JP2016014376A - Internal combustion engine exhaust emission control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To purify NOx discharged from a NOx occlusion reduction catalyst at a time of a NOx emission control in an internal combustion engine exhaust emission control device.SOLUTION: An exhaust emission control device comprises: a NOx occlusion reduction catalyst; and an electrochemical reactor provided at a downstream position of the catalyst. The device executes a lean control to supply exhaust gas at a lean air-fuel ratio to the catalyst by controlling an air-fuel ratio of an air-fuel mixture within a combustion chamber to a lean air-fuel ratio when an emission condition to emit NOx from the catalyst is not satisfied. On the other hand, the device executes a NOx emission control to supply exhaust gas at a rich air-fuel ratio to the catalyst over a predetermined period when the emission condition is satisfied. Furthermore, the device starts applying a voltage to the reactor at a time of or after starting the NOx emission control, and then detects timing just before start of decomposition of a solid electrolyte of the reactor on the basis of an impedance of the reactor and stops applying the voltage to the reactor at timing of the detection.

Description

  The present invention relates to an exhaust gas purification apparatus for an internal combustion engine that purifies NOx by a NOx storage reduction catalyst and an electrochemical reactor.

  2. Description of the Related Art An exhaust gas purification device for an internal combustion engine (hereinafter also referred to as “conventional device”) having a NOx storage reduction catalyst (NOx catalyst) in an exhaust passage is known (see, for example, Patent Document 1). This NOx occlusion reduction catalyst occludes NOx (nitrogen oxide) in the exhaust gas when the air-fuel ratio of the exhaust gas flowing into the catalyst is a lean air-fuel ratio larger than the stoichiometric air-fuel ratio. On the other hand, when the air-fuel ratio of the exhaust gas flowing into the catalyst becomes a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio, the catalyst releases stored NOx and reduces it to unburned HC (hydrocarbon) in the exhaust gas. The released NOx is purified by the reaction.

JP 2005-171853 A JP 2009-209833 A

  By the way, when a large amount of NOx is occluded in the NOx occlusion reduction catalyst, the amount of NOx that can be further occluded by the catalyst decreases. Therefore, in order to recover the amount of NOx that can be stored by the catalyst, in the conventional apparatus, exhaust gas having a rich air-fuel ratio is supplied to the catalyst, and thereby NOx stored in the catalyst is released. . Most of the released NOx is reduced and purified by the catalyst. However, in a specific period immediately after the rich air-fuel ratio exhaust gas starts to flow into the catalyst, part of the NOx released from the catalyst may flow out of the catalyst without being reduced by the catalyst. . Such outflow of NOx reduces the NOx purification rate of the exhaust purification device.

  Thus, as one of the reasons why NOx flows out from the NOx storage reduction catalyst, the rate at which NOx is released from the catalyst during the specific period (NOx release rate) is the rate at which NOx is reduced by the catalyst (NOx reduction). It is conceivable that it is larger than (speed).

  The present invention has been made to address the above-described problems. That is, one of the objects of the present invention is an exhaust gas purification apparatus for an internal combustion engine provided with a NOx storage reduction catalyst, and when NOx flowing out from the catalyst is supplied when rich air-fuel ratio exhaust gas is supplied to the catalyst. An object of the present invention is to provide an exhaust emission control device that can be purified (hereinafter referred to as “the device of the present invention”).

  The apparatus of the present invention includes a NOx occlusion reduction catalyst disposed in an exhaust passage of an internal combustion engine, and an electrochemical reactor disposed in a position downstream of the catalyst in the exhaust passage. The electrochemical reactor has a solid electrolyte part, a first electrode part, and a second electrode part. The solid electrolyte portion is made of a solid electrolyte having oxygen ion conductivity. The first electrode part includes a NOx occlusion part that is disposed on the surface of the solid electrolyte part and occludes NOx. The second electrode part is disposed on the surface of the solid electrolyte part. Further, the electrochemical reactor electrochemically reduces and purifies NOx in the first electrode portion when a voltage is applied between the first electrode portion and the second electrode portion.

  The device of the present invention further includes a voltage application unit. The voltage application unit can apply a voltage between the first electrode unit and the second electrode unit.

  The device of the present invention further includes a first control unit. The first control unit sets the air-fuel ratio of the air-fuel mixture in the combustion chamber of the engine from the stoichiometric air-fuel ratio when a condition (release condition) for releasing NOx stored in the catalyst from the catalyst is not satisfied. The lean air-fuel ratio is also controlled to be larger. As a result, a lean air-fuel ratio exhaust gas is supplied to the catalyst. Such control is also referred to as “lean control” for convenience. During lean control, NOx in the exhaust gas flowing into the catalyst is occluded by the catalyst.

  The release condition is, for example, that the amount of NOx occluded in the catalyst “has reached the limit value of the amount of NOx occluded by the catalyst” or “has reached a value close to the limit value” Is.

  On the other hand, when the release condition is satisfied, the first control unit executes NOx release control for supplying exhaust gas having a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio to the catalyst over a predetermined period. When this NOx release control is executed, NOx stored in the catalyst is released from the catalyst. Most of the released NOx is reduced and purified by unburned HC in the exhaust gas in the catalyst. In other words, a part of the released NOx is not reduced in the catalyst but flows out downstream of the catalyst and flows into the electrochemical reactor.

  In the NOx release control, the device according to the present invention, for example, controls the air-fuel ratio of the air-fuel mixture in the combustion chamber of the engine to a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio, or the catalyst in the exhaust passage. The rich air-fuel ratio exhaust gas is supplied to the catalyst by supplying a reducing agent (for example, hydrocarbon) into the exhaust gas at an upstream position.

  The device of the present invention further includes a second control unit. The second control unit starts application of the voltage using the voltage application unit at the start of the NOx release control or after the start of the NOx release control. That is, the device of the present invention applies a voltage between the first electrode portion and the second electrode portion of the electrochemical reactor at or after the start of the NOx release control. Thus, NOx flowing into the electrochemical reactor is electrochemically reduced and purified at the first electrode portion.

  As described above, NOx may flow out from the catalyst in a specific period immediately after the rich air-fuel ratio exhaust gas starts to be supplied to the catalyst (immediately after it starts to flow in). However, at this time, since a voltage is applied between the first electrode portion and the second electrode portion of the electrochemical reactor, NOx flowing out from the catalyst and flowing into the electrochemical reactor is reduced and purified in the reactor. The Thus, the device of the present invention can purify NOx flowing out from the catalyst.

  As described above, after the start of NOx release control, NOx flows out of the NOx storage reduction catalyst and flows into the electrochemical reactor. The amount of NOx flowing into the electrochemical reactor (inflowing NOx amount) gradually increases and then gradually decreases. On the other hand, the electrochemical reactor reduces NOx occluded in its own NOx occlusion unit and NOx released by NOx release control and flowing into the electrochemical reactor. Accordingly, the impedance measured by the impedance measuring unit decreases with an increase in the inflow NOx amount after a while from the start of voltage application, and then increases with a decrease in the inflow NOx amount. When the inflow NOx amount becomes substantially “0”, only NOx stored in its own NOx storage unit is reduced, and when the reduction of NOx stored in the NOx storage unit is completed, oxygen in the solid electrolyte is moved. Start letting. That is, if the voltage is applied even after the reduction of NOx occluded in the NOx occlusion unit is completed, and the air-fuel ratio of the exhaust gas flowing into the electrochemical reactor at that time is a rich air-fuel ratio, the solid electrolyte Decomposition begins. This decomposition of the solid electrolyte reduces the performance of the electrochemical reactor.

  Therefore, the device of the present invention further includes an impedance measuring unit that measures “impedance of the solid electrolyte part between the first electrode part and the second electrode part”. Then, the second control unit detects a time point immediately before starting the decomposition of the solid electrolyte based on the impedance measured by the impedance measurement unit after starting the application of the voltage using the voltage application unit. . Further, the second control unit stops the application of the voltage using the voltage application unit at the detected “time immediately before the start of the decomposition of the solid electrolyte”.

  Therefore, the device of the present invention can reliably stop the application of voltage at the time after the reduction of NOx released by the NOx release control and before the decomposition of the solid electrolyte starts. Therefore, according to the device of the present invention, NOx released by NOx release control can be purified reliably and decomposition of the solid electrolyte can be avoided.

  In this case, the second control unit obtains an impedance measured by the impedance measurement unit as an initial value when application of the voltage using the voltage application unit is started, and then measured by the impedance measurement unit. It is configured to detect a time point when the initial value is increased after the impedance is reduced and reached the initial value as a time point immediately before the decomposition of the solid electrolyte starts.

  In addition, when starting the application of the voltage using the voltage application unit after the start of the NOx release control, the second control unit flows out of the catalyst by the NOx release control after the start of the NOx release control. It is preferable that the application of the voltage is started before the time point when NOx starts to flow into the electrochemical reactor. According to this, when the application of the voltage is started after the start of the NOx release control, the NOx flowing out from the catalyst can be more reliably reduced and purified in the electrochemical reactor. In addition, since the voltage application time can be shortened compared to the case where the voltage application is started simultaneously with the start of the NOx release control, the power consumption can be reduced.

  Furthermore, according to these, since the time immediately before the start of the decomposition of the solid electrolyte can be reliably detected, the voltage application can be terminated at an appropriate timing.

  Other objects, other features and attendant advantages of the present invention will be readily understood from the description of the embodiments of the present invention described with reference to the following drawings.

FIG. 1 is an overall view of an internal combustion engine to which an exhaust emission control device according to an embodiment of the present invention is applied. 2A is a cross-sectional view of the electrochemical reactor shown in FIG. 1 in an enlarged manner, and FIG. 2B is an enlarged view of the reactor partition shown in FIG. 2A. It is sectional drawing of the same partition. FIG. 3 is a view similar to FIG. 2B, and FIG. 3A is a view showing the reactor when NOx is stored in the NOx storage layer in the reactor shown in FIG. B) is a diagram showing the reactor when NOx is reduced and purified in the reactor. FIG. 4 is a time chart showing the voltage applied to the reactor shown in FIG. 2, the impedance of the reactor, and the like. FIG. 5 is a flowchart showing a rich spike routine executed by the CPU of the ECU shown in FIG. FIG. 6 is a flowchart showing a reactor voltage application routine executed by the CPU of the ECU shown in FIG. FIG. 7 is a diagram similar to FIG. 4, and is a time chart for explaining the operation of the exhaust emission control device according to the modification of the embodiment of the present invention. FIG. 8 is a flowchart showing a reactor voltage application routine according to a modification of the embodiment of the present invention executed by the CPU of the ECU shown in FIG.

  Hereinafter, an exhaust emission control device for an internal combustion engine according to an embodiment of the present invention will be described with reference to the drawings. The exhaust emission control device according to this embodiment is applied to the internal combustion engine (engine) 10 shown in FIG. The engine 10 is a multi-cylinder (in this example, in-line four cylinders), 4-cycle, piston reciprocating type, and diesel engine. The engine 10 includes an engine body 20, a fuel supply system 30, an intake system 40 and an exhaust system 50.

  The engine main body 20 includes a main body 21 including a cylinder block, a cylinder head, a crankcase, and the like. Four cylinders (combustion chambers) 22 are formed in the main body 21. A fuel injection valve (injector) 23 is disposed above each cylinder 22. The injection valve 23 opens in response to an instruction from an engine ECU (electronic control unit) 80 described later, and directly injects fuel into the cylinder 22.

  The fuel supply system 30 includes a fuel pressurization pump (supply pump) 31, a fuel delivery pipe 32, and a common rail (accumulation chamber) 33. The discharge port of the pump 31 is connected to the fuel delivery pipe 32. The fuel delivery pipe 32 is connected to the common rail 33. The common rail 33 is connected to the fuel injection valve 23. The pump 31 is configured to pump up the fuel stored in a fuel tank (not shown) and then pressurize the pump 31 to supply the pressurized high-pressure fuel to the common rail 33 through the fuel delivery pipe 32.

  The intake system 40 includes an intake manifold 41, an intake pipe 42, an air cleaner 43, a compressor 44a of a supercharger 44, an intercooler 45, a throttle valve 46, and a throttle valve actuator 47.

  The intake manifold 41 includes a “branch portion connected to each cylinder 22” and a “collection portion where the branch portions are gathered”. The intake pipe 42 is connected to the collecting portion of the intake manifold 41. The intake manifold 41 and the intake pipe 42 constitute an intake passage. In the intake pipe 42, an air cleaner 43, a compressor 44a, an intercooler 45, and a throttle valve 46 are sequentially arranged from the upstream side to the downstream side of the flow of intake air. The throttle valve actuator 47 changes the opening degree of the throttle valve 46 in accordance with an instruction from the ECU 80.

  The exhaust system 50 includes an exhaust manifold 51, an exhaust pipe 52, a turbine 44b of the supercharger 44, and an exhaust purification device 53.

  The exhaust manifold 51 includes a “branch portion connected to each cylinder 22” and a “collection portion where the branch portions are gathered”. The exhaust pipe 52 is connected to a collecting portion of the exhaust manifold 51. The exhaust manifold 51 and the exhaust pipe 52 constitute an exhaust passage. In the exhaust pipe 52, a turbine 44b and an exhaust purification device 53 are sequentially arranged from the upstream to the downstream of the flow of the exhaust gas.

  The exhaust purification device 53 is directed from the upstream to the downstream of the flow of the exhaust gas. The NOx storage reduction catalyst (hereinafter referred to as “NSR catalyst”) 54, the electrochemical reactor 55, and the diesel particulate filter (hereinafter referred to as “DPF”) 56) in order. The DPF 56 collects particulates in the exhaust gas.

  The NSR catalyst 54 stores NOx in the exhaust gas when the air-fuel ratio of the exhaust gas flowing into the NSR catalyst 54 is a lean air-fuel ratio larger than the stoichiometric air-fuel ratio. On the other hand, the catalyst 54 releases the stored NOx and reduces and purifies the released NOx by the unburned HC in the exhaust gas when the air-fuel ratio of the exhaust gas flowing into it becomes a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio.

  As shown in FIG. 2A, the electrochemical reactor 55 includes a plurality of partition walls 55w extending in parallel with each other and a plurality of partition walls (not shown) extending in parallel with each other in the vertical direction with respect to the partition walls 55w. A member having a so-called “honeycomb structure” (hereinafter, referred to as “honeycomb member”) 55h having a large number of formed passages 55p is included. The honeycomb member 55h is disposed in the exhaust pipe 52. More specifically, the honeycomb member 55h is disposed in the exhaust pipe 52 so that the outer wall surface 55o is in close contact with the inner wall surface 52i of the exhaust pipe 52 over the entire circumference. The exhaust gas flowing in the exhaust pipe 52 passes through the passage 55p of the honeycomb member 55h, and then flows out from the member 55h.

  As shown in FIG. 2B, which shows “part 55w ′ of the partition wall 55w” surrounded by a chain line in FIG. 2A, the partition wall 55w of the electrochemical reactor 55 is a solid electrolyte layer (solid Electrolyte part) 60, cathode electrode layer 61 c, anode electrode layer 61 a, NOx occlusion layer (NOx occlusion part) 62, and power supply circuit (voltage application part) 63.

  The solid electrolyte layer 60 is made of a known porous solid electrolyte having oxygen ion conductivity, for example, yttria stabilized zirconium (YSZ).

  The anode electrode layer 61 a is disposed on one surface of the solid electrolyte layer 60. The electrode layer 61a is made of a porous material made of cermet of yttria stabilized zirconium and silver (Ag). The cathode electrode layer 61 c is disposed on one surface of the electrolyte layer 60 opposite to the anode electrode layer 61 a with respect to the solid electrolyte layer 60. The electrode layer 61c is also made of a porous material made of cermet of yttria stabilized zirconium and silver (Ag).

  The NOx occlusion layer 62 is disposed on the surface of the electrode layer 61c opposite to the solid electrolyte layer 60 with respect to the cathode electrode layer 61c. The NOx storage layer 62 is made of a porous material made of cermet of yttria stabilized zirconium and barium (Ba). The cathode electrode layer 61c and the NOx storage layer 62 are collectively referred to as “first electrode part”, and the anode electrode layer 61a is also referred to as “second electrode part”.

The NOx occlusion layer 62 occludes NO ₂ in the exhaust gas in the form of barium nitrate through the following chemical reaction (1) (see also FIG. 3A).
2BaO + 4NO ₂ + O ₂ → 2Ba (NO ₃ ) ₂ (1)

Furthermore, when NO in the exhaust gas is oxidized to NO ₂ through the following chemical reaction (2), it is stored in the NOx storage layer 62 in the form of barium nitrate through the chemical reaction (1).
2NO + O ₂ → 2NO ₂ (2)

  As shown in FIG. 2B, the power supply circuit 63 includes a DC power supply 64, a switch 65, and a current detector 74. In the reactor 55 according to the present embodiment, the DC power supply 64, the switch 65, the current detector 74, the cathode electrode layer 61c, the solid electrolyte layer 60, and the anode electrode layer 61a are connected in series from the upstream to the downstream of the electron flow. Thus, a circuit is formed. Therefore, as shown in FIG. 3B, when the switch 65 is closed, a voltage is applied between the anode electrode layer 61a and the cathode electrode layer 61c by the DC power source 64, that is, to the solid electrolyte layer 60. The switch 65 is on / off controlled by a CPU of an ECU 80 described later.

When a voltage is applied to the solid electrolyte layer 60, the NOx occluded in the NOx occlusion layer 62 is reduced and purified in the cathode electrode layer 61c through the following chemical reaction (3) (see also FIG. 3B). .
2Ba (NO ₃ ) ₂ + 4e ⁻ → 2BaO + 2O ² + + N ₂ + 2O ₂ (3)

Furthermore, NO _{2 that} has not been stored in the NOx storage layer 62 and has directly reached the cathode electrode layer 61c through the storage layer 62 is reduced to NO through the following chemical reaction (4) in the electrode layer 61c.
NO ₂ + 2e ⁻ → NO + O ²⁻ (4)

In addition, “NO generated through the chemical reaction (4)” and “NO that has not been stored in the NOx storage layer 62 and has directly reached the cathode electrode layer 61c through the storage layer 62” are stored in the electrode layer 61c. It is reduced and purified through the following chemical reaction (5).
2NO + 4e ⁻ → N ₂ + 2O ²⁻ (5)

In such a reduction and purification reaction of NOx (the above chemical reactions (3) and (5)), oxygen ions O ²⁻ are generated. The oxygen ions O ²⁻ move from the cathode electrode layer 61c through the solid electrolyte layer 60 to the anode electrode layer 61a. Oxygen ions O ²⁻ emit electrons e ⁻ to the anode electrode layer 61a in the anode electrode layer 61a and become oxygen molecules O ₂ . As a result, a current flows through the power supply circuit 63. The current detector 74 detects this current.

  Referring back to FIG. 1, the ECU 80 is an electronic circuit including a well-known microcomputer, and includes a CPU, a ROM, a RAM, a backup RAM, an interface, and the like. The ECU 80 is connected to sensors which will be described later, and receives signals from these sensors. Further, the ECU 80 sends instruction (drive) signals to various actuators and the switch 65 of the power supply circuit 63.

  The ECU 80 is connected to an air flow meter 71, a crank angle sensor 72, an accelerator opening sensor 73, and a current detector 74 of the power supply circuit 63 of the electrochemical reactor 55.

  The air flow meter 71 measures the mass flow rate (intake air amount) Ga of intake air passing through the intake passage, and outputs a signal representing the intake air amount Ga. Further, the air flow meter 71 detects the temperature of the intake air and outputs a signal indicating the temperature of the intake air.

  The crank angle sensor 72 outputs a signal corresponding to the rotational position (that is, crank angle) of a crankshaft (not shown) of the engine 10. The ECU 80 acquires the crank angle (absolute crank angle) of the engine 10 based on the compression top dead center of a predetermined cylinder based on signals from “this crank angle sensor 72 and a cam position sensor (not shown)”. Further, the ECU 80 acquires the engine speed based on the signal from the crank angle sensor 72.

  The accelerator opening sensor 73 detects the opening (accelerator pedal operation amount) of an accelerator pedal (not shown) and outputs a signal representing the accelerator pedal opening. As described above, the current detector 74 of the power supply circuit 63 of the electrochemical reactor 55 detects the current flowing through the solid electrolyte layer 60 of the reactor 55 and outputs a signal representing the value of the current.

  Next, an outline of the operation of the exhaust emission control device according to the present embodiment will be described with reference to FIG. When the NOx release condition for releasing NOx stored in the NSR catalyst 54 from the catalyst 54 (this condition will be described in detail later) is not satisfied (the period TL), the CPU of the ECU 80 The air-fuel ratio of the air-fuel mixture in the combustion chamber 22 is controlled to a lean air-fuel ratio (hereinafter simply referred to as “lean air-fuel ratio”) larger than the stoichiometric air-fuel ratio (lean control is executed). In this case, the lean air-fuel ratio exhaust gas containing NOx flows into the NSR catalyst 54, and NOx in the exhaust gas is occluded by the catalyst 54.

  The CPU operates in the engine operating state (in this example, the operating state of the engine 10 (in this example, “the amount of intake air Ga detected by the air flow meter 71”) and “the amount of fuel injected from the fuel injection valve 23 ( The amount of NOx discharged from the combustion chamber 22 per unit time (hereinafter referred to as “unit discharged NOx amount”) is estimated based on the fuel injection amount Qa ”etc. Further, the CPU By accumulating the estimated value of the unit exhaust NOx amount, the amount of NOx stored in the NSR catalyst 54 (hereinafter referred to as “NSR storage NOx amount”) is estimated, as shown in FIG. When the NOx release condition is not satisfied (period TL), the estimated value of the NSR occluded NOx amount (hereinafter referred to as “NSR occluded NOx estimated value”) NOXnsr-a gradually increases with time. Increase to .

  There is a limit to the amount of NOx that the NSR catalyst 54 can store. When the NSR occlusion NOx amount reaches “the limit value of the NOx amount that can be occluded by the catalyst 54 (hereinafter also referred to as“ NSR occlusion limit value ”), NOx that has flowed into the catalyst 54 is not occluded by the catalyst 54. In other words, the catalyst 54 flows out downstream.

  Therefore, when the NSR storage NOx estimated value NOXnsr-a reaches a value “NOXth1” (hereinafter referred to as “first determination value”) smaller than the NSR storage limit value by a predetermined value (time t0). , It is determined that the NOx release condition is satisfied. When the CPU determines that the NOx release condition is satisfied, the fuel injection amount is increased, so that the air-fuel ratio of the air-fuel mixture in the combustion chamber 22 is reduced to a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio (hereinafter simply “rich air-fuel ratio”). The engine 10 is operated as “fuel ratio”. That is, the CPU starts a so-called rich spike. As a result, the air-fuel ratio of the exhaust gas flowing into the NSR catalyst 54 (hereinafter referred to as “NSR inflow air-fuel ratio”) A / Fnsr decreases toward the rich air-fuel ratio. When a certain period (period T1) elapses from the start of the rich spike (time t1), the air-fuel ratio A / Fnsr becomes the rich air-fuel ratio.

  When the NSR inflow air-fuel ratio A / Fnsr becomes a rich air-fuel ratio, NOx occluded in the NSR catalyst 54 is released, and the released NOx is reduced and purified by unburned HC in the exhaust gas. Further, immediately after the inflow air-fuel ratio A / Fnsr becomes a rich air-fuel ratio, the “NOx reduction rate in the catalyst 54” is slower than the “NOx release rate from the catalyst 54”, so that the inflow air-fuel ratio A / Fnsr was released from the catalyst 54. Part of the NOx flows out of the catalyst 54 without being reduced and purified in the catalyst 54.

  The amount of NOx flowing out from the NSR catalyst 54 (hereinafter referred to as “NSR outflow NOx amount”) is t1 when the NSR inflow air-fuel ratio A / Fnsr becomes a rich air-fuel ratio, as shown in FIG. It increases at a stretch from. On the other hand, the NOx reduction rate gradually increases after the inflow air-fuel ratio A / Fnsr becomes a rich air-fuel ratio, and a certain period T2 elapses from the time t1 when the air-fuel ratio A / Fnsr becomes the rich air-fuel ratio. When this occurs (time t3), the NOx release rate becomes greater. Therefore, at this time point t3, the NSR outflow NOx amount starts to decrease, and thereafter becomes “0” when a certain period T3 has elapsed (time t5).

  During the rich spike (period TR), the CPU determines that the “from NSR catalyst 54” is based on the engine operating state (in this example, “NSR inflow air-fuel ratio A / Fnsr”, “flow rate of exhaust gas flowing into NSR catalyst 54”, etc.). The amount of NOx released per unit time (hereinafter referred to as “unit released NOx amount”) ”is estimated. The CPU newly estimates the NSR storage NOx amount by subtracting the estimated value of the unit release NOx amount from the current NSR storage NOx estimated value NOXnsr-a. During the rich spike, there is almost no NOx newly stored in the NSR catalyst 54. Therefore, after the start of the rich spike, the NSR occlusion NOx estimated value NOXnsr-a gradually decreases as shown in FIG.

  The CPU ends the rich spike when the NSR occlusion NOx estimated value NOXnsr-a reaches “a second determination value NOXth2 smaller than the first determination value NOXth1” (time t7). That is, the rich spike is a control for supplying the rich air-fuel ratio exhaust gas to the NSR catalyst 54 over a predetermined period TR from time t0 to time t7. The second determination value NOXth2 is “a value greater than“ 0 ”and close to“ 0 ””.

  Since the NSR inflow air-fuel ratio A / Fnsr is the rich air-fuel ratio from the end point t7 of the rich spike until the lapse of a certain period T4, the NOx reduction in the NSR catalyst 54 continues. Therefore, even if the rich spike is ended, the NSR occlusion NOx estimated value NOXnsr-a decreases until the predetermined period T4 elapses. Since the inflow air-fuel ratio A / Fnsr becomes a lean air-fuel ratio at the time point t8 when the predetermined period T4 has elapsed, the NSR occlusion NOx estimated value NOXnsr-a gradually increases after the simultaneous point t8.

  As described above, when the NSR inflow air-fuel ratio A / Fnsr becomes a rich air-fuel ratio, a part of NOx released from the NSR catalyst 54 flows out from the catalyst 54 downstream. These NOx flow into the electrochemical reactor 55. In order to reduce and purify these NOx by the reactor 55, the CPU applies a voltage between the cathode electrode layer 61c and the anode electrode layer 61a of the reactor 55 simultaneously with the start of the rich spike (time t0). In this example, the voltage (hereinafter referred to as “reactor applied voltage” or simply “applied voltage”) Vrct applied between the electrode layers 61c and 61a is maintained at a constant voltage Vc.

  The constant voltage Vc is “when all the NOx occluded in the NOx occlusion layer 62 of the reactor 55 is reached before NOx starts to flow into the reactor 55 when voltage application is started simultaneously with the start of the rich spike. The voltage is set so low that it does not reduce.

  There is almost no NOx flowing out of the NSR catalyst 54 from the voltage application start time t0 to the time t1 when the NSR inflow air-fuel ratio A / Fnsr becomes the rich air-fuel ratio. Therefore, at the start point t0 of voltage application, the NOx stored in the NOx storage layer 62 of the reactor 55 is electrochemically reduced and purified by the reactor 55. At this time, the CPU is based on “the value of the current flowing through the solid electrolyte layer 60 (hereinafter referred to as“ reactor current value ”or simply“ current value ”) Irct” and “reactor applied voltage Vrct (= Vc)”. Then, the impedance value Zrct of the solid electrolyte layer 60 is calculated and stored as the initial value Zstart.

  An NSR catalyst 54 is disposed upstream of the reactor 55. Therefore, when the rich spike is not executed (period TL), NOx discharged from the engine 10 is occluded in the NSR catalyst 54, so that almost no NOx flows into the reactor 55. On the other hand, according to the “voltage application to the reactor 55” according to the present embodiment, as described below, the amount of NOx occluded in the NOx occlusion layer 62 of the reactor 55 at the end of voltage application (hereinafter referred to as “the voltage application to the reactor 55”). This is referred to as “reactor occlusion NOx amount.”) Voltage application is performed so that NOXrct-a becomes “a value greater than 0 but very small”.

  Therefore, the reactor occluded NOx amount NOXrct-a at the voltage application start time t0 is very small. Therefore, the reactor current value Irct at the simultaneous point t0 is small, and as a result, the initial value Zstart becomes a large value.

It should be noted that the air-fuel ratio (reactor inflow air-fuel ratio) A / Frct of the exhaust gas flowing into the reactor 55 is a lean air-fuel ratio at the voltage application start time t0. For this reason, a large amount of oxygen exists in the reactor 55. Therefore, these oxygens receive electrons e ⁻ in the cathode electrode layer 61 c and become oxygen ions O ²⁻ , and these oxygen ions O ²⁻ may move in the solid electrolyte layer 60. In this case, the reactor current at the voltage application start time t0 includes a current resulting from the reduction of oxygen. However, NOx (reactor storage NOx) stored in the NOx storage layer 62 is present closer to the cathode electrode layer 61c than the oxygen. For this reason, in the cathode electrode layer 61c, the reactor occluded NOx is preferentially reduced. Therefore, most of the reactor current at the start of voltage application is a current resulting from the reduction of the NOx stored in the reactor.

  As described above, there is almost no NOx flowing into the reactor 55 in the period T1 from the voltage application start time t0 to the time t1 when a certain period has elapsed. Therefore, during the same period T1, only reduction and purification of NOx (reactor storage NOx) stored in the NOx storage layer 62 is performed. As a result, the reactor storage NOx amount NOXrct-a decreases. Accordingly, the reactor current value Irct also decreases, and as a result, the reactor impedance Zrct increases.

  At a time t1 when a certain period T1 has elapsed from the voltage application start time t0, NOx begins to flow into the reactor 55, and this NOx amount (hereinafter referred to as “reactor inflow NOx amount”) NOXrct-i gradually increases. Increase. As the inflow NOx amount NOXrct-i increases, the amount of NOx reduced and purified by the reactor 55 increases, and the reactor current value Irct also increases. As a result, the reactor impedance Zrct decreases.

  On the other hand, after the time t1, the inflow until the reactor inflow NOx amount NOXrct-i exceeds the “maximum NOx amount that can be reduced by the cathode electrode layer 61c per unit time (maximum NOx reduction amount) NOXmax”. Even if the NOx amount NOXrct-i increases, most of the NOx is reduced and purified by the cathode electrode layer 61c. Accordingly, the reactor occluded NOx amount NOXrct-a increases after time t1, but this increase is very slight. That is, the stored NOx amount NOXrct-a hardly changes.

  Thereafter, when the reactor inflow NOx amount NOXrct-i exceeds the maximum NOx reduction amount NOXmax (time t2), the reactor 55 can reduce and purify only NOx of the maximum NOx reduction amount NOXmax. Therefore, the current value Irct is constant at a current value corresponding to the maximum NOx reduction amount. As a result, the impedance Zrct is also constant. Of course, NOx that is not reduced and purified at this time is stored in the NOx storage layer 62.

  On the other hand, after the time t2, until the reactor inflow NOx amount NOXrct-i reaches a peak at time t3, the NOx amount occluded in the NOx occlusion layer 62 increases at a stretch, so the reactor occluded NOx amount NOXrct-a becomes It increases at a stretch. Thereafter, even if the inflow NOx amount NOXrct-i starts to decrease from the peak, the reactor 55 does not stop until the inflow NOx amount NOXrct-i decreases to the maximum NOx reduction amount NOXmax (from time t3 to time t4). Since all NOx flowing into the reactor cannot be reduced and purified, the NOx amount NOXrct-a stored in the reactor gradually increases.

  After the inflow NOx amount NOXrct-i increases and reaches the maximum NOx reduction amount NOXmax (time t2), until the inflow NOx amount NOXrct-i decreases and reaches the maximum NOx reduction amount NOXmax (time t4). The reactor current value Irct is substantially constant, and as a result, the reactor impedance Zrct is also substantially constant.

  After the inflow NOx amount NOXrct-i reaches the maximum NOx reduction amount NOXmax at time t4, the NOx amount reduced in the cathode electrode layer 61c gradually decreases. For this reason, the reactor current value Irct gradually decreases, and as a result, the reactor impedance Zrct gradually increases.

  Although the inflow NOx amount NOXrct-i becomes “0” at time t5, NOx is occluded in the NOx occlusion layer 62, and therefore the NOx reduction reaction in the cathode electrode layer 61c is continued. Therefore, at time t5, the reactor current value Irct does not become “0”, and gradually decreases as the reactor NOx occlusion amount NOXrct-a decreases after time t5. For this reason, the reactor impedance Zrct gradually increases after time t5. Thereafter, when the reactor impedance Zrct reaches the initial value Zstart (time t6), the CPU stops the voltage application.

  After the NOx amount around the cathode electrode layer 61c (that is, the reactor storage NOx amount NOXrct-a) becomes “0”, voltage application is continued in a state where the reactor inflow air-fuel ratio A / Frct is a rich air-fuel ratio. Then, oxygen atoms in the material constituting the solid electrolyte layer 60 (that is, the solid electrolyte) are ionized and move toward the anode electrode layer 61a. That is, chemical decomposition of the solid electrolyte layer 60 occurs.

  On the other hand, as described above, the initial value Zstart is a value corresponding to “a very small reactor storage NOx amount NOXrct-a”. Accordingly, when the voltage application is stopped when the reactor impedance Zrct reaches the initial value Zstart (time t6), before the NOx amount around the cathode electrode layer 61c becomes “0”, that is, the solid state of the solid electrolyte layer 60. The voltage application is stopped immediately before the start of electrolyte decomposition. Therefore, according to the exhaust gas purification apparatus according to the present embodiment, decomposition of the solid electrolyte layer 60 is avoided.

  It is considered that the solid electrolyte layer 60 is likely to be decomposed as the amount of NOx around the cathode electrode layer 61c decreases. However, since the NOx reduction reaction is more likely to occur than the decomposition reaction of the solid electrolyte layer 60, it is assumed that the decomposition of the solid electrolyte layer 60 does not start until the amount of NOx around the cathode electrode layer 61c becomes “0”. Is done.

  In the above-described embodiment, a voltage may be applied to the solid electrolyte layer 60 so that the current flowing through the solid electrolyte layer 60 is constant. In this case, the CPU controls the voltage applied to the solid electrolyte layer 60 so that the current flowing through the solid electrolyte layer 60 is constant. In this case, the CPU measures the impedance based on the voltage value and the current value.

  Next, the actual operation of the exhaust emission control device according to this embodiment will be described. The CPU of the ECU 80 repeatedly executes the “rich spike routine” shown in FIG. 5 every elapse of a predetermined time. Therefore, when the predetermined timing is reached, the CPU starts the process from step 50 in FIG. 5 and determines in step 51 whether or not the NSR occlusion NOx estimated value NOXnsr-a is greater than or equal to the first determination value NOXth1.

  The NSR occlusion NOx estimated value NOXnsr-a is calculated by an NSR occlusion NOx estimated value calculation routine (not shown). In this calculation routine, during the lean control, the CPU determines “the previous time of this calculation routine based on the intake air amount Ga and the fuel injection amount Qa (load of the engine 10) and the lookup table Map (Ga, Qa). The amount of NOx discharged from the engine 10 between the start time of this time and the current start time (discharge NOx amount) is calculated. Next, the CPU calculates the latest estimated NSR storage NOx value NOXnsr-a by adding the calculated exhausted NOx amount to the “NSR storage NOx estimated value NOXnsr-a at that time”.

  On the other hand, during the rich spike, the CPU calculates “this calculation based on the flow rate SVrct of the exhaust gas flowing into the NSR catalyst 54, the air-fuel ratio A / Frct of the exhaust gas, and the lookup table Map (SVrct, A / Frct). The amount of NOx released from the NSR catalyst 54 (the amount of released NOx) between the previous start time of the routine and the current start time is calculated. Next, the CPU calculates the latest estimated NSR storage NOx value NOXnsr-a by subtracting the calculated released NOx amount from the “NSR storage NOx estimated value NOXnsr-a at that time”.

  If the NSR occlusion NOx estimated value NOXnsr-a is not equal to or greater than the first determination value NOXth1, the CPU makes a “No” determination at step 51 to proceed to step 55 where the air-fuel ratio of the air-fuel mixture in the combustion chamber 22 of the engine 10 is determined. Is controlled to a lean air-fuel ratio larger than the stoichiometric air-fuel ratio, the lean control for supplying the NSR catalyst 54 with the exhaust gas having the same lean air-fuel ratio is executed, and this routine is temporarily terminated.

  On the other hand, if the estimated value NOXnsr-a is greater than or equal to the first determination value NOXth1, the CPU makes a “Yes” determination at step 51 to proceed to step 52 to start a rich spike (mixing in the combustion chamber 22 of the engine 10). The air / fuel ratio of the air is set to a rich air / fuel ratio smaller than the stoichiometric air / fuel ratio) and “1” is input to the rich spike flag Frs (the flag Frs is set). Thereafter, the CPU proceeds to step 53 to determine whether or not the NSR storage NOx estimated value NOXnsr-a is equal to or smaller than the second determination value NOXth2.

  If the NSR occlusion NOx estimated value NOXnsr-a is not less than or equal to the second determination value NOXth2, the CPU makes a “No” determination at step 53 to return to step 53. On the other hand, if the estimated value NOXnsr-a is equal to or smaller than the second determination value NOXth2, the CPU makes a “Yes” determination at step 53 to proceed to step 54 to end the rich spike and set the rich spike flag Frs to “0”. "(Flag Frs is reset), and this routine is terminated once.

  According to this, NOx discharged from the engine 10 can be reduced and purified without causing the NSR storage NOx amount to reach its limit value.

  Further, the CPU repeatedly executes the “reactor voltage application routine” shown in FIG. 6 every elapse of a predetermined time. Therefore, when the predetermined timing is reached, the CPU starts the process from step 60 in FIG. 6 and determines in step 61 whether or not the rich spike flag Frs is “1”. The flag Frs is a flag that is set at step 52 in FIG. 5 (Frs ← 1) and reset at step 54 in FIG. 5 (Frs ← 0).

  When the flag Frs is not “1” (that is, when the flag Frs is “0”), the CPU makes a “No” determination at step 61 to end the present routine tentatively. On the other hand, when the flag Frs is “1”, the CPU makes a “Yes” determination at step 61 to proceed to step 62 to start voltage application to the solid electrolyte layer 60. Thereafter, the CPU proceeds to step 63, acquires the reactor impedance Zrct at that time as “initial value Zstart”, and proceeds to step 64. If voltage application has already started at the time when the routine proceeds to step 62, the CPU proceeds to step 64 without performing processing in step 62 and step 63.

  In step 64, the CPU determines whether or not the reactor impedance Zrct is increasing after it has once decreased to a peak value (see period Ti in FIG. 4). In the case where the impedance Zrct is increasing and then increasing after reaching the peak value, the CPU makes a “Yes” determination at step 64 to proceed to step 65. On the other hand, if the impedance Zrct is not increasing while it once decreases and reaches the peak value, the CPU makes a “No” determination at step 64 to end the present routine tentatively.

  In step 65, the CPU determines whether or not the reactor impedance Zrct is greater than or equal to the initial value Zstart acquired in step 63. If the impedance Zrct is equal to or greater than the initial value Zstart, the CPU makes a “Yes” determination at step 65 to proceed to step 66 to stop the voltage application to the solid electrolyte layer 60 and end this routine once. On the other hand, if the impedance Zrct is not equal to or greater than the initial value Zstart, the CPU makes a “No” determination at step 65 to end the present routine tentatively.

  According to this, NOx flowing out from the NSR catalyst 54 during the rich spike can be reduced and purified while avoiding decomposition of the solid electrolyte layer 60.

  Further, the exhaust purification apparatus according to the above-described embodiment is arranged such that a second catalyst capable of reducing and purifying NOx with a reducing agent is disposed at a position downstream of the NSR catalyst 54 in the exhaust passage. This is also advantageous compared to the case where NOx flowing out from the NSR catalyst 54 is reduced and purified.

  More specifically, for example, “the same catalyst as the NSR catalyst 54” can be used as the second catalyst. However, when such a second catalyst is used, it is necessary to supply a reducing agent to the second catalyst when releasing NOx stored in the second catalyst, and the second catalyst It is also necessary to maintain the temperature above its activation temperature. Alternatively, a so-called SCR catalyst capable of reducing and purifying NOx using ammonia as a reducing agent can be used as the second catalyst. However, when NOx flowing out from the NSR catalyst 54 is reduced by such a second catalyst, it is necessary to supply a reducing agent (ammonia) to the second catalyst, and the temperature of the second catalyst is set to the second catalyst. It is also necessary to maintain above the activation temperature. In addition, even if the supply of the reducing agent to the second catalyst is started simultaneously with the start of the NOx release control, the NOx reduction of the second catalyst is started when NOx starts to flow into the second catalyst. It is possible that the ability is not high enough.

  On the other hand, the electrochemical reactor 55 only needs to apply a voltage, and further exhibits a reduction function simultaneously with the application of the voltage. Therefore, the above-described necessity does not occur in the electrochemical reactor 55. For these reasons, the exhaust purification apparatus according to this embodiment is more advantageous than the case where NOx flowing out from the NSR catalyst 54 is reduced and purified by the second catalyst. However, the present invention does not exclude placing another catalyst in addition to the electrochemical reactor 55 downstream of the NSR catalyst 54.

  By the way, when determining the time point at which the voltage application to the electrochemical reactor 55 is stopped, a fixed threshold value is prepared in advance for the impedance, and when the impedance reaches the fixed threshold value, the voltage application to the reactor 55 is performed. It is also possible to configure the exhaust emission control device so as to stop (see, for example, Patent Document 2).

  However, the impedance value corresponding to the time immediately before the start of the decomposition of the solid electrolyte is, for example, “Aging change of NOx purification characteristics of the electrochemical reactor 55” and “Ambient environment of the electrochemical reactor that changes with each voltage application”. May change due to impact. Therefore, the time when the impedance value reaches the fixed threshold prepared in advance may be later than the time when the decomposition of the solid electrolyte starts. In this case, the deterioration of the solid electrolyte proceeds.

  On the other hand, the impedance value measured during voltage application is a value affected by the NOx purification characteristics of the electrochemical reactor 55 and the surrounding environment of the electrochemical reactor 55 at that time. For this reason, as in the exhaust emission control device according to the present embodiment, if the “time immediately before the decomposition of the solid electrolyte” is detected based on the behavior of the impedance measured during voltage application, the solid electrolyte is more accurately detected. It is possible to detect a time point just before the decomposition starts.

  As described above, the exhaust gas purification apparatus according to this embodiment includes the NOx storage reduction catalyst (NSR catalyst) 54, the electrochemical reactor 55, and the first electrode portion (the cathode electrode layer 61c and the NOx storage) of the reactor 55. A voltage application unit (power supply circuit 63) capable of applying a voltage between the layer 62) and the second electrode unit (anode electrode layer 61a).

  Furthermore, the exhaust purification apparatus according to the present embodiment does not satisfy the release condition (see step 51 in FIG. 5) for releasing NOx stored in the catalyst from the catalyst (period in FIG. 4). TL and see the case of “No” in step 51 of FIG. 5), by controlling the air-fuel ratio of the air-fuel mixture in the combustion chamber 22 of the engine 10 to a lean air-fuel ratio that is larger than the stoichiometric air-fuel ratio. When the lean control for supplying the exhaust gas with the same lean air-fuel ratio (see step 55 in FIG. 5) is executed and the release condition is satisfied (“Yes” at time t0 in FIG. 4 and step 51 in FIG. 5). (See the case.) NOx release control is performed in which exhaust gas having a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio is supplied to the catalyst over a predetermined period (see period TR in FIG. 4) (step 52 in FIG. 5 is performed). Refer to.) First control Equipped with a.

  In addition, the exhaust gas purification apparatus according to the present embodiment includes an impedance measurement unit (current detection unit) that measures an impedance Zrct of the solid electrolyte unit (solid electrolyte layer) 60 between the first electrode unit and the second electrode unit. And the application of the voltage using the voltage application unit at the start of the NOx release control (see time t0 in FIG. 4 and step 62 in FIG. 6), and then the measurement. Impedance (see step 63 in FIG. 6) Based on Zstart, a time immediately before the start of decomposition of the solid electrolyte is detected and detected (at time t6 in FIG. 4 and step 65 in FIG. 6). In the case of “Yes”, the application of the voltage using the voltage application unit is stopped (see time t6 in FIG. 4 and step 66 in FIG. 6). Provided.

  Next, an exhaust purification device according to a modification of the above-described embodiment will be described. The exhaust emission control device according to the above-described embodiment uses the reactor impedance Zrct (initial value Zstart) at the start of voltage application to the reactor 55 as an index for determining when to stop the reactor application voltage Vrct. However, the exhaust emission control device according to this modification uses the increase rate Rzrct of the reactor impedance Zrct as the index.

  More specifically, as shown in FIG. 7, the reactor inflow NOx amount NOXrct-i gradually decreases and reaches the maximum NOx reduction amount NOXmax after time t3 (time t4). As described above, the reactor occluded NOx amount NOXrct-a gradually decreases thereafter. Thus, while the occluded NOx amount NOXrct-a decreases, the reactor current value Irct also gradually decreases, and as a result, the reactor impedance Zrct gradually increases. At this time, the reactor current value Irct decreases relatively slowly until the reactor occluded NOx amount NOXrct-a is very close to “0” (until time t6), and the rate of decrease is substantially constant. Accordingly, the reactor impedance Zrct increases relatively slowly, and the increase rate Rzrct is substantially constant.

  However, when the reactor occluded NOx amount NOXrct-a becomes very close to “0” (time t6), the reactor current value Irct starts to decrease rapidly as indicated by reference symbol A in FIG. Accordingly, the reactor impedance Zrct starts to increase rapidly as indicated by reference numeral B in FIG. That is, the increase rate Rzrct of the impedance Zrct increases rapidly.

  The exhaust emission control apparatus according to this modification uses such a change in the increase rate Rzrct of the impedance Zrct. That is, the exhaust gas purification device stops the voltage application to the reactor 55 when the increase rate Rzrct (time differential value dZrct / dt of the impedance Zrct) of the reactor impedance Zrct becomes equal to or greater than a predetermined threshold (time t6).

  Next, the actual operation of the exhaust emission control device according to this modification will be described. Instead of the “reactor voltage application routine” shown in FIG. 6, the CPU repeatedly executes the “reactor voltage application routine” shown in FIG. 8 every elapse of a predetermined time. Therefore, when the predetermined timing comes, the CPU starts the process from step 80 in FIG. 8 and determines whether or not the rich spike flag Frs is “1” in step 81. The flag Frs is a flag that is set at step 52 in FIG. 5 (Frs ← 1) and reset at step 54 in FIG. 5 (Frs ← 0).

  When the flag Frs is not “1” (that is, when the flag Frs is “0”), the CPU makes a “No” determination at step 81 to end the present routine tentatively. On the other hand, if the flag Frs is “1”, the CPU makes a “Yes” determination at step 81 to proceed to step 82, starts applying a voltage to the solid electrolyte layer 60, and proceeds to step 83. If voltage application has already started when the routine proceeds to step 82, the CPU proceeds to step 83 without performing any processing in step 82.

  In step 83, the CPU determines whether or not the reactor impedance Zrct is increasing after it once decreases and reaches a peak value (see period Ti in FIG. 7). If the impedance Zrct is not increasing while it once decreases and reaches the peak value, the CPU makes a “No” determination at step 83 to end the present routine tentatively. On the other hand, if the impedance Zrct is decreasing and once increasing and then increasing, the CPU makes a “Yes” determination at step 83 to proceed to step 84 where the reactor impedance Zrct at that time The rate of increase Rzrct is obtained.

  Next, the CPU proceeds to step 85 to determine whether or not the increase rate Rzrct of the reactor impedance Zrct acquired in step 84 is equal to or greater than a threshold value Rth. If the increase rate Rzrct is greater than or equal to the threshold value Rth, the CPU makes a “Yes” determination at step 85 to proceed to step 86 to stop the voltage application to the solid electrolyte layer 60 and end this routine once. On the other hand, if the increase rate Rzrct is not greater than or equal to the threshold value Rth, the CPU makes a “No” determination at step 85 to end the present routine tentatively.

  According to this, NOx flowing out from the NSR catalyst 54 during the rich spike can be reduced and purified while avoiding decomposition of the solid electrolyte layer 60.

  As described above, the exhaust gas purification apparatus according to this modification example determines whether or not to stop voltage application to the reactor 55 depending on whether or not the increase rate Rzrct of the reactor impedance Zrct is equal to or greater than the predetermined threshold value Rth. To decide. However, instead of this, when the exhaust gas purification device detects the increase rate Rzrct, the difference (impedance) between the detected increase rate Rzrct and the “increase rate Rzrct detected immediately before the detection point of the same increase rate Rzrct” When the second-order differential value of Zrct) is equal to or greater than a predetermined value, the voltage application to the reactor 55 can be stopped.

  Note that the exhaust emission control device according to the above-described embodiment can also be applied to a gasoline engine. Furthermore, in the exhaust emission control device according to the above-described embodiment, the DPF 56 may not be provided.

  Furthermore, in the exhaust emission control device according to the above-described embodiment, the voltage application to the solid electrolyte layer 60 is started simultaneously with the start of the rich spike, but the voltage application to the solid electrolyte layer 60 is started after the start of the rich spike. May be. In this case, the voltage application to the solid electrolyte layer 60 is started after the start of the rich spike and until the time when NOx flowing out of the NSR catalyst 54 starts flowing into the electrochemical reactor 55 due to the rich spike. It is preferable. According to this, NOx flowing out from the NSR catalyst 54 can be reduced and purified in the electrochemical reactor 55 more reliably. In addition, since the voltage application time can be shortened compared with the case where the voltage application is started simultaneously with the start of the rich spike, the power consumption can be reduced.

  In addition, in the above-described implementation device, a sensor (NOx sensor) capable of detecting the NOx concentration is disposed in the exhaust passage between the NSR catalyst 54 and the voltage application reactor 55, and “the NOx concentration detected by the sensor is detected. Application of voltage to the reactor 55 is started when it becomes larger than “0” (at the start of NOx release control) or “when the NOx concentration exceeds a predetermined threshold (after the start of NOx release control)”. You may make it do.

  DESCRIPTION OF SYMBOLS 10 ... Internal combustion engine, 54 ... NOx occlusion reduction catalyst (NSR catalyst), 55 ... Electrochemical reactor, 60 ... Solid electrolyte layer, 61a ... Anode electrode layer, 61c ... Cathode electrode layer, 62 ... NOx occlusion layer, 64 ... DC power supply , 65 ... switch, 74 ... current detector, 80 ... ECU

Claims (3)

An exhaust purification device for an internal combustion engine,
A NOx storage reduction catalyst disposed in the exhaust passage of the engine;
An electrochemical reactor disposed at a position downstream of the catalyst in the exhaust passage, comprising a solid electrolyte portion made of a solid electrolyte having oxygen ion conductivity, disposed on a surface of the solid electrolyte portion, and storing NOx A first electrode part including a NOx occlusion part and a second electrode part disposed on the surface of the solid electrolyte part, and a voltage is applied between the first electrode part and the second electrode part. An electrochemical reactor that electrochemically reduces and purifies NOx at the first electrode part,
A voltage application unit capable of applying a voltage between the first electrode unit and the second electrode unit;
By controlling the air-fuel ratio of the air-fuel mixture in the combustion chamber of the engine to a lean air-fuel ratio larger than the stoichiometric air-fuel ratio when a release condition for releasing NOx stored in the catalyst from the catalyst is not satisfied. Performing lean control for supplying exhaust gas with the same lean air-fuel ratio to the catalyst, and NOx release control for supplying exhaust gas with a rich air-fuel ratio smaller than the stoichiometric air-fuel ratio to the catalyst over a predetermined period when the release condition is satisfied A first control unit to be executed;
An impedance measuring unit for measuring the impedance of the solid electrolyte part between the first electrode part and the second electrode part;
Application of the voltage using the voltage application unit is started at the start of the NOx release control or after the start of the NOx release control, and then immediately before the decomposition of the solid electrolyte is started based on the measured impedance. A second control unit that detects a time point and stops application of the voltage using the voltage application unit at the time point when the time point is detected;
Exhaust gas purification device. The exhaust emission control device according to claim 1,
The second controller is
Obtaining the impedance measured by the impedance measurement unit as an initial value when application of the voltage using the voltage application unit is started, and then increasing after the impedance measured by the impedance measurement unit decreases Configured to detect a time point when the initial value is reached as a time point immediately before the decomposition of the solid electrolyte starts,
Exhaust purification device. The exhaust emission control device according to claim 1 or 2,
The second controller is
The application of the voltage is started after the start of the NOx release control and until the time when NOx flowing out of the catalyst by the NOx release control starts to flow into the electrochemical reactor,
Exhaust purification device.

Images