## Abstract

We discuss thin optical structures that allow chromatic aberrations to be avoided in the THz domain. The paper contains the theoretical considerations, computer modeling and experimental evaluation of the high order kinoform diffractive elements in the THz range. According to the obtained results application of the high order kinoforms enables broadband operation in the THz range.

© 2014 Optical Society of America

## 1. Motivation

Active and passive systems working in the terahertz (THz) radiation range require optical elements, like lenses or mirrors, which efficiently collect energy and concentrate properly the incoming radiation. This is compulsory as most of the THz sources and detectors exhibit a rather low efficiency [1–4]. Thus a sufficient THz energy, exceeding the noise level, must reach a single element detector or each pixel of a matrix detector. Therefore, the optical system transmission must be optimized. Simultaneously, the geometrical aberrations of the optical system must be low in order to both concentrate the energy at the focal plane and produce sharp images. Typically, the size of pixels in available THz matrix detectors is of the order of the wavelength, or even sub-wavelength in sensors derived from the infrared technology (in that case, the delivered signal corresponds to an average over several pixels, which allows noise processing) [5–8]. Therefore, the optics must operate at the limit of diffraction. In practice, the aperture of imaging or focusing optical component is often large and its diameter usually spreads from 50 mm up to 300 mm or even more. To reach the diffractive resolution limit, the focal length of such components is often equal to their diameter. Moreover, chromatic aberrations must also be reduced in many practical cases. For example, a chemical identification of the observed items, in view of security applications, may be performed by recording the images at several different wavelengths, for which the items show a differentiated THz absorption [9, 10]. Another relevant example is given by pulsed THz systems, like time-domain spectroscopy set-ups, which deliver broadband spectra spreading over several frequency decades [11–14].

Common THz optical components include refractive dielectric devices and metallic mirrors. Metallic mirrors are known to be achromatic and to show a very high efficiency, as actual metals in the THz range behave almost as ideal conductors [15–18]. However, reflecting and focusing mirrors are very sensitive to any misalignment and they do not work properly off axis. Thus systems based on focusing mirrors show high geometrical aberrations, larger than systems based on refractive lenses. Moreover, large numerical aperture mirrors are heavy and have significant sizes. On the other hand, dielectric lenses are easy and cheap to manufacture, especially when made from common organic plastics like high density polyethylene (HDPE) or polytetrafluoroethylene (PTFE). They are approximately achromatic because the refractive index *n* of these materials is almost constant over the whole THz range. Unfortunately, devices with the requested f-number (*N* ≈1) and focal lengths (100~300 mm) are tremendously thick. For example, a plano-convex lens with 300 mm diameter and *N* = 1 made from HDPE (*n* ≈1.54) is 100 mm thick! Not only such lenses are heavy, but they do not well transmit high frequency signals, as the absorption of polymers is rather substantial over 1 THz. In addition, we have already demonstrated that strong absorption of the medium boosts aberrations of bulky lenses [19]. In the THz range of radiation the first order kinoform can be used [20, 21], but they suffer from chromatic aberration. A good improvement of the bulky design is given by the so-called high order kinoform (HOK) scheme [22, 23]. HOK lenses consist of zones with a thicknesses being of the order of the THz wavelength, while classical Fresnel lenses have much thicker zones (in comparison with the wavelength) and can be described by a ray-tracing approach. By a precise control of the phase shift introduced by each zone, a constructive interference of all the waves diffracted by the HOK structure occurs only at the focal spot, while destructive interferences make the signal dropping to zero elsewhere in the focal plane. Thus, as compared to Fresnel lenses, HOK show much better optical performance.

We already proposed HOK [19, 24] as a relatively thin diffractive optical element for THz beams. Such an element is lightweight and offers lower attenuation due to its smaller thickness. The goal of the present paper is to give a theoretical outline of the HOK response with significantly suppressed chromatic aberration. Numerical simulations and experiments illustrate focusing in the broadband 300-600 GHz. This whole study demonstrates that thin HOK refractive elements help at nicely manipulating THz beams and moreover exhibit fairly good achromatic properties over the whole THz spectrum.

## 2. Theoretical background

First, we describe the difference between few lenses to facilitate the understanding of the
optical response of the HOK element, also discussed in [25]. A classical refractive lens (here a plano-convex) is a thick element [Fig. 1(a)]. Thanks to the original idea by Fresnel, its thickness can be strongly reduced by removing
circular layers whose thicknesses correspond to phase differences equal to multiples of
2*π*. In Fresnel lenses, the radial width of such zones is larger than
1000 times the wavelength. Kinoforms [26] [Fig. 1(b)] are improved devices based on a similar scheme. The
radial width of each zone if of the order of the wavelength and the phase shifts introduced by
all the zones are precisely controlled in order to achieve constructive interference at the
focal spot. In a classical kinoform (1st order kinoform), the phase jump between two adjacent
zones is 2*π*, resulting in a very thin and thus highly transparent
structure. However, because of the single 2*π* phase jump, 1st order
kinoform reveals very strong wavelength dependence. This inconvenience is minimized by designing
HOK [22, 23]
[Fig. 1(c)], whose phase jump is
2*pπ*, with the order *p* >> 1. The HOK element
remains still a thin component, leading to a low attenuation of the illuminating beam as
required for THz applications.

Let us briefly discuss the chromatic properties of the HOK in a case of a broadband illuminating radiation. We consider a HOK that is designed for the *p*^{th} order at wavelength *λ* and we study its chromatic aberrations, i.e. its focusing properties for a smaller wavelength *λ*’ of the spectrum. Generally, the phase *φ* of a kinoform can be arbitrary chosen to design the desired optical properties. Here, for a better clarity, we concentrate on a HOK focusing lens [23, 25], with the well-known phase function of a thin converging lens in the paraxial approximation:

*r*denotes the radial coordinate and

*f*is the focal length of the lens. In the THz range, it is reasonable to assume a constant refractive index

*n*of the kinoform medium. The local phase shift induced by the propagation through the lens material of thickness

*h*is equal to:Then, in a case of the illumination by the smaller wavelength

*λ*’ (

*λ*’<

*λ*), the phase shift is $\Delta \phi \text{'}=\frac{\lambda}{\lambda \text{'}}\Delta \phi $. The transmitted wave of wavelength

*λ*is multiplied by the phase transmittance of the lens $T\left(\phi \right)=\mathrm{exp}\left(i\phi \right)$. Thus, for a different wavelength

*λ*’, the transmittance converts into the form ${T}_{1}\left(\phi \right)=\mathrm{exp}\left(i\phi \frac{\lambda}{\lambda \text{'}}\right)$ being a function of

*φ*. We design the HOK in such a way that phase returns to zero each time when

*φ*exceeds $d=2p\text{\hspace{0.17em}}\pi $. Then the function

*T*

_{1}(

*φ*) defines the HOK transmittance within the period $d=2p\text{\hspace{0.17em}}\pi $and we can expand this transmittance for a wavelength

*λ*’ into the following Fourier series:

*m*denotes the diffractive order with the amplitude given as follows:

*m*-th order, i.e. the square of the modulus of the Fourier coefficient, is:

*m*fulfilling the following condition:where $\xi \in \left[0,1\right]$ and $m\ge p$ according to the assumption $\lambda \text{'}$<

*λ*. This leads to:As $\xi \in \left[0,1\right]$, weaker orders contribute less to the output wave field. The phase ${\varphi}_{m}$ at wavelength

*λ*’ of the

*m*-th order is deduced from (3):

*φ*the value of

_{m}*φ*given by (1) for the wavelength

*λ*. The focal length $f\text{'}$ for the wavelength $\lambda \text{'}$ obeys the same expression (1) but the phase is defined by

*φ*. Finally, we obtain:

_{m}*λ*and

*λ*’.

Although we have only discussed the case of a converging lens, the same conclusion can be drawn for an arbitrary phase function *φ* that is inversely proportional to the wavelength *λ*. The above method can be applied to design a large variety of HOK components as for example spherical, cylindrical, elliptical, hyperbolical, conical lenses, or axicons, or prisms. Relation (8) indicates that the geometry of the reconstructed wave front is almost independent of the wavelength for a sufficient large *p*. Hence chromatic properties of such HOKs can be substantially avoided.

## 3. High order kinoform lens for the THz range – numerical modeling

The numerical modeling is performed using a non-paraxial version of the modified convolution method [27, 28] combined with a non-paraxial modification of the beam propagation algorithm [29]. The volume element is divided into slices, each of them is 50 µm wide. The studied HOK structure exhibits a diameter *d* = 300 mm and the desired THz focal distance is *f* = 300 mm. The HOK is supposed to be made from HDPE, whose refractive index is *n* = 1.54 and the attenuation coefficient is *α* = 0.23 cm^{−1} at 0.3 THz (*λ* = 1 mm). The HOK is designed to be optimized for the *p* = 8 order at *λ* = 1 mm. We perform the modeling of the HOK focusing response in the wavelength range from 0.5 mm to 1 mm.

Generally Eq. (6) can be rewritten in the following
way: *λ*’ = *λp*/*m*’,
where *m*’ is a number equal to *m* ± ξ/2.
According to Eq. (5), the designed HOK lens shows
one main order of diffraction for the wavelengths corresponding to the integer
*m*’ (ξ = 0). Table 1
gives the *λ*’s in the range 0.5-1 mm, i.e. for
*m*’ varying from 8 to 16. For these wavelengths, the focal spots are diffraction limited (the spot diameter
is almost equal to the wavelength) and, as the lens is achromatic, they lie in the same focal
plane. Figure 2 shows the computed images of the beam in
the focal plane for different *λ*’, i.e. when
*m*’ varies from 8 to 16. For integer values of *m*’, the light beam is properly focused down
to the limit of diffraction. The focal spot size increases with the wavelength as expected from
the diffraction law. According to the description given in the former section, a non-integer
number *m*’ defining *λ*’ by the equation
*λ*’ = *λp*/*m*’
should lead to a focal shift. This effect is well observed in Fig. 2 where spots corresponding to non-integer *m*’ are
defocused. On the other hand, for practical imagery applications, the chosen order
*p* = 8 is sufficiently high to maintain the concentrated beam within an assumed
detector area, which is a 5 × 5 mm^{2} square. Detectors with this dimensions are
available and were used in our experiments.

## 4. High order kinoform lens for the THz range – experiment

We characterized HDPE HOK lenses that have the same geometrical parameters as the ones used in
the modeling as described above. The experimental setup is presented in Fig. 3. A sub-THz beam is delivered by a Virginia Diodes Inc. (VDI) all-electronic system. The
frequency is tunable in the range 0.14-1 THz. The output of the VDI source is a diagonal horn. A
*f* = 75 mm HDPE lens (diameter 5.08 cm) focuses the beam on a 2.5 mm diameter
pinhole. The pinhole was imaged in the 2*f*-2*f* configuration by
the HOK lens, thus leading to the point spread function (PSF) of the HOK lens in the focal
plane. The image plane was scanned with a movable detector (Zero Bias Detector – VDI
Schottky Diode) attached to a 5 × 5 mm^{2} horn. The image was recorded within a
25 × 25 mm^{2} paraxial region at the focal plane by steps of 5 mm. Due to the
relatively large geometrical size of the detector, a translation of the lens or the detector by
10 mm along the optical axis did not change significantly the readout

The recorded images are presented in Fig. 4. Each image was recorded for a given wavelength *λ*’, which
corresponds to the *m*’ parameter ranging between 8 and 10, and is
normalized due to the different efficiency of the source and detector for different frequencies.
The average transmission of the volume of the HOK (0.3 THz) is 83% of the illumination energy
(taking into account only the attenuation of the lens material). Additionally, this percentage
will be smaller due to the Fresnel losses. Moreover, the attenuation of HDPE changes with
frequency. The spatial resolution of the set-up is enough to allow the recording of the PSF even
when the THz beam is strongly focused. However, the diameter (3 mm) of the recorded central peak
is here limited by the set up resolution, and its actual size is smaller, i.e. certainly close
to the diffraction limit (~1 mm). The recorded images are in a very good agreement with the
computed ones, as it may be seen in Fig. 2. It should be
noticed that experimental results have the sampling (pixel size) 5x5 mm^{2} and
numerical simulations 0,1x0,1 mm^{2}. That is why in the experiment we cannot see big
differences between the consecutive images. The HOK lens was designed to work properly for the
detector having the diameter 5x5 mm^{2}, which means to focus well on this area for all
frequencies. Even when in modelling we can notice some focal shifts for non-integer numbers of
*m*’, in the experiment they are slightly visible due to the averaging of
the detector. The differences between all the recorded PSFs are not significant even if, for the
fractional values of *m*’, they exhibit additional blur (see for example
the cases *m*’ = 8.50, *m*’ = 8.75,
*m*’ = 9.5). This proves that chromatic aberrations are strongly reduced
for this HOK lens, and thus this component can be usefully employed in THz broadband imaging
systems.

## 5. Conclusion

The paper analyzes the usefulness of the transmissive focusing HOK lenses for broadband THz radiation. In agreement with the above theoretical outline, these HOK lenses exhibit negligible chromatic aberrations and good focusing properties in the THz range. Although theory specifies that ideal achromatism is achieved for HOKs running in the large *p* order regime, calculation and experiment show that acceptable effects are obtained even with a relatively small value of *p* (here *p* = 8). This small value permits the fabrication of thin HOK lenses, and thus to reduce the absorption losses in the lens material. In conclusion, such HOK components constitute an interesting alternative to more classical THz optics, with many advantages like improved transparency, low chromatic aberrations and reduced weight.

## Acknowledgments

This work was supported by the Polish Ministry of Science and Higher Education under the Project O N515 020140 and by the Polish National Science Center under grant N N515 498840, with a complementary support from the European Union in the framework of European Social Fund through the Warsaw University of Technology Development Program.

## References and links

**1. **M. Tonouchi, “Cutting-edge terahertz technology,” Nat. Photonics **1**(2), 97–105 (2007). [CrossRef]

**2. **Y.-S. Lee, *Principles of Terahertz Science and Technology* (Springer 2009).

**3. **F. F. Sizov and A. Rogalski, “THz detectors,” Prog. Quantum Electron. **34**(5), 278–347 (2010). [CrossRef]

**4. **E. Bründermann, H.-W. Hübers, and M. Kimmitt, *THz techniques* (Springer 2012).

**5. **N. Oda, “Uncooled bolometer-type terahertz focal plane array and camera for real-time imaging,” C. R. Phys. **11**(7-8), 496–509 (2010). [CrossRef]

**6. **A. Rogalski, “Semiconductor detectors and focal plane arrays for far-infrared imaging,” Opto-Electron. Rev. **21**(4), 406–426 (2013). [CrossRef]

**7. **M. A. Dem’yanenko, D. G. Esaev, I. V. Marchishin, V. N. Ovsyuk, B. I. Fomin, B. A. Knyazev, and V. V. Gerasimov, “Application of uncooled microbolometer detector arrays for recording radiation of the terahertz spectral range,” Optoelectron. Instrum. Data Process. **47**(5), 508–512 (2011). [CrossRef]

**8. **A. W. Lee and Q. Hu, “Real-time, continuous-wave terahertz imaging by use of a microbolometer focal-plane array,” Opt. Lett. **30**(19), 2563–2565 (2005). [CrossRef] [PubMed]

**9. **H. Zhong, A. Redo-Sanchez, and X.-C. Zhang, “Identification and classification of chemicals using terahertz reflective spectroscopic focal-plane imaging system,” Opt. Express **14**(20), 9130–9141 (2006). [CrossRef] [PubMed]

**10. **J. Chen, Y. Chen, H. Zhao, G. J. Bastiaans, and X.-C. Zhang, “Absorption coefficients of selected explosives and related compounds in the range of 0.1-2.8 THz,” Opt. Express **15**(19), 12060–12067 (2007). [CrossRef] [PubMed]

**11. **P. Y. Han and X.-C. Zhang, “Free-space coherent broadband terahertz time-domain spectroscopy,” Meas. Sci. Technol. **12**(11), 1747–1756 (2001). [CrossRef]

**12. **A. D. Burnett, W. Fan, P. C. Upadhya, J. E. Cunningham, M. D. Hargreaves, T. Munshi, H. G. M. Edwards, E. H. Linfield, and A. G. Davies, “Broadband terahertz time-domain spectroscopy of drugs-of-abuse and the use of principal component analysis,” Analyst (Lond.) **134**(8), 1658–1668 (2009). [CrossRef] [PubMed]

**13. **I. Katayama, R. Akai, M. Bito, H. Shimosato, K. Miyamoto, H. Ito, and M. Ashida, “Ultrabroadband terahertz generation using 4-N,N-dimethylamino-4′-N′-methyl-stilbazolium tosylate single crystals,” Appl. Phys. Lett. **97**(2), 021105 (2010). [CrossRef]

**14. **E. Matsubara, M. Nagai, and M. Ashida, “Coherent infrared spectroscopy system from terahertz to near infrared using air plasma produced by 10-fs pulses,” J. Opt. Soc. Am. B. **30**(6), 1627–1630 (2013). [CrossRef]

**15. **M. Nazarov and J.-L. Coutaz, “Terahertz Surface Waves Propagating on Metals with Sub-wavelength Structure and Grating Reliefs,” J. Infrared Millimeter, Terahertz, Waves **32**(10), 1054–1073 (2011). [CrossRef]

**16. **M. Naftaly and R. Dudley, “Terahertz reflectivities of metal-coated mirrors,” Appl. Opt. **50**(19), 3201–3204 (2011). [CrossRef] [PubMed]

**17. **J. Lloyd-Hughes and T.-I. Jeon, “A review of the terahertz conductivity of bulk and nano-materials,” J. Infrared Millimeter, Terahertz, Waves **33**(9), 871–925 (2012). [CrossRef]

**18. **M. A. Ordal, R. J. Bell, R. W. Alexander Jr, L. L. Long, and M. R. Querry, “Optical properties of fourteen metals in the infrared and far infrared: Al, Co, Cu, Au, Fe, Pb, Mo, Ni, Pd, Pt, Ag, Ti, V, and W,” Appl. Opt. **24**(24), 4493–4499 (1985). [CrossRef] [PubMed]

**19. **M. Sypek, J.-L. Coutaz, A. Kolodziejczyk, M. Makowski, and J. Suszek, “Aberrations of the large aperture attenuating THz lenses,” Proc. SPIE **8261**, 826110 (2012). [CrossRef]

**20. **E. D. Walsby, J. Alton, C. Worrall, H. E. Beere, D. A. Ritchie, and D. R. Cumming, “Imprinted diffractive optics for terahertz radiation,” Opt. Lett. **32**(9), 1141–1143 (2007). [CrossRef] [PubMed]

**21. **B. A. Knyazev, V. S. Cherkassky, Y. Y. Choporova, V. V. Gerasimov, M. G. Vlasenko, M. A. Dem’yanenko, and D. G. Esaev, “Real-Time Imaging Using a High-Power Monochromatic Terahertz Source: Comparative Description of Imaging Techniques with Examples of Application,” J. Infrared Millimeter, Terahertz, Waves **32**(10), 1207–1222 (2011). [CrossRef]

**22. **J. C. Marron, D. K. Angell, and A. M. Tai, “Higher-Order Kinoforms,” Proc. SPIE **1211**, 62–66 (1990). [CrossRef]

**23. **L. N. Hazra and C. A. Delsile, “Higher order kinoform lenses: diffraction efficiency and aberrational properties,” Opt. Eng. **36**(5), 1500–15007 (1997). [CrossRef]

**24. **M. Sypek, M. Makowski, E. Hérault, A. Siemion, A. Siemion, J. Suszek, F. Garet, and J. L. Coutaz, “Highly efficient broadband double-sided Fresnel lens for THz range,” Opt. Lett. **37**(12), 2214–2216 (2012). [CrossRef] [PubMed]

**25. **D. W. Sweeney and G. E. Sommargren, “Harmonic diffractive lenses,” Appl. Opt. **34**(14), 2469–2475 (1995). [CrossRef] [PubMed]

**26. **J. A. Jordan Jr, P. M. Hirsch, L. B. Lesem, and D. L. Van Rooy, “Kinoform Lenses,” Appl. Opt. **9**(8), 1883–1887 (1970). [PubMed]

**27. **M. Sypek, “Light propagation in the Fresnel region. New numerical approach,” Opt. Commun. **116**(1-3), 43–48 (1995). [CrossRef]

**28. **M. Sypek, C. Prokopowicz, and M. Górecki, “Image multiplying and high-frequency oscillations effects in the Fresnel region light propagation simulation,” Opt. Eng. **42**(11), 3158–3164 (2003). [CrossRef]

**29. **Z. Jaroszewicz, A. Kołodziejczyk, M. Sypek, and C. Gomez-Reino, “Non-paraxial analytical solution for the generation of focal curves,” J. Mod. Opt. **43**(3), 617–637 (1996). [CrossRef]