Open Access

Tunable optical Kerr effects of DNAs coupled to quantum dots

Nanoscale Research Letters20127:660

DOI: 10.1186/1556-276X-7-660

Received: 15 October 2012

Accepted: 12 November 2012

Published: 29 November 2012


The coupling between DNA molecules and quantum dots can result in impressive nonlinear optical properties. In this paper, we theoretically demonstrate the significant enhancement of Kerr coefficient of signal light using optical pump-probe technique when the pump-exciton detuning is zero, and the probe-exciton detuning is adjusted properly to the frequency of DNA vibration mode. The magnitude of optical Kerr coefficient can be tuned by modifying the intensity of the pump beam. It is shown clearly that this phenomenon cannot occur without the DNA-quantum dot coupling. The present research will lead us to know more about the anomalous nonlinear optical behaviors in the hybrid DNA-quantum dot systems, which may have potential applications in the fields such as DNA detection.


Kerr effects DNA-quantum dot system Optoelectronic material


Biomaterials are now drawing more and more attention since they often present special properties which are not easily obtained from traditional inorganic or organic materials. In addition, biomaterials come from renewable resources and are usually biodegradable. Among biomaterials, researches have been interested in DNA for various reasons, such as potential applications of DNA assembly in molecular electronic devices [1], nanoscale robotics [2], and DNA-based computation [3]. One of the most interesting applications in DNA is to use DNA as a kind of optoelectronic material. Thin film of DNA-CTMA has been used successfully in various applications such as organic light emitting diodes, a cladding and host material in nonlinear optical devices, and organic field-effect transistors because of its nature of large dielectric constant and large band gap [4]. DNA-based polymers are utilized in optically pumped organic solid-state lasers [5]. A better understanding of the nonlinear optical properties of DNA materials will undoubtedly lead us to more exciting applications. So, many researches on nonlinear optical properties of DNA materials have been undertaken. Samoc et al. have studied the nonlinear refractive index and the two-photon absorption coefficient of native (sodium salt) DNA [6]. Second harmonic generation of DNA assemblies in the form of DNA-CTMA has been characterized both theoretically and experimentally by Wanapun et al. [7]. Krupka et al. investigated the third-order nonlinear optical properties of thin films of DNA-based complexes with optical third harmonic generation technique [8]. Nonlinear optical properties of different materials based on DNA are under investigation currently.

In this paper, we theoretically propose and analyze some nonlinear optical properties in a DNA-quantum dot coupling system, which have remained unexplored to date. We investigate DNA molecules coupled to the peptide quantum dot with the optical pump-probe technique. This technique has been realized by several groups [913], which shows the probability for experimental realization. Since photodetection technology is well developed, for instance with the assistance from quantum dot [14], we can expect to observe some properties of DNA molecules by detecting the weak probe beam. However, toxicity should always be cared about when DNA molecules are used together with nanomaterials as has been tested in [15], so a problem we need to pay attention to is that the metallic quantum dots used in biological assays are always toxic. Recently, Amdursky et al. [16, 17] have shown that the peptide quantum dot is nontoxic to the environment and biological tissues. This kind of quantum dot is a good choice of new labeling materials in biological and biomedical experiments. Most recently, the coherent optical spectrum in such a quantum dot-DNA system has been studied by Li and Zhu [18].

In the system, the vibration mode of DNA molecules makes a great contribution to this coupled system so that the optical Kerr effect can be enhanced significantly. This optical Kerr effect can also be switched by adjusting the intensity of the pump beam while leaving the other parameters unchanged. In view of these novel properties, we propose a method to measure the frequency of the vibration mode of DNA molecules.


To understand our system, we consider one of the large amount of DNA-quantum dots (DNA-QDs) in actual reagent as shown in Figure 1. The DNA-QD system is driven by a strong pump field and a weak probe field. A two-level system (the ground state |g >  and the excited state |ex > ) can be chosen as the model for quantum dot, which are dressed by the DNA vibration mode as shown in Figure 1. This two-level system can be described with the pseudo-spin 1 and the corresponding operators are σ+ , σ and σ z . The Hamiltonian of quantum dot can be written as H QD = ω eg σ z , where ωeg = ωex − ω g is the exciton frequency of quantum dot.
Figure 1

DNA and peptide quantum dot coupling system. A peptide QD coupled to DNA molecules in the simultaneous presence of two optical fields. The energy levels of QD when dressing the vibrational modes of DNA molecules are also shown in this figure.

The DNA molecules in our study are modeled as harmonic oscillators [19] and almost have no difference between them, which is not difficult to realize with biological techniques. The Hamiltonian of DNA molecules is
H D = j = 1 n p j 2 2 m j + 1 2 m j ω j 2 q j 2 ,

where the commutation relation [ q j , p j ] = iℏ is satis fied [20].

The damping of the longitudinal vibration mode of the DNA molecules is fairly small in a small volume of aqueous solution [20], though the DNA vibrational modes decay quickly. Therefore, in small volume of aqueous solution, the only vibration mode we care about is the longitudinal vibration mode. In addition, flexion of DNA molecules will result in extensions and compressions of the model, which will finally lead to the modification of the quantum dot levels through the longitudinal strain [21, 22]. The Hamiltonian caused by the coupling of DNA molecules and a quantum dot has the form as follows:
H QD-DNA = σ z j = 1 n κ j q j ,

where κ j is the coupling strength between quantum dot and the j th DNA molecule, and the quantum dot is coupled to n DNA molecules. Because of the diluted aqueous solution of DNA molecules, we do not take the coupling between DNA molecules into consideration [20].

The coupling between QD and optical fields is
H QD f = μ [ E p σ + exp ( i ω p t ) + E p σ exp ( i ω p t ) ] μ [ E s σ + exp ( i ω s t ) + E s σ exp ( i ω s t ) ] ,

where μ is the electric dipole moment of the exciton and Ep(Es) and ωp(ωs) are the amplitude and frequency of the pump-probe field, respectively.

Now we get the Hamiltonian of the QD-DNA system,
H = H QD + H D + H QD-DNA + H QD f = ω eg σ z + i = 1 n p i 2 2 m i + 1 2 m i ω i 2 q i 2 + σ z i = 1 n κ i q i μ [ E p σ + exp ( i ω p t ) + E p σ exp ( i ω p t ) ] μ [ E s σ + exp ( i ω s t ) + E s σ exp ( i ω s t ) ] .
In the rotating frame at ω p , the Hamiltonian becomes
H = Δ p σ z + i = 1 n p i 2 2 m i + 1 2 m i ω i 2 q i 2 ( Ω p σ + + Ω p σ ) + ℏϑ σ z μ [ E s σ + exp ( iδt ) + E s σ exp ( iδt ) ] ,

where Δ p  = ωegωp, ϑ = j = 1 n κ j q j , Ω p = μ E p / is the Rabi frequency and δ = ωs − ωcis the probe-pump detuning.

With this Hamiltonian, we can obtain the equations of motion for σ z , σ, and ϑ via Heisenberg equation and introduce some damping parameters such as Γ1, Γ2 and τD[23]. Γ1 is the exciton relaxation rate and Γ2 is the dephasing rate. τD is the vibrational lifetime of DNA. By introducing the corresponding damping and noise terms [24, 25], the equations are as follows:
d σ z dt = Γ 1 ( σ z + 1 ) + i Ω p ( σ + σ ) + E s exp ( iδt ) σ + E s exp ( iδt ) σ ,
d σ dt = ( i Δ p + + Γ 2 ) σ 2 i Ω p σ z 2 E s exp ( iδt ) σ z + F n ,
d 2 ϑ d t 2 + τ D dt + ω D 2 ϑ = λ ω D 2 σ z + ξ n ,
where λ = j = 1 n κ j 2 m j ω D 2 is the coupling strength of DNA molecules and quantum dot. ωD is the frequency of DNA longitudinal vibrational modes. The δ-correlated Langevin noise operator F n represents the coupling between ϑ and σ, the main cause of the decay of vibration mode. F n has zero mean value < F n  > = 0 and the correlation relation < F n ( t ) F n + ( t ) > δ ( t t ) . The operator ξ n stands for the Brownian stochastic force, since the thermal bath of Brownian and non-Markovian processes will affect the vibration mode of DNA molecules [24, 26]. The quantum effects on the DNA are only observed in the case ωDτD > > 1. The Brownian noise operator can be modeled as Markovian with the decay rate 1/τ D of the vibration mode. Therefore, the Brownian stochastic force has zero mean value < ξ n  > = 0 and can be expressed as [26]
< ξ + ( t ) ξ ( t ) > = 1 τ D ω D 1 + coth ( ℏω 2 k B T ) 2 Π ω e ( t t ) dω.
With the standard methods of quantum optics, the steady-state solution of Equations 6, 7, and 8 are expressed as follows when setting all the time derivatives to zero:
σ 0 = 2 i Ω p σ 0 z σ 0 z Γ 2 i Δ p , ϑ 0 = λ σ 0 z ,
where σ0z is determined by Equation 15. To extend this formalism beyond weak coupling, we can always rewrite each Heisenberg operator as the sum of its steady-state mean value and a small fluctuation with zero mean value as follows: σ = σ0 + δ σ, σ z  = σ0z + δ σ z , and ϑ = ϑ0 + δϑ, which should be substituted into Equations 6, 7, and 8. We can neglect the nonlinear term δϑδ σ safely. Since the optical drives are weak and classical, we will identify all the operators with their expectation values and omit the quantum and thermal noise terms [9]. Then the linearized Langevin equations can be written as follows:
< δ σ z ̇ > = i Ω p ( < δ σ > < δ σ > ) Γ 1 < δ σ z > + E s exp ( iδt ) < δ σ > E s exp ( iδt ) < δ σ > ,
< δ σ ̇ > = ( i Δ p + Γ 2 ) < δ σ > 2 i Ω p < δ σ z > i ( ϑ 0 < δ σ > + < δϑ > σ 0 ) 2 E s exp ( iδt ) < δ σ z > ,
< δϑ ̈ > + < δϑ ̇ > τ D + ω D 2 < δϑ > = λ ω D 2 < δ σ z > .
From the approximations < δ σ z > = σ z + exp ( iδt ) + σ z exp ( iδt ) , < δ σ > = σ + exp ( iδt ) + σ exp ( iδt ) and < δϑ > = ϑ+ exp(−iδt) + ϑexp(iδt) [27], we can obtain:
( Γ 1 λ 2 ) σ 0 z 3 + ( Γ 1 λ 2 2 + 2 Γ 1 λ Δ p ) σ 0 z 2 + ( Γ 1 Δ p 2 Γ 1 Γ 2 2 + Γ 1 λ Δ p 4 Γ 2 Ω p 2 ) σ 0 z = 1 2 Γ 1 Γ 2 2 + 1 2 Γ 1 Δ p 2 .
Now we get σ0z. Then σ0and ϑ0 are also known. All of the equations can then be solved completely. We finally obtain the part we are interested in, the equation:
σ = Z ( δ ) σ z ,
where the equations used are G ( δ ) = ( λ ω D 2 ) / ( δ 2 + / τ D ω D 2 ) , Z ( δ ) = ( σ 0 G + 2 Ω p ) / ( i Γ 2 δ Δ p ϑ 0 ) , F(δ) = (σ0G + 2Ω p )/(i Γ2 + δ − Δ p  − ϑ0), σ 0 = 2 i Ω p σ 0 z Γ 2 + i Δ p σ 0 z , ϑ0 = −λ σ0z and finally
σ z + = E s σ 0 ( i Γ 2 + δ Δ p ϑ 0 ) 2 E s Ω p σ 0 z ( iG Ω p i Z Ω p + Γ 1 ) ( i Γ 2 + δ Δ p ϑ 0 ) .
We can use the equations above and σ z = σ z + to obtain the nonlinear optical susceptibility:
χ ( ω s ) eff ( 3 ) = N μ 3 σ 3 ε 0 2 Ω p 2 E s = m χ 3 ( ω s ) ,

where N is the number density of DNA-QDs and m = N μ 4 3 ε 0 3 Γ 2 3 .

Results and discussion

To show the numerical results, we choose the realistic quantum dot-DNA system, in which a peptide quantum dot is coupled to several DNA molecules as illustrated in Figure 1. Although the DNA molecules in solution form can be distorted in mess, one can extend these molecules into linear form with electromagnetic field or fluid force [28]. In addition, the longitudinal vibrational frequency can be affected by the length of DNA molecules, which could just be considered as a factor affecting vibration frequency. In the theoretical calculation, we choose ωD = 40 GHz and τD = 5 ns as the vibration frequency and lifetime of DNA molecules [22, 2931]. For our study, we can safely select the decay rate of the peptide quantum dot as Γ1 = 16 GHz for any practical purpose [32].

Figure 2a plots the optical Kerr coefficient Re χ eff ( 3 ) (solid curve) and nonlinear absorption Im χ eff ( 3 ) (dash curve) as functions of probe-exciton detuning Δ s  = ωs − ωeg with Δ p  = 0 and λ = 0, while Figure 2b shows optical Kerr coefficient Re χ eff ( 3 ) (solid curve) and nonlinear absorption Im χ eff ( 3 ) (dash curve) as functions of probe-exciton detuning Δs = ωs − ωegwith Δ p  = 0 but λ = 2 GHz. It demonstrates that if we fix the pump beam on-resonance with the exciton frequency and scan the probe beam, we can obtain the large enhanced optical Kerr effect at ωs = ωeg − ωD and ωs = ωeg + ωD. The origin of this phenomenon is the quantum interference between the vibration mode of DNA molecules and the beat of the two optical fields via the exciton when probe-pump detuning δ is adjusted equal to the frequency of the vibration mode of DNA molecules. If we ignore the coupling, λ = 0, the enhancement of optical Kerr effect will disappear completely as has been demonstrated in Figure 2a. Therefore, the importance of the coupling between quantum dot and DNA molecules is obvious since the enhancement of optical Kerr effect could not occur in such a system when λ = 0. Furthermore, we can propose a scheme to measure the frequency of the vibration mode of DNA molecules by making use of the phenomenon above. From Figure 2b, we can clearly see that as the frequency of the vibration mode is ωD = 40 GHz, the two sharp peaks at ±40 GHz just match the mode frequency. This means that if we first adjust pump beam properly and scan the probe frequency across the exciton frequency in the spectrum, we can easily obtain the accurate vibration frequency of DNA, which implies some future potential applications.
Figure 2

The optical Kerr coefficient of probe beam (in units of m ) with pump beam on-resonance. (a) The optical Kerr coefficient and nonlinear absorption as functions of probe-exciton detuning Δ s in the case λ=0. (b) The optical Kerr coefficient and nonlinear absorption as functions of probe-exciton detuning Δ s in the case λ = 2 GHz.

To explore the phenomenon above more carefully, we show the optical Kerr coefficient Re χ eff ( 3 ) as functions of probe-exciton detuning Δ s  = ω sω egwith Δ p  = 0 and different coupling strengths and vibration lifetimes in Figure 3a,b respectively. In Figure 3a, we see that the larger the coupling strength is, the higher the optical Kerr coefficient peak will be. Figure 4a shows that the optical Kerr coefficient peak increases monotonously with vibration lifetime τD. We should not feel surprised about these results. Since the optical Kerr coefficient peak is caused by the coupling between DNA molecules and quantum dot, the peak will become more and more obvious when the coupling makes stronger. These results demonstrate that the coupling plays a key role in such a coupled system.
Figure 3

Optical Kerr coefficient of probe beam with various coupling strengths and vibration lifetimes. (a) The optical Kerr coefficient (in units of m ) as functions of probe-exciton detuning Δ s with pump beam on-resonance (Δ p  = 0) and different coupling strengths. (b) The optical Kerr coefficient (in units of m ) as functions of probe-exciton detuning Δ s with pump beam on-resonance (Δ p  = 0) and different vibration lifetimes.
Figure 4

Optical Kerr coefficient of probe beam with various Rabi frequencies of pump beam. (a) The optical Kerr coefficient (in units of m ) as functions of probe-exciton detuning Δ s with pump beam off-resonance(Δ p  = ωD) and different Rabi frequencies. (b) The detailed part of (a) showed the relation between optical Kerr coefficient and Rabi frequency>

Figure 4a presents optical Kerr effects as functions of Δ s with Δ p  = ωD and different Rabi frequencies of the pump field, whose detail is shown in Figure 4a. We first notice that the probe beam experiences different optical Kerr coefficients when appearing in the pump beams with different intensities. When we pay attention to the detail (shown in Figure 4a), we find that by increasing the intensity of the pump beam, the optical Kerr effect will be weakened significantly. Therefore, we can see that the magnitude of optical Kerr effect can be tuned by controlling the light intensity, implying a method for regulating the nonlinear optical features of DNAs via coupling to quantum dots.


In conclusion, we have proposed a theoretical model for DNA-quantum dot hybrid system in the presence of a strong pump laser and a weak probe laser. The coupling leads to the great enhancement of probe beam Kerr coefficient at two off-resonant points, which may be of potential use in frequency measurement. Furthermore, the relation between the optical Kerr coefficient of the probe beam and intensity of the pump beam may be utilized to control the strength of optical nonlinearity of the system. We believe that such a phenomenon may lead to a more profound understanding of nonlinear optical properties of the hybrid quantum dot-DNA system. We expect our consequences can be checked experimentally in the near future.



This work was supported by the National Natural Science Foundation of China (numbers 10974133 and 11274230) and the Ministry of Education Program for Ph.D.

Authors’ Affiliations

Key Laboratory of Artificial Structures and Quantum Control (Ministry of Education), Department of Physics, Shanghai Jiao Tong University


  1. Robinson BH, Seeman NC: The design of a biochip: a self-assembling molecular-scale memory device. Protein Eng 1987, 1: 295–300. 10.1093/protein/1.4.295View ArticleGoogle Scholar
  2. Yan H, Zhang XP, Shen ZY, Seeman NC: A robust DNA mechanical device controlled by hybridization topology. Nature 2002, 415: 62–65. 10.1038/415062aView ArticleGoogle Scholar
  3. Turberfield A: DNA as an engineering material. Phys World 2003, 16: 43–46.View ArticleGoogle Scholar
  4. Birendra ST, Serdar SN, James GG: Bio-organic optoelectronic devices using DNA. Adv Polym Sci 2010, 223: 189–212.Google Scholar
  5. Yu Z, Li W, Hagen JA, Zhou Y, Klotzkin D, Grote JG, Steckl AJ: Photoluminescence and lasing from deoxyribonucleic acid (DNA) thin films doped with sulforhodamine. Appl Opt 2007, 46: 1507–1513. 10.1364/AO.46.001507View ArticleGoogle Scholar
  6. Samoc M, Samoc A, Grote JG: Complex nonlinear refractive index of DNA. Chem Phys Lett 2006, 431: 132–134. 10.1016/j.cplett.2006.09.057View ArticleGoogle Scholar
  7. Wanapun D, Hall VJ, Begue NJ, Grote JG, Simpson GJ: DNA-based polymers as chiral templates for second-order nonlinear optical materials. Chem Phys Chem 2009, 10: 2674–2678. 10.1002/cphc.200900303Google Scholar
  8. Krupka O, Ghayoury AE, Rau I, Sahraoui B, Grote JG, Kajzar F: NLO properties of functionalized DNA thin films. Thin Sol Films 2008, 516: 8932–8936. 10.1016/j.tsf.2007.11.089View ArticleGoogle Scholar
  9. Weis S, Rivieere R, Deleeglise S, Gavartin E, Arcizet O, Schliesser A, Kipperberg TJ: Optomechanically induced transparency. Science 2010, 330: 1520–1523. 10.1126/science.1195596View ArticleGoogle Scholar
  10. Teufel JD, Li D, Allman MS, Cicak K, Sirois AJ, Whittaker JD, Simmonds RW: Circuit cavity electromechanics in the strong-coupling regime. Nature 2011, 471: 204–208. 10.1038/nature09898View ArticleGoogle Scholar
  11. Safavi-Naeini AH, Alegre TPM, Chan J, Eichenfield M, Winger M, Lin Q, Hill JT, Chang DE, Painter O: Electromagnetically induced transparency and slow light with optomechanics. Nature 2011, 472: 69–73. 10.1038/nature09933View ArticleGoogle Scholar
  12. Li JJ, Zhu KD: A scheme for measuring vibrational frequency and coupling strength in a coupled annomechancial resonator-quantum dto system. Appl Phys Lett 2009, 94: 063116. 10.1063/1.3072599View ArticleGoogle Scholar
  13. He W, Li JJ, Zhu KD: Coupling-rate determination based on radiation pressure-induced normal mode splitting in cavity optomechanical systems. Opt Lett 2010, 35: 339–341. 10.1364/OL.35.000339View ArticleGoogle Scholar
  14. Thomas G: Injector quantum dot molecule infrared photodetector: a concept for efficient carrier injection. Nano-Micro Lett 2011, 3: 121–128.View ArticleGoogle Scholar
  15. S.Karthick RN, Gnanendra KE, Reepika R: Synthesis of silver nanoparticles by Lactobaciluus acidophilus 01 strain and evaluation of its in vitro genomic DNA toxicity. Nano-Micro Lett 2010, 2: 160–163.View ArticleGoogle Scholar
  16. Amdursky N, Molotskii M, Gazit E, Rosenman G: Self-assembled bioinspired quantum dots: optical properties. Appl Phys Lett 2009, 94: 261907. 10.1063/1.3167354View ArticleGoogle Scholar
  17. Amdursky N, Molotskii M, Gazit E, Rosenman G: Elementary building blocks of self-assembled peptide nanotubes. J Am Chem Soc 2010, 132: 15632–15636. 10.1021/ja104373eView ArticleGoogle Scholar
  18. Li JJ, Zhu KD: Coherent optical spectroscopy in a biological semiconductor quantum dot-DNA hybrid system. Nano Res Lett 2012, 7: 1–7. 10.1186/1556-276X-7-1View ArticleGoogle Scholar
  19. Van Zandt LL: Resonant microwave absorption by dissolved DNA. Phys Rev Lett 1986, 57: 2085–2087. 10.1103/PhysRevLett.57.2085View ArticleGoogle Scholar
  20. Dorfman BH: The effects of viscous water on the normal mode vibrations of DNA. Dissert Abstr Int 1984, 45: 2213.Google Scholar
  21. Edwards GS, Davis CC, Saffer JD, Swicord ML: Resonant microwave absorption of selected DNA molecules. Phys Rev Lett 1984, 53: 1284–1287. 10.1103/PhysRevLett.53.1284View ArticleGoogle Scholar
  22. Edwards GS, Davis CC, Saffer JD, Swicord ML: Microwave-field-driven acoustic modes in DNA. Biophys J 1985, 47: 799–807. 10.1016/S0006-3495(85)83984-9View ArticleGoogle Scholar
  23. Donega CM, Bode M, Meijerink A: Size- and temperature-dependence of exciton lifetimes in CdSe quantum dots. Phys Rev B 2006, 74: 085320.View ArticleGoogle Scholar
  24. Gardiner CW, Zoller P: Quantum kinetic theory. V. Quantum kinetic master equation for mutual interaction of condensate and noncondensate. Phys Rev A 2000, 61: 033601/1–26.Google Scholar
  25. Milburn GJ, Jacobs K, Walls DF: Quantum-limited measurements with the atomic force microscope. Phy Rev A 1994, 50: 5256–5263. 10.1103/PhysRevA.50.5256View ArticleGoogle Scholar
  26. Giovannetti V, Vitali D: Phase-noise measurement in a cavity with a movable mirror undergoing quantum Brownian motion. Phy Rev A 2001, 63: 023812.View ArticleGoogle Scholar
  27. Boyd RW: Nonlinear Optics. Amsterdam: Academic Press; 2008.Google Scholar
  28. Marko JF, Siggia ED: Stretching DNA. Macromolecules 1995, 28: 8759–8770. 10.1021/ma00130a008View ArticleGoogle Scholar
  29. Yuan CL, Chen HM, Lou XW, Archer LA: DNA bending stiffness on small length scales. Phys Rev Lett 2008, 100: 018102.View ArticleGoogle Scholar
  30. Gill R, Willner I, Shweky I, Banin U: Fluorescence resonance energy transfer in CdSe/ZnS-DNA conjugates: probing hybridization and DNA cleavage. J Phys Chem B 2005, 109: 23715–23719. 10.1021/jp054874pView ArticleGoogle Scholar
  31. Adai BK: Vibrational resonances in biological systems at microwave. Biophys J 2002, 82: 1147–1152. 10.1016/S0006-3495(02)75473-8View ArticleGoogle Scholar
  32. Tsay MJ, Trzoss M, Shi LX, Kong XX, Selke M, Jung EM, Weiss S: Singlet oxygen production by peptide-coated quantum dot-photosensitizer conjugates. Am Chem Soc 2007, 129: 6865–6871. 10.1021/ja070713iView ArticleGoogle Scholar


© Li and Zhu; licensee Springer. 2012

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.