JP3102811B2 - Ultrafast optical waveform measurement method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for measuring an ultrafast optical waveform such as an ultrashort optical pulse, and more particularly to a method for measuring an optical pulse waveform from 100 fs to 100 ps. The present invention relates to an ultra-high-speed optical waveform measurement method capable of measuring by converting into an optical signal.

[0002]

2. Description of the Related Art A technique for generating an ultrashort optical pulse is 198.
It progressed rapidly in the 0's, and it is now possible to obtain tunable and stable subpicosecond pulses. Such an ultrashort light pulse greatly contributes to fields such as physics and chemistry as a light source for time-resolved spectroscopy and nonlinear spectroscopy. Furthermore, in recent years, a technique for freely shaping the pulse waveform of an ultrashort light pulse has begun to appear, and a way is being opened for application as a more advanced optical measurement means. In addition, if information is written into an ultrashort optical pulse using waveform shaping, Tbit
Since an optical signal of s / s or more can be generated, there is a possibility that optical communication and optical information processing at a speed several orders of magnitude higher than at present can be realized. However, there is no point in generating such a high-speed optical signal unless it can be read. The response speed of current photodetectors (PIN photodiodes, avalanche photodiodes, etc.) is as fast as several
Since 0 GHz is the limit, an optical signal of Tbits / s or more cannot be directly read.

[0003] Typical methods for reading the light waveform in a region faster than picosecond include a method using a streak camera and a non-linear correlation method (SHG autocorrelation, cross correlation, etc.). The former has the advantage that the optical waveform can be measured directly, but has the disadvantage that the time resolution is from picoseconds to sub-picoseconds and that it is very expensive. The latter has no problem in the time resolution, but it is necessary to sweep the probe pulse, so that it is not applied to optical communication or the like that requires instantaneous reading.

[0004]

SUMMARY OF THE INVENTION The present invention has been made in view of such a situation, and an object of the present invention is to solve the above-mentioned drawbacks of the prior art so that the time resolution is not limited to electronics. It is an object of the present invention to provide an ultra-fast optical waveform measurement method based on a completely new principle that is inexpensive and does not require sweeping of a probe pulse.

[0005]

According to the present invention, there is provided an ultrahigh-speed optical waveform measuring method according to the present invention, in which a measuring time waveform is incident on a diffraction grating, and diffracted light of a predetermined order other than the 0th order is spatially Fourier-transformed. After the phase term is erased by optically mixing it with the probe light pulse in a nonlinear substance, the phase term is erased, and then Fourier-transformed again spatially, and the measured time waveform is read out as a spatial pattern similar to that. It is a method.

In this case, the mixing of the probe light pulse and the light in the nonlinear material is a sum frequency light mixing, and
In the case of four-wave mixing, in the case of four-wave mixing, the probe light pulse is made incident on another diffraction grating before being made incident on the nonlinear material, and diffracted light of a predetermined order other than the 0th order is made into the nonlinear material. The light may be incident to perform four-wave mixing. Also, each spatial Fourier transform can be performed optically.

[0007] This ultra-high-speed optical waveform measurement method can be applied to convert a temporal serial optical signal into a spatially parallel optical signal.

[0008]

According to the present invention, a measurement time waveform is made incident on a diffraction grating, and diffracted light of a predetermined order other than the 0th order is spatially Fourier-transformed, and its phase term is optically mixed with a probe light pulse in a nonlinear material. After the phase term is erased, the Fourier transform is performed spatially again and the measured time waveform is read out as a spatial pattern similar to that, so the time waveform can be read instantaneously, and the time resolution is not limited to electronics. It is inexpensive and does not require sweeping of probe pulses. By this method, 100 fs to 100
The waveform of the optical pulse of ps can be measured.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The ultra-high-speed optical waveform measurement method of the present invention is basically realized by utilizing the inverse process of the waveform shaping technique. First, the principle will be described below.

In the present invention, a time waveform cannot be directly measured as an electric signal using a photodetector for an optical waveform faster than picoseconds. Therefore, in the present invention, the time waveform is converted into a spatial pattern to be used as a spatial pattern of light. Take the reading method.
In this method, a temporal waveform is projected into space using a diffraction grating, and the spatial pattern is fixed and read by nonlinear light mixing on a Fourier transform plane. FIG. 1 shows the principle diagram. Hereinafter, the process of reading the time waveform as a spatial pattern will be described in detail.

The time waveform of the light pulse to be measured is U (t). For simplicity, the spatial pattern of the measured light pulse is assumed to be a uniform function R (x) over a sufficiently wide range. That is, the incident pulse is set as follows: E _input (x, t) = R (x) U (t) (1) U (t) is a pulse waveform to be measured. When this pulse hits the diffraction grating G, the first-order diffracted light is expressed by the following equation immediately after the diffraction grating.

E ₁ (t, x) = R (sx) U (t−αx) (2) where x is a space coordinate perpendicular to the traveling direction of light and parallel to the paper, and α and s are This is a diffraction parameter determined by the groove spacing of the diffraction grating and the incident angle. Here, the conversion from t to t-αx indicates that the pulse is inclined with respect to the traveling direction of the diffracted light as shown in FIG. Since the light diffracted from the adjacent grooves of the diffraction grating G has different optical path lengths by the length of the wavelength, the diffracted pulse is inclined as shown in FIG. 1, and the time waveform is projected on the x-axis at each time. Have been. t →
This is converted to t-αx. However, as can be seen from FIG. 1, since the projected position moves with time, even if the spatial pattern is observed by integrating the time at this stage, the time waveform cannot be known.

Next, the electric field is converted to a lens L having a focal length f.
_When spatially Fourier-transformed by using ₁ , at the focal plane, E ₂ (t, x) ∝R (st / α) V (2πx / αλf) × exp (−i2πxt / αλf) (3) . Here, V (Δω) is a Fourier transform of U (t). E ₂ (t, x) is the product of the Fourier transform of the waveform to be measured and the phase term exp (−i2πxt / αλf). If this is directly Fourier transformed by the lens L ₂ , E _{3 ′} (t, x) x) ∝R (st) U (t + αx) (4 ′) This is because the slope is opposite to the equation (2), but nothing is known.

Therefore, the phase term exp (-i2πxt
/ Αλf) is eliminated and Fourier-transformed again by the lens L ₂ , the following equation is obtained: E _output (t, x) ∝R (st / α) U (αx) (4) Looking at equation (4), the time waveform U of the incident pulse
It can be seen that (t) is projected on the x-axis. That is,
Since the time waveform is projected at the same position regardless of the time, observing the spatial pattern by performing time integration yields a pattern similar to the time waveform.

In order to eliminate the phase term exp (-i2πxt / αλf), it is sufficient to optically mix the photoelectric field having the complex-conjugated phase term with the nonlinear substance N. In the case of FIG.
The light for that is represented as E _P (t, x). There are several ways to create E _P , but here, only those utilizing second-order and third-order nonlinearities will be described.

FIG. 2A shows an arrangement in the case of utilizing second-order nonlinearity. E those obtained by Fourier transformation of much shorter pulses as compared to the measured waveform in the diffraction grating G ₁ and the lens L ₃
Is _P, to obtain a spatial pattern by inverse Fourier transform of the sum frequency mixing light E _P E ₂ and E _2. Diffraction grating G ₁ and lens L ₃
, If used as having the same parameters as used for the waveform to be measured, the direction of diffracting Conversely, E _P phase e
Since it has xp (+ i2πxt / αλf), the phase term is canceled.

FIG. 2B shows an arrangement in the case of utilizing the third-order nonlinearity. As with the case of using the second-order nonlinear, a material obtained by Fourier transformation of much shorter pulses as compared to the measured waveform in the diffraction grating G ₂ and lens L ₄ and E _P, 4-wave mixing light E _P E _P E Inverse Fourier transform of ₂ ^* to get a spatial pattern. In this case, either the groove spacing of the diffraction grating G ₂ used in E _P to that of half the grating G using the waveform to be measured, or, lens L ₄ having a focal length 2f
, The phase factor of E _P is half of E ₂ ,
The phase term is cancelled.

In FIG. 2 (c), the short pulse is simply referred to as E _P , and the four-wave mixed light E _P E _P E with E ₂ is used.
₂ ^* is Fourier inverse transformed. In other words, by reading by cutting a portion of the E ₂ in a very short pulses, a method that would ignore the phase term. This method
Since only a part of E ₂ is used, it seems that the efficiency is low. However, since the short pulse is used as E _P , the peak power becomes large, and the four-wave mixing light E _P E _P E ₂ ^* is shown in FIG.
It becomes stronger than the case of (b). The optical system is mirror M
Is very simple.

By the above-described method, ideally, a similar shape of the measured waveform can be obtained as a spatial pattern as expressed by the equation (4). A factor is reducing the accuracy of waveform measurement. The factors and countermeasures are summarized below.

1) The assumption that the spatial spread of the pulse to be measured is uniform over a range sufficiently larger than the time width multiplied by the speed of light c is used to derive equation (2) from equation (2). ing. In practice, the spatial spread (ie, beam diameter)
Is finite, the Fourier transform of equation (3) is a Fourier transform with a weight of a function R (x) representing a spatial spread. If the time width of the measured waveform is 100 ps or less, this effect can be ignored if the beam diameter on the diffraction grating surface is several cm. When the time width is 100 ps or more, the size of the diffraction grating is finite, so that this effect cannot be ignored, and the accuracy of waveform measurement decreases.

[0021] 2) When the actual light mixing, in order to separate the light mixing light and original optical, to intersect at an angle to the E ₂ and E _P. This makes the cancellation of the phase term incomplete. However, this effect is 100 fs
(Fs: 10 ⁻¹⁵ seconds) It has been clarified from the result of the numerical calculation that if the pulse is measured longer than 10 degrees, no problem occurs if the angle is set to 10 degrees or less.

3) Since the short pulse for producing E _P has a finite width _UP (t) instead of the δ function, the spatial pattern actually obtained is not an exact analog of the waveform to be measured. , convolution of the _{U P (t) U * U} P (α
x).

In summary, the measured waveform is 100 fs
From 100 ps to a shorter pulse U
If prepared _P (t), the time waveform accuracy of the width of the U _P (t) can be read as a spatial pattern. This means that an ultrafast optical signal obtained by shaping the waveform of an ultrashort optical pulse can be read out by optically mixing it with the original ultrashort optical pulse. This corresponds to serial-to-parallel conversion of an optical signal, and an actual example will be described below. When it is not ready it is shorter pulse than the waveform to be measured (such as if you want to know the mode locked pulse waveform) is, by using the measured pulse itself as E _P, a convolution or autocorrelation waveform of a pulse waveform You can know. Examples will be described later.

An example of waveform measurement will be described below. In the arrangement of FIG. 2 (c), the illustrating an example of measurement of the waveform using the measured pulse itself as E _P. Because of the use of four-wave mixing of E _P E _P E ₂ ^*, as the physical mechanisms, E
E _P is diffracted by the dynamic diffraction grating E _P E ₂ ^* formed by _P and E ₂ . At this time, the relaxation magnitude of the time T ₁ of the dynamic diffraction grating, the resulting space pattern is different. If the relaxation time is much longer than the measured pulse width,
A signal proportional to the autocorrelation function of the incident photoelectric field is obtained.
Conversely, if it is much shorter, a third-order intensity autocorrelation waveform of the pulse is obtained. In any case, a signal corresponding to the pulse waveform can be obtained in real time (without the need for scanning unlike the SHG autocorrelation method).

In the arrangement shown in FIG. 3, the pulse waveform of a picosecond pulse obtained from a mode-locked dye laser (Rhodamine 6G) was measured. As a non-linear substance N, a dye solution (DODCI) having a concentration of 10 ^-5 M / L in ethanol having a thickness of 1
What was put in the cell of mm was used. Further, a diffraction grating G of 2400 lines / mm was used, and the lens L ₁ ,
And using the cylindrical lens of focal length 300mm as L _2. The spatial pattern was observed with a CCD camera. FIG. 4 shows the obtained spatial pattern. Since the relaxation time of DODCI is much longer than the pulse width to be measured, the autocorrelation function of the optical electric field has been obtained. The horizontal axis in FIG. 4 is CCD
The position of the camera is represented, but is converted to the corresponding time. In this method, since only the dye solution and the CCD camera are required, a very inexpensive ultrashort light pulse measurement system is obtained.

Next, an example in which the method of the present invention is used for serial-to-parallel conversion of a high-speed optical signal will be described. As described above, if the information is written by shaping the waveform of the ultrashort optical pulse, an optical signal having a capacity of Tbits / s or more can be generated. However, there is currently no photodetector capable of reading such a high-speed optical signal, so that the time waveform is converted into a spatial pattern and read. That is, it is necessary to perform serial-to-parallel conversion of the optical signal. Considering that the above-described waveform measurement method is used for such a purpose, the relaxation time of the nonlinear material becomes a problem. The length of the signal that can be generated by waveform shaping is 10
Since it is up to about 0 ps, to maximize the transmission capacity, 1
It is necessary to perform serial-to-parallel conversion by repeating 00 ps. That is, the relaxation time of the nonlinear material is 100p
s or less is required. As a non-linear material that satisfies the requirements, ZnS epitaxially grown on a GaAs substrate
e thin film was used.

ZnSe has an exciton resonance at a wavelength of 442 nm, has a large third-order nonlinear susceptibility (χ ⁽³⁾ -10 ⁻² esu) at a temperature of 9 ° K, and has a relaxation time of a generated dynamic diffraction grating. Is reported to be about 20 ps. A 2.4 μm-thick ZnSe is prepared, and a reflection type 4
Was made serial → parallel conversion using wave mixing light E _P E _P E ₂ ^*.

FIG. 5 shows an arrangement for this purpose. From one ultrashort optical pulse, as shown in the figure, two probe pulses separated by 35 ps and 3
A bit serial signal was created. The wavelength of the light source is Zn
The thickness is set to 442 nm in accordance with the exciton resonance of Se. In addition, a ZnSe thin film is used to increase the exciton resonance effect.
Cool to 9 ° K. Four-wave mixing light E _P E _P E ₂ ^* of the space pattern is detected by a CCD camera.

It was confirmed that the spatial pattern detected by the CCD camera coincided with the serial signal waveform. FIG. 6 shows an example of the detection pattern. (C) of FIG.
The probe pulses “a” and “b” read the serial signals “011” and “100”, respectively, and the sum is detected. (Because the CCD camera detects the time-integrated signal, it becomes the sum of the two signals. .). At this time, if the probe pulse "b" is eliminated, only the "011" signal is converted. Conversely, if the probe pulse "a" is eliminated, only the "110" signal is converted. ) It turns out that the probe pulse works as a gate pulse for conversion, and its response speed is 35 ps or less.

Using this system, how fast the conversion cycle can be taken was evaluated as follows. The time lag t
Another probe pulse was prepared with a delay of _p . The first signal pulse and the probe pulse form a dynamic grating, but the dynamic grating must disappear before the arrival of the second delayed probe pulse. Otherwise, an erroneous signal will be obtained even though there is no signal signal synchronized with the second probe pulse. FIG. 7 shows the result of measuring this error signal while changing the time delay. When the time delay is 13 ps or more, no error signal is observed. This confirms that the conversion cycle of this system extends over 70 GHz.

The ultrashort light pulse used in this evaluation is 4p
s, it could only be created up to 0.12 THz as a serial signal, but if subpicosecond pulses are used, T
Hz or higher can be converted in a cycle of several tens of GHz.

The conversion efficiency of this system depends on the diffraction efficiency of a dynamic grating formed in a nonlinear material.
Since the third-order nonlinear optical effect is used, the diffraction efficiency depends on the intensity of the input light. In the above case, the value was about 0.1%. This value is by no means large but sufficient for the sensitivity of the photodetector (in this case: CCD camera).

The real-time waveform measurement method using the waveform shaping method has been described above. This method is effective for a pulse of 100 fs to 100 ps. If a probe pulse shorter than the pulse to be measured is prepared, the waveform can be instantaneously read as a spatial pattern. Further, even when the measured pulse itself is used as the probe pulse, a waveform convolution or a third-order correlation waveform can be obtained.

[0034]

As described above, according to the ultrafast optical waveform measuring method of the present invention, the following effects can be obtained. 1) Since a dye can be used as a nonlinear substance, an inexpensive waveform measurement system can be provided over a wide range of wavelengths.

2) It does not require scanning and can be used for measurement of a single event.

3) Even if the measured pulse itself is used as the probe pulse, the spatial pattern obtained is a third-order correlated waveform, so that the asymmetry of the waveform can be known.

[Brief description of the drawings]

FIG. 1 is a principle diagram for explaining the measurement principle of the ultrafast optical waveform measurement method of the present invention.

FIG. 2 shows several arrangements for canceling phase terms.

FIG. 3 is a diagram showing an arrangement for measuring a pulse waveform of a picosecond pulse obtained from a mode-locked dye laser.

FIG. 4 is a diagram showing an example of measuring a waveform by the arrangement of FIG. 3;

FIG. 5 is a diagram showing an arrangement for serial-to-parallel conversion.

6 is a diagram showing an example of a detection pattern according to the arrangement of FIG.

FIG. 7 is a diagram showing measurement results for evaluating the speed of a conversion cycle according to the arrangement of FIG. 5;

[Explanation of symbols]

G, G _1, G ₂ ... diffraction grating _{L 1, L 2, L 3} , L 4 ... Lens N ... nonlinear material M ... mirror

Continuing from the front page (56) References JP-A-1-201122 (JP, A) Kazuhiro Ema et al. "Ultrafast Optical Waveform Measurement Method Using Waveform Shaping and Its Application" Annual Report of Research Institute, Faculty of Engineering, The University of Tokyo (Japan) Tokyo University Faculty of Engineering Laboratory, September 1991, Volume 50, pp. 79-85 Ema "Real-Time Ultrashort Pulse and then Pulse-Shape Measurement Use a Dynamic Grating" Japan Announcement of Journal of Applied Physics. 30 No. 12A pp. 2046-2049 Kazuhiro Ema "Control of Ultrashort Optical Pulse Waveforms" Japan Society of Applied Physics (Japan) Japan Society of Applied Physics Issued January 5, 1991 Vol. 60 No. 1 pp. 41-44 Eka Kazuhiro et al. On Real-Time Waveform Shaping Method of the University of Tokyo ”Annual Report of the Institute of Engineering, The University of Tokyo (Japan) The Institute of Engineering, The University of Tokyo Published September 1990, Vol. 49, pp. 93-96 (58) Fields surveyed (Int. Cl. ⁷⁾ G01J 11/00 G01J 1/00 G01J 1/42-1/44

Claims (6)

(57) [Claims] 1. A measurement time waveform is made incident on a diffraction grating, and diffracted light of a predetermined order other than the 0th order is spatially Fourier-transformed, and its phase term is optically mixed with a probe light pulse in a non-linear substance to be erased. An ultra-high-speed optical waveform measurement method characterized by spatially Fourier transforming again after erasing a phase term and reading out a measurement time waveform as a spatial pattern similar to the waveform. 2. The ultrafast optical waveform measurement method according to claim 1, wherein the optical mixing of the probe light pulse and the nonlinear substance in the nonlinear material is a sum frequency optical mixing. 3. The method according to claim 1, wherein the mixing of the probe light pulse and the light in the nonlinear material is four-wave mixing.
The ultrafast optical waveform measurement method described. 4. The method according to claim 1, wherein the probe light pulse is made incident on another diffraction grating before being made incident on the nonlinear material, and diffracted light of a predetermined order other than the zero order is made incident on the nonlinear material to perform four-wave mixing. 4. The method for measuring an ultrafast optical waveform according to claim 3, wherein: 5. The ultrafast optical waveform measurement method according to claim 1, wherein each of the spatial Fourier transforms is performed optically. 6. An ultra-high-speed optical waveform measurement method according to claim 1, wherein a temporal serial optical signal is converted into a spatial parallel optical signal.