LiNb03 photonic foundry

Application Notes

• Introduction

  Alenia Marconi Systems is an industrial company in the radar and defence systems market, where its technological innovation and engineering expertise have made it a major international leader.

  The “Integrated Photonic Device” product line has been developed in order to satisfy the high demanding requirements for military applications in the optical processing of microwave signals and fiber optic sensor fields.
  A wide range of Integrated Photonic Devices (IPDs) are offered based on state of the art TAPE (Thermally Annealed Proton Exchanged) and TiD (Titanium Diffusion) technologies on LiNbO 3 substrates, in which the Research Department at Alenia Marconi Systems is among the world leaders. The IPDs utilise electro-optical and acousto-optical interactions with modulating signals to offer advanced and high performance devices for fiber optic communication systems, fiber sensors and scientific applications. Among the most representative devices:

  • wide bandwidth phase, amplitude and frequency modulators (20-40 GHz);
  • high extinction ratio (>50 dB) amplitude modulators;
  • high optical radiation damage threshold;
  • polarisation insensitive acousto-optically tuneable filters;
  • polarisation splitters and filters.

  Custom designed devices are offered to adapt performances to custom requirements and to integrate more functions in the same IPD.
  A new generation of modulators has been developed in order to satisfy the growing request of modulators with a bandwidth wider than 20 GHz up to the Millimeter Wave region (40 GHz).

• Fabrication Technologies

  The devices are realised on LiNbO3 (Lithium Niobate) exploiting state of the art technologies developed by Alenia Marconi Systems for military and space applications, and are the leader for advanced applications in the fiber optic communication systems.
  The X cut of LiNbO3 is the preferred substrate, but it is also possible to create custom devices on both the additional standard cuts Y and Z. The choice of the X cut is based on the higher thermal stability, one order of magnitude higher than that obtained with the Z cut substrates.
  The Thermal Annealed Proton Exchange (TAPE) is the preferred technology for the optical waveguide fabrication, due to its numerous advantages over the more traditional Titanium Diffusion (TiD), among which:

  • higher resistance (2¸3 orders of magnitude) to the damage induced by the optical power;
  • single polarisation operation (the proton exchange induces the increase of the extraordinary index only, without significantly affecting the ordinary index), with residual polarisation crosstalk lower than -60 dB;
  • the extra low polarisation crosstalk, along with other technological advances, allow us to obtain very high extinction ratios (>50 dB) in the interferometric intensity modulators.

  TiD technology is also available as single waveguide technology or integrated with TAPE in the same manufacturing process to realise advanced devices, such as Polarisation Insensitive Acousto-Optical Tunable Filters (AOTF) or Polarisation Splitter/Filters (PS/F).

• Principles of Operation

  3.1 Electro-Optical Modulators

  These modulators exploit the electro-optical properties of the LiNbO 3 . In the linear electro-optic effect the presence of an electrical field induces the change of the refractive index of the material, and subsequently changes the propagation speed of the radiation. Thin film metallic electrodes are deposited onto the substrate surface and used to couple the modulating electrical signal to the optical carrier (figure 1).

  

  3.1.1 Phase Modulators

  The electro-optic effect is used to modulate the optical carrier fase. In this device a couple of electrodes are located on the sides of a single waveguide (figure 1) .
  The modulator transfer function is simply given by:

  

  Where Δф is the relative phase delay, V is the voltage applied to the electrodes and Vπ is the necessary voltage required to obtain a phase delay equal to π.

  3.1.2 Coupling Electrodes

  When the signal is time varying, the simple capacitive electrode configuration, in which they can be considered a capacitor C in parallel with the terminating resistor R, is frequency limited by the RC cut. The value of this cut is usually in the range of 1-2 GHz. In order to obtain modulators efficient enough for the microwave frequencies (20 GHz), coplanar transmission line configurations are adopted for the electrodes, allowing Travelling Wave (TW) coupling between the modulating signal and the optical carrier.

  The higher frequency limit of the TW electrode configuration is possible due to:

  • the difference between the propagation speeds of the modulating signal (c/n RF ) and the optical carrier (c/n opt );
  • the resistive/dielectrical losses experimented by the modulating in the coupling electrodes.

  The difference between the propagation speeds is the major limitation to the bandwidth in the case of conventional TW electrodes. The -3dB bandwidth (BW)is given in this case by the following expression:

  

  where L is the length of the coupling between the TW electrodes and the optical waveguide Δn=nRF-nopt is the difference between the two propagation indexes (typically Δn=2). Because the efficiency of the modulator (Vπ) decreases with L, wider bandwidths imply lower electro-optic efficiencies.

  Alenia Marconi Systems has developed a new TW electrode configuration enabling us to overcome this limitation. In this new configuration, called “Velocity Matched” (VM) the propagation speed of the modulating signal is adapted to that of the optical carrier. The bandwidth of the modulator is now limited by resistive/dielectric losses only, which increase with frequency. Moreover, the electro-optic efficiency is practically bandwidth independent (figure 2) . This makes the VM increasingly advantageous, in terms of efficiency, over the conventional TW electrodes as the bandwidth increases. The electro-optic efficiency of the conventional TW still remains higher than that of the VM configuration for a bandwidth lower than a threshold value, typically in the range of 3¸5 GHz. A further advantage of the VM configuration is the value of the characteristic impedance close to 50 Ohms, perfectly matched with the external signal line, in spite of the 25-35 Ohms of the conventional TW electrodes.

  
  Figure 2: Electro-optic efficiency vs bandwidth

  3.1.3 Amplitude Modulators

  The intensity modulation of the optical carrier is obtained by the phase modulation in the two arms of an integrated interferometer (figure 3). The symmetric configuration of the coplanar TW electrodes allows the push-pull phase modulation of the two arms on the Mac Zehnder interferometer. The modulator transfer function is given by the following expression:

  

  Γ is the optical insertion loss, фRF = πVRF / Vπ and фDC are the phase differences between the two arms of the interferometer induced by signals applied to the RF and the DC electrodes. The DC electrode is usually used to polarise the modulator transfer function at its quadrature point. It has to be noted that the value of the Vp at the RF electrode is now twice the value of the single phase modulator.

  
  Figure 3: Intensity modulator

  3.2 Acousto-optical Tunable Filters

  These devices exploit the collinear interaction between the optical radiation and a Surface Acoustic Wave, which induces the resonant coupling between the TE and the TM polarised optical radiation. The integration of polarisation splitter and filter allows us to obtain a filter transfer function of which the peak wavelength (lf ) is an inverse linear function of the SAW frequency f (figure 4).

  
  Figure 4: AOTF transfer function

  The advanced fabrication process, in which both the TAPE and the TiD technologies are exploited, permit high performances and the possibility of adopting many different configurations to comply to the various customer requirements. An example of these configurations is given in figure 5.

  
  Figure 4: AOTF transfer function

• Main Options

  The technology developed by Alenia Marconi Systems results in advanced and high level performances in the most relevant parameters. The custom integration of a wide range of relevant options in the integrated photonic devices is a common practice for us, allowing the best exploitation of our technological expertise.

  4.1 – High Extinction Ratio (Intensity Modulators)
  Results have proved that using the standard process, the value of the extinction ratio, defined in the intensity modulator as the ratio between the maximum and the minimum transmission, is typically less than 30 dB. The adoption of specific technological advances in the fabrication process allows us to obtain extinction ratios higher than 50 dB, without affecting the other modulator parameters.

  4.2 – Zero Bias Polarisation
  The modulator is passively biased to work close to the requested polarisation point (quadrature, extinction, ….) of its transfer function. Varying lengths of the two interferometer arms is adopted to obtain this passive biasing. The DC voltage demand to keep the modulator at its polarisation point is now less than ± 1 v. This greatly simplifies the electronics for the closed loop stabilisation of the polarisation point, and practically eliminates the so called DC Drift. The Dc drift is a very well know effect in LiNbO 3 devices inducing the drift of the DC voltage vs. time and temperature. This effect, being characterised by a threshold typically close to 1-2 v, is not activated adopting the Zero Bias Polarisation option.

  
  Figure 6: Dual outputs intensity modulator

  4.3 – Dual Output Intensity Modulators
  In this option the “Y” junction at the rear of the interferometer is substituted by a -3 dB directional coupler (figure 6). Both of the two outputs are equivalent to the single output of a conventional interferometer. The two transfer functions are given by the following expression (figure 7):

  

  Both the outputs are polarised at their quadrature points with фDC = 0.

  
  Figure 7: Dual Output modulator transfer functions

  4.4 – Monitoring Output
  This option is useful to monitor the polarisation status and the functionality of the devices integrated in the optical circuit. Part of the optical power (-N dB) is decoupled from the main waveguide and sent to a secondary independent optical output. This is shown in the example in figure 8. In this case the transfer function of the secondary output (IB) is given by the following expression:

  IB = αIA

  In which a is the coupling ratio and IA is the transfer function of the main output.

• Multifunctional Integration

  The technology developed by Alenia Marconi Systems has allowed us to develop a wide range of custom devices in which numerous functions are integrated into the same optical circuit. Furthermore, we are prepared to collaborate with the buyer in the design and optimisation of the devices with the aim of satisfying custom requirements by using our technological expertise.

  
  Figure 8: Example of multifunctional circuit

• Advanced Devices

  Advanced Devices

  The performance of the intensity modulators has been improved adopting a new “velocity matched” configuration for the electro-optical coupling electrodes. The bandwidth has been extended to more than 20 GHz, and further developments are being carried our to extend it to the millimeter wave range (> 40 GHz). The particular electrode configuration and the use of low dielectric constant, low dielectric loss buffer material allows us to obtain 50 ohm impedance and wide bandwidth with high electro-optic efficiency (V p <7 v for 20 GHz modulators at 1550 nm ). This new generation of modulators exploit state of the art TAPE technology on LiNbO3, and benefit all the additional advanced characteristics: elevated extinction ratio (>50 dB), high optical power operation (>300 mw CW at 1300-1550), passive biasing to the quadrature,….. . Fiber optic links are greatly improved in terms of reduced noise figure, lower losses and low ripple in the frequency response by the use of this new generation of modulators.

  6.2 Advanced Functions
  
  Improved devices have been developed for the advanced processing of the optical carrier.
  Examples of these devices are:

  • SSB and Carrier Suppresed Modulators (used for extended dynamic range processing)
  • Frequency Shifters (necessary in coherent architectures).

  
  Figure 9: SEM image of VM electrodes

• Definition of the parameters

  7.1 Optical Insertion Loss
  The Optical Insertion Loss (IL) is defined as the ratio between the maximum output optical power transmitted by the modulator and the optical input power (figure 10). These optical powers are considered in the fiber, therefore the IL includes the propagation loss and the losses at the device-fiber interfaces.

  
  Figure 10: Intensity modulator transfer function
  (quadrature polarisation)

  7.2 – Optical Back Reflections
  It is the ratio between the optical power sent to the input or output of the device and the optical power reflected in the fiber when in a condition of total absorption of the radiation transmitted to the output or input respectively.

  7.3 – V p (RF)
  It is the DC voltage, or alternatively, the amplitude of a low frequency signal (typically at 1 KHz), that is applied to the RF electrodes belonging to a phase or amplitude modulator, to obtain a phase delay equal to p in the phase modulator or to move the intensity modulator transfer function from a maximum transmission to the nearest minimum transmission condition.

  7.4 – S11
  It is defined as the reflection coefficient for the RF power (PRF,in ) at the RF input vs. frequency.

  
  Figure 11: The reference link

  7.5 – S21 (electro-optical)
  This electrical parameter is defined by including the modulator inside an ideal reference fiber optic link (figure 11) characterised by a monomode laser source at a specified wavelength, a zero loss fiber optic transmission line in which the modulator is included, and a photodetector characterised by unitary responsivity (1 A/w) and loaded by a 50W resistor. In these conditions the electro-optic S21 of the modulator is defined as the ratio between the RF power at the output of the link and the RF power of the signal at the input of the modulator, vs. frequency.

  

  This parameter is usually normalised by its value at low frequency. The variation of the S21 vs. frequency describes the change of the electro-optic efficiency of the modulator (figure 12.a and 12.b) . A decay D S21(dB) of the S21 value, measured between two frequencies, f min and f max , corresponds to an increase of the V p vs. frequency. In other terms, the ratio between the voltage amplitudes of the signal, which, at two frequencies, allow the modulation between the maximum and the minimum transmission, is given by the following expression:

  

  The decay of the S21 vs. frequency is caused mainly by two factors:

  • an increase in the resistive/dielectric losses of the signal in the electrode;
  • the decay of the modulation efficiency caused by the increase of the mismatch between the propagation speeds of the modulating signals and the optical carrier.

  It is again evident the advantages in using the “Velocity Matched” configuration in order to eliminate the latter of these two limiting factors.

  
  Figure 12.a

  
  Figure 12.b

  7.6 – Bandwidth
  The bandwidth is defined by the frequency at which the S21 decay is measured as -3 dB.

  7.7 – Extinction Ratio (Intensity Modulators)
  The extinction ratio in the intensity modulator is defined as the ratio between the maximum and the minimum transmission of the modulator. The value of this parameter is usually certified by applying a DC signal to the DC electrode. For a signal at frequency f, the value of the extinction ratio is defined as the ratio between the maximum and the minimum transmission when the voltage amplitude is set at exactly Vp of that frequency, and the modulator is polarised exactly at its quadrature point. This value of the extinction ratio does not change with the frequency.

• Vπ measurement

  Introduction

  The Vπ of an interferometric integrated optical amplitude modulator is defined as the voltage change requested to move the modulator from the minimum to the maximum optical transmission. It is a function of frequency f of the modulating signal V(f), and is defined in the modulator optical transmission function:

  

  Where G is the optical insertion loss of the modulator.

  Vπ measurement: low frequency

  The modulator transmission function is characterised by including the modulator in a standard test bed in which a laser, at the design wavelength, and a photodetector, with the necessary bandwidth, are both integrated into a fiber optic circuit. Regarding the frequency, at which the Vp has to be measured, it allows us to drive the modulator with a good quality saw-tooth signal (f < 1 MHz). The observation of the electrical signal at the photodetector output with a standard oscilloscope will allow us to measure the Vπ.

  

  The input and output electrical signals are given to the X and Y inputs of the oscilloscope. In this condition it will be sufficient to adjust the amplitude VS and the offset of the Saw-tooth signal in order to obtain a perfect sinusoidal at the oscilloscope, subsequently the Vp will be given by:

  

  Vπ measurement: high frequency

  The Vπ can also be measured at any frequency adopting the following method.
  Two RF signals at slightly different frequencies (w1, w2>>Dw) and equal powers ( ), are fed into the modulator’s electrical input. Assuming the voltage of the applied signals is much less than that of the Vp at the frequency f (Vπ(f)>> V1=V2=Vin(f)= ) , it is possible to demonstrate the following equation:

  

  Where Pout,1 and Pout,3 are the RF output signals at the fundamental frequency w1 (or w2), and at the intermodulation 2w1-w2 (or 2w2-w1) respectively.