 4:1401 doi: 10.1038/ncomms2412 (2013).

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/

Web End =http://creativecommons.org/licenses/by-nc-sa/3.0/

6 NATURE COMMUNICATIONS | 4:1401 | DOI: 10.1038/ncomms2412 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

& 2013 Macmillan Publishers Limited. All rights reserved.