Radio frequency device test structure

Abstract

The invention relates to the field of semiconductor radio frequency devices, and discloses a radio frequency device test structure which comprises a first grounding bonding pad, a first signal bonding pad and a second grounding bonding pad which form a first port. The third grounding bonding pad, the second signal bonding pad and the fourth grounding bonding pad form a second port; the dielectric layer of at least one of an area between the first grounding pad and the first signal pad, an area between the second grounding pad and the first signal pad, an area between the third grounding pad and the second signal pad, and an area between the fourth grounding pad and the second signal pad is provided with a groove. The radio frequency device test structure can improve the measurable frequency range of the S parameter of the radio frequency device.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor radio frequency, and particularly to a radio frequency test structure for S-parameter measurement. Background Art

[0002] In the measurement of radio frequency semiconductor devices, it is very necessary to accurately obtain the S-parameters of the device under test (DUT) for both device modeling and device characteristic verification. The S-parameters of the DUT are generally obtained in two steps: First, the GSG (ground-signal-ground) pad test structure is used to measure the DUT, and then the parasitic capacitance and inductance of the pad and transmission line are removed through de-embedding technology to calculate the S-parameters of the DUT. Therefore, the GSG pad test structure is crucial for obtaining the S-parameters of the DUT. In the GSG pad test structure, the open-thru method is a commonly used de-embedding means (CN10716772A discloses a de-embedding method for small-signal measurement).

[0003] Figure 1 、 Figure 4 is a traditional GSG pad test structure and a thru calibration component (GSG-thru, Figure 2 ) and an open calibration component (GSG-open, Figure 3 ) structure of a radio frequency device. It generally includes 6 layers in structure, namely a back gold layer (20), a GaAs substrate layer 10, and a dielectric layer 30. The GSG-thru structure consists of 4 Ground pads (G1, G2, G3, G4), 2 Signal pads (S1, S2), 1 transmission line (L), and 4 metal back holes (TV). The GSG-open structure consists of 4 Ground pads (G1, G2, G3, G4), 2 Signal pads (S1, S2), and 4 metal back holes (TV). For GaAs devices applied below 100 GHz, the thickness of the GaAs substrate in the industry is generally set to a typical value of 100 μm. This traditional GSG-thru test structure has problems of large reflection coefficient (S11) and small transmission coefficient (S21) under high-frequency conditions. For example, when the frequency is greater than 76 GHz, its S11 is greater than -10 dB. And the traditional GSG-open test structure has problems of decreasing reflection coefficient (S11), increasing electromagnetic leakage, and increasing transmission coefficient (S21) in a high-frequency range. For example, when the frequency is greater than 92 GHz, its S11 decreases to below -1 dB, resulting in the inability to measure the S-parameters of the DUT at frequencies above 76 GHz through this GSG pad test structure, which limits the upper limit of the measurable frequency of the device.

[0004] Regarding the above problems, there are usually two existing solutions: one is to redesign the GSG-thru structure to solve the impedance matching problem in the high-frequency band. However, under the premise of the limitation of the GSG probe pitch (usually 150μm or 100μm), the design difficulty is relatively large. The other is to use the method of thinning the substrate to increase the upper limit of the operating frequency of the semiconductor device and the GSG pad test structure. For example, the industry usually reduces the thickness of the GaAs substrate from 100μm to 50μm to meet the application requirements of 100GHz. This method will cause many problems in the actual manufacturing process: the substrate is extremely prone to warping and cracking due to the reduction of the tensile strength of the substrate; it is extremely easy to cause metal and chemical reagent contamination to GaAs during the grinding process, etc. (CN102543665A - A rapid thinning method for improving gallium arsenide substrate, CN102543665B - A rapid thinning method for improving gallium arsenide substrate), resulting in a reduction in the substrate yield rate, and an obvious increase in the manufacturing difficulty and cost. Summary of the Invention

[0005] In order to overcome the defects of the prior art, the technical problem to be solved by the present invention is to provide a radio frequency device test structure.

[0006] The specific implementation manner of the solution of the present invention is as follows: A radio frequency device test structure includes a semiconductor base layer and first to fourth ground pads (G1 - G4), a first signal pad (S1), and a second signal pad (S2) provided on the semiconductor base layer (10). The first ground pad (G1), the first signal pad (S1), and the second ground pad (G2) form port one; the third ground pad (G3), the second signal pad (S2), and the fourth ground pad (G4) form port two; the first to fourth ground pads (G1 - G4) are connected; the device under test (DUT) is arranged between the first signal pad (S1) and the second signal pad (S2). The input end of the first signal pad (S1) and the DUT is connected by a wiring metal, and the signal pad (S2) and the output end of the DUT are connected by a wiring metal; at least one region of the dielectric layer (30) in the region between the first ground pad (G1) and the first signal pad (S1), the region between the second ground pad (G2) and the first signal pad (S1), the region between the third ground pad (G3) and the second signal pad (S2), and the region between the fourth ground pad (G4) and the second signal pad (S2) is provided with a groove (50). Further, preferably, the device under test (DUT) is arranged at the middle position of the first to fourth ground pads (G1 - G4), the first signal pad (S1), and the second signal pad (S2).

[0007] Further preferably, in the above test structure, a dielectric layer (30) is provided between the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2).

[0008] Further preferably, in the above test structure, a groove (50) is provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2). A groove (50) is provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2).

[0009] In an embodiment of the present invention, in the above test structure, optionally, the projections of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) form a 2×2 array, and chamfer angles (90) are respectively formed at positions close to the center of the array for the first to fourth ground pads (G1 to G4).

[0010] Furthermore, in the above test structure, the dielectric layer is single-layer or multi-layer; the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively composed of single-layer or multi-layer metal, the thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively greater than the thickness of the dielectric layer, and the thicknesses of the wiring metals are respectively less than or equal to the thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2).

[0011] Furthermore, the above test structure further includes a back gold layer (20). The projected areas of the first to fourth ground pads (G1 to G4) on the semiconductor substrate layer (10) are all larger than the projected areas of the first signal pad (S1) or the second signal pad (S2) on the semiconductor substrate layer 10. The first to fourth ground pads (G1 to G4) are respectively connected to the back gold layer (20) through metal vias (40); the groove is filled with air or a metal material.

[0012] Furthermore, in the above test structure, the number of grooves (50) provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is single or multiple; the projected pattern of the groove on the semiconductor substrate layer (10) is circular, rectangular or oval; the depth of the groove is greater than or equal to the thickness of the dielectric layer.

[0013] The present invention also discloses a radio frequency device test structure, including a semiconductor substrate layer and the first to fourth ground pads (G1 to G4), a first signal pad (S1), and a second signal pad (S2) provided on the semiconductor substrate layer (10). The first ground pad (G1), the first signal pad (S1), and the second ground pad (G2) form port one; the third ground pad (G3), the second signal pad (S2), and the fourth ground pad (G4) form port two; the first to fourth ground pads (G1 to G4) are connected; the first signal pad (S1) and the second signal pad (S2) are connected by a wiring metal; a dielectric layer (30) is provided between the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2); at least one area of the dielectric layer (30) in the area between the first ground pad (G1) and the first signal pad (S1), the area between the second ground pad (G2) and the first signal pad (S1), the area between the third ground pad (G3) and the second signal pad (S2), and the area between the fourth ground pad (G4) and the second signal pad (S2) is provided with a groove (50).

[0014] Further preferably, in the above test structure, a dielectric layer (30) is provided between the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1); a groove (50) is provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2).

[0015] In an embodiment of the present invention, in the above test structure, optionally, the projections of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) form a 2×2 array, and chamfer angles (90) are respectively formed at positions of the first to fourth ground pads (G1 to G4) close to the center of the array.

[0016] Furthermore, in the test structure, the dielectric layer is single-layer or multi-layer; the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively composed of single-layer or multi-layer metal, the thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively greater than the thickness of the dielectric layer, and the thicknesses of the wiring metals are respectively less than or equal to the thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2).

[0017] Furthermore, the above test structure further includes a back gold layer (20), the projection areas of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) are all greater than the projection areas of the first signal pad (S1) or the second signal pad (S2) on the semiconductor base layer 10, the first to fourth ground pads (G1 to G4) are respectively connected to the back gold layer (20) through metal vias (40); the grooves are filled with air or filled with metal materials.

[0018] Further, in the above test structure, the number of grooves (50) provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is one or more; the number of grooves (50) provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is one or more; the number of grooves (50) provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is one or more; the number of grooves (50) provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is one or more; the projected pattern of the grooves on the semiconductor substrate layer (10) is circular, rectangular or elliptical; the depth of the grooves is greater than or equal to the thickness of the dielectric layer.

[0019] The present invention also discloses a radio frequency device test structure, including a semiconductor substrate layer and first to fourth ground pads (G1-G4), a first signal pad (S1), and a second signal pad (S2) provided on the semiconductor substrate layer (10). The first ground pad (G1), the first signal pad (S1), and the second ground pad (G2) form port one; the third ground pad (G3), the second signal pad (S2), and the fourth ground pad (G4) form port two; the first to fourth ground pads (G1-G4) are connected; a dielectric layer (30) is provided between the first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2); at least one region of the dielectric layer (30) in the regions between the first ground pad (G1) and the first signal pad (S1), between the second ground pad (G2) and the first signal pad (S1), between the third ground pad (G3) and the second signal pad (S2), and between the fourth ground pad (G4) and the second signal pad (S2) is provided with grooves (50).

[0020] More preferably, in the above test structure, a dielectric layer (30) is provided between the first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2); the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is provided with grooves (50); the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is provided with grooves (50); the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is provided with grooves (50); the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is provided with grooves (50).

[0021] In an embodiment of the present invention, for the above test structure, the projections of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) form a 2×2 array, and chamfer angles (90) are respectively formed at positions close to the center of the array for the first to fourth ground pads (G1 to G4).

[0022] Furthermore, for the above test structure, the dielectric layer is single-layer or multi-layer; the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively composed of single-layer or multi-layer metal. The thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively greater than the thickness of the dielectric layer, and the thicknesses of the wiring metals are respectively less than or equal to the thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2).

[0023] Furthermore, the above test structure further includes a back gold layer (20). The projected areas of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) are all larger than the projected areas of the first signal pad (S1) or the second signal pad (S2) on the semiconductor base layer 10. The first to fourth ground pads (G1 to G4) are respectively connected to the back gold layer (20) through metal vias (40); the grooves are filled with air or metal materials.

[0024] Furthermore, for the above test structure, the number of grooves (50) provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is single or multiple; the projected pattern of the grooves on the semiconductor base layer (10) is circular, rectangular, or elliptical; the depth of the grooves is greater than or equal to the thickness of the dielectric layer.

[0025] The present invention also provides a semiconductor wafer, which includes a plurality of semiconductor radio frequency chips and the above three types of radio frequency test structures.

[0026] The present invention provides a method for improving the measurable frequency range of an existing radio frequency device GSG pad test structure. Without changing the size of the existing GSG pad test structure and the thickness of the semiconductor substrate layer, that is, without re-designing the existing GSG pad test structure, by adding air holes or filling metal materials around the pads of the GSG pad test structure, the reflection coefficient of the existing GSG-thru test structure at higher frequencies is reduced and its transmission coefficient is increased, and the reflection coefficient of the existing GSG-open test structure at higher frequencies is increased and its transmission coefficient is reduced, thereby achieving an increase in the upper limit of the measurable frequency of the existing GSG pad test structure. The present invention has the characteristics of simple structure, effectiveness, and compatibility with existing manufacturing processes, and provides a new method for the design of radio frequency device GSG pad test structures at higher frequencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figures 1 - 3 Schematic diagrams of the radio frequency GSG pad test structure, GSG-thru structure, and GSG-open structure of an existing radio frequency device respectively;

[0028] Figure 4 Schematic cross-sectional view of the radio frequency GSG pad test structure of an existing radio frequency device in the A-A' direction;

[0029] Figures 5 - 7 Schematic diagrams of the GSG pad test structure, GSG-thru test structure, and GSG-open test structure of the radio frequency device in Embodiment 1 of the present invention respectively;

[0030] Figure 8 Schematic cross-sectional view of the radio frequency GSG pad test structure of Embodiment 1 of the present invention in the B-B' direction;

[0031] Figure 9 S-parameter curves of S(1,1) and S(2,1) of the GSG-thru test structure in Embodiment 1 and the existing comparative example;

[0032] Figure 10 S-parameter curves of S(1,1) and S(2,1) of the GSG-open test structure in Embodiment 1 and the existing comparative example;

[0033] Figure 11 Schematic diagram of two layers of pad metal layers of the GSG pad test structure, GSG-thru test structure, and GSG-open test structure in the embodiment of the present invention;

[0034] Figure 12Schematic diagram of two-layer dielectric layers for the GSG pad test structure, GSG-thru test structure, and GSG-open test structure according to embodiments of the present invention;

[0035] Figure 13 Schematic diagram of three-layer dielectric layers for the GSG pad test structure, GSG-thru test structure, and GSG-open test structure according to embodiments of the present invention;

[0036] Figure 14 Equivalent circuit diagram of parasitic effects for the GSG-thru and GSG-open test structures implemented according to the present invention;

[0037] Figure 15 Cross-sectional schematic diagram of Embodiment 2 of the present invention;

[0038] Figure 16 Cross-sectional schematic diagram of Embodiment 3 of the present invention;

[0039] Figure 17 S-parameter curves of the GSG-thru test structure of Embodiment 3 and S(1,1), S(2,1) of the existing comparative example;

[0040] Figure 18 S-parameter curves of the GSG-open test structure of Embodiment 3 and S(1,1), S(2,1) of the existing comparative example;

[0041] Figures 19 - 21 Schematic diagrams of the radio frequency GSG pad test structure, GSG-thru test structure, and GSG-open test structure according to Embodiment 4 of the present invention;

[0042] Figures 22 - 24 Schematic diagrams of the radio frequency GSG pad test structure, GSG-thru test structure, and GSG-open test structure according to Embodiment 5 of the present invention;

[0043] Figure 25 S-parameter curves of the GSG-thru test structure of Embodiment 4 and S(1,1), S(2,1) of the existing comparative example;

[0044] Figure 26 S-parameter curves of the GSG-open test structure of Embodiment 4 and S(1,1), S(2,1) of the existing comparative example;

[0045] Figure 27 S-parameter curves of the GSG-thru test structure of Embodiment 5 and S(1,1), S(2,1) of the existing comparative example;

[0046] Figure 28S-parameter curves of the GSG-open test structure of Example 5 and S(1,1) and S(2,1) of the existing comparative example;

[0047] Figure 29 S-parameter curves of the GSG-thru test structure of Example 6 and S(1,1) and S(2,1) of the existing comparative example;

[0048] Figure 30 S-parameter curves of the GSG-open test structure of Example 6 and S(1,1) and S(2,1) of the existing comparative example;

[0049] Figures 31 - 33 Schematic diagrams of the GSG pad test structure, GSG-thru test structure, and GSG-open test structure of the radio frequency device of Example 7 of the present invention;

[0050] Figure 34 S-parameter curves of the GSG-thru test structure of Example 7 and S(1,1) and S(2,1) of the existing comparative example;

[0051] Figure 35 S-parameter curves of the GSG-open test structure of Example 7 and S(1,1) and S(2,1) of the existing comparative example;

[0052] Figure 36 、 Figure 37 Schematic diagram of the GSG pad test structure of the radio frequency device of other embodiments of the present invention;

[0053] Figure 38 Schematic diagram of the semiconductor wafer of the present invention. Detailed implementation manners

[0054] The technical solution of the present invention will be further described below with reference to the accompanying drawings and through specific implementation manners. In the description drawings of the present invention, only schematic diagrams are shown. It should be understood that the dimensions of the various parts shown in the drawings are not drawn according to the actual proportional relationship. Without conflict, the features in the embodiments of the present application can be combined with each other.

[0055] Example 1

[0056] The present invention provides one of a radio frequency device GSG pad test structure (ground-signal-ground test structure), a corresponding radio frequency device GSG-thru test structure (ground-signal-ground through test structure, also called a through calibration component), and a radio frequency device GSG-open test structure (ground-signal-ground open test structure, also called an open calibration component):

[0057] The radio frequency device GSG pad test structure is as Figure 5 、8 As shown, it includes a back gold layer 20, a semiconductor substrate layer 10, and a dielectric layer 30 arranged in sequence. The first to fourth ground pads G1 - G4, the first signal pad S1, and the second signal pad S2 are respectively arranged on the semiconductor substrate layer 10. The semiconductor substrate layer of the present invention is a GaAs substrate; the first ground pad G1, the first signal pad S1, and the second ground pad G2 form port one of the ground - signal - ground (GSG) structure; the third ground pad G3, the second signal pad S2, and the fourth ground pad G4 form port two of the ground - signal - ground (GSG) structure; the first to fourth ground pads G1 - G4 are respectively connected to the back gold layer 20 through metal vias 40; the projected areas of the first to fourth ground pads G1 - G4 on the semiconductor substrate layer 10 are generally greater than the projected areas of the first signal pad S1 or the second signal pad S2 on the semiconductor substrate layer 10. The projected areas of the first to fourth ground pads G1 - G4 on the semiconductor substrate layer 10 are rectangular and form a 2×2 array. The first signal pad S1 is a signal input pad, and the second signal pad S2 is a signal output pad. The characteristic impedance of the two - port of this RF device test structure is designed to be 50 ohms to match the impedance of common test equipment, cables, probes, etc.

[0058] The device under test (DUT) is arranged between the first signal pad S1 and the second signal pad S2. Preferably, the device under test (RF device) DUT is arranged at the middle position among the first to fourth ground pads G1 - G4, the first signal pad S1, and the second signal pad S2. The input end of the first signal pad S1 and the DUT is connected by a wiring metal, and the output end of the second signal pad S2 and the DUT is connected by a wiring metal. The thickness of the wiring metal is respectively less than or equal to the thickness of the pads (the first to fourth ground pads G1 - G4, the first signal pad S1, and the second signal pad S2). A dielectric layer 30 is provided between the first to fourth ground pads G1 - G4, the first signal pad S1, and the second signal pad S2; the thickness of the pads (the first to fourth ground pads G1 - G4, the first signal pad S1, and the second signal pad S2) is greater than the thickness of the dielectric layer. It should be noted that the middle position referred to in the present invention can be the exact center position or a position in the approximate middle area.

[0059] A dielectric layer 30 between a first ground pad G1 and a first signal pad S1 is provided with a plurality of grooves 50, and the plurality of grooves 50 are arranged at intervals in sequence; a dielectric layer 30 at a corresponding position between a second ground pad G2 and the first signal pad S1 is provided with a plurality of grooves 50, and the plurality of grooves 50 are arranged at intervals in sequence; a dielectric layer 30 at a corresponding position between a third ground pad G3 and a signal pad S2 is provided with a plurality of grooves 50, and the plurality of grooves 50 are arranged at intervals in sequence; a dielectric layer 30 at a corresponding position between a ground pad G4 and the signal pad S2 is provided with a plurality of grooves 50, and the plurality of grooves 50 are arranged at intervals in sequence; the depth of the groove is less than or equal to the thickness of the dielectric layer 30.

[0060] Correspondingly, as shown in the appendix Figure 6 As shown, the radio frequency device GSG-thru test structure includes a first ground pad G1, a first signal pad S1, and a second ground pad G2 to form port one of a ground-signal-ground (GSG) structure; a third ground pad G3, a second signal pad S2, and a fourth ground pad G4 form port two of a ground-signal-ground (GSG) structure; compared with the radio frequency device GSG pad test structure, the difference is that no device under test DUT is provided, but the first signal pad S1 and the second signal pad S2 are connected by a wiring metal.

[0061] Correspondingly, as shown in the appendix Figure 7 As shown, the radio frequency device GSG-open test structure includes a first ground pad G1, a first signal pad S1, and a second ground pad G2 to form port one of a ground-signal-ground (GSG) structure; a third ground pad G3, a second signal pad S2, and a fourth ground pad G4 form port two of a ground-signal-ground (GSG) structure; compared with the radio frequency device GSG-thru test structure, the difference is that no wiring metal is provided between the first signal pad S1 and the second signal pad S2.

[0062] The shape of the projection of the above-mentioned groove 50 on the semiconductor substrate layer 10 is circular, and the groove is filled with air, that is, the groove 50 forms an air dielectric hole through a cylindrical air dielectric. It should be noted that the formation of the air dielectric holes in this embodiment is arranged at periodic intervals, and it can also be arranged at non-periodic intervals.

[0063] Usually, the probe pitch is a fixed value such as 100μm, 150μm; in the embodiment of the present invention, the pitch between the ground pad and the signal pad is 36μm; in the implementation of the present invention, the pitch between the holes is 28μm, the size of the round hole: the radius of the round hole is 10μm, the depth of the round hole is 1.6μm, that is, the depth of the groove is 1.6μm, and the depth of the groove 50 is less than the thickness of the dielectric layer 30.

[0064] Figure 9(a) The S-parameter curve of S(1,1) for the GSG-thru test structure (red curve below) of this embodiment and the existing comparative example (black curve above). For the GSG-thru test structure, compared with the existing comparative example, the reflection coefficient S11 of this embodiment decreases within the 100 GHz frequency range. Especially around 50G - 72 GHz, the reflection coefficient S11 is less than -20 dB, and within the frequency band of 76G - 100 GHz, its reflection coefficient S11 is reduced to below -10 dB. Figure 9 (b) The S-parameter curve of S(2,1) for the GSG-thru test structure (red curve above) of this embodiment and the existing comparative example (black curve below). Compared with the existing comparative example, the transmission coefficient S21 of this embodiment increases by about 0.5 - 26.5 dB within the frequency band of 76G - 100 GHz, and S21 is greater than -1 dB (corresponding transmission rate nearly 90%) within the frequency range below 96 GHz. This means that within the frequency band of 76G - 96 GHz of the present invention, the GSG-thru test structure can also effectively improve the transmission of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0065] Figure 10 (a) The S-parameter curve of S(1,1) for the GSG-open test structure (red curve above) of this embodiment and the existing comparative example (black curve below); Figure 10 (b) The S-parameter curve of S(2,1) for the GSG-open test structure (red curve) of this embodiment and the existing comparative example (black curve); For the GSG-open test structure, S11 of this embodiment increases significantly within the frequency band of 85 GHz - 100 GHz, and S11 is greater than -1 dB (corresponding reflectivity nearly 90%) within the frequency range below 98 GHz. While its S21 is basically smaller than that of the existing comparative example within 60G - 100 GHz, and S21 is -80 dB at about 64 GHz. This means that within the frequency band of 76G - 98 GHz of the present invention, the GSG-open test structure can effectively reduce the leakage of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0066] Based on the results of the above GSG-thru test structure and GSG-open test structure, the embodiment of the present invention can increase the measurable frequency of the GSG pad test structure from 76 GHz to about 96 GHz.

[0067] In the first embodiment of the present invention, the first to fourth ground pads G1 to G4, the first signal pad S1, and the second signal pad S2 can be composed of a single-layer metal layer. It should be noted that the pad can be composed of a single-layer metal layer or a multi-layer metal stack, and its thickness is greater than the thickness of the dielectric layer 30. The multi-layer metal layer, such as a two-layer metal layer, as shown in the attached Figure 11 figure, includes metal layer M1 and metal layer M2. The dielectric layer 30 can be composed of a single layer or multiple layers. The multi-layer dielectric layer, such as a two-layer dielectric layer, as shown in the attached Figure 12 figure.

[0068] The greater the depth of the groove 50, the better. In theory, beneficial effects can be obtained as long as the depth of the groove is greater than 0. However, in the actual process manufacturing, the dielectric layer 30 is generally composed of a first dielectric layer 301 and a second dielectric layer 302 to form two layers ( Figure 12 ) or is composed of a third dielectric layer 311, a fourth dielectric layer 312, and a fifth dielectric layer 313 to form three layers ( Figure 13 ) of polymers such as silicon nitride. To be compatible with the manufacturing process, the depth of the groove 50 can be equal to the thickness of the second dielectric layer 302, or the sum of the thicknesses of the first dielectric layer 301 and the second dielectric layer 302 ( Figure 12 ); it can also be equal to the thickness of the fifth dielectric layer 313 or the sum of the thicknesses of the fourth dielectric layer 312 and the fifth dielectric layer 313, or the sum of the thicknesses of the third dielectric layer 311, the fourth dielectric layer 312, and the fifth dielectric layer 313 ( Figure 13 ). In the embodiment of the present invention, the thicknesses of the third dielectric layer 311, the fourth dielectric layer 312, and the fifth dielectric layer 313 are 1.6 μm, 0.16 μm, and 1.05 μm, respectively.

[0069] The dielectric layer 30 is generally a polymer dielectric such as silicon nitride with electromagnetic loss. In the embodiment of the present invention, the relative dielectric constant of the polymer of the fifth dielectric layer 313 is 2.9, and the loss tangent is 0.001; the relative dielectric constants of the nitrides of the third dielectric layer 311 and the fourth dielectric layer 312 are 6.9, and the loss tangent is 0.001. The groove 50 is air, and the relative dielectric constant of air is close to 1, and the loss tangent is close to 0. Analyzing from the equivalent circuit angle of the GSG-thru and GSG-open test structures, as shown in the attached Figure 14As shown in FIGS. (a) and 14(b), between the first ground pad G1 and the first signal pad S1, between the second ground pad G2 and the first signal pad S1, between the third ground pad G3 and the second signal pad S2, and between the fourth ground pad G4 and the second signal pad S2, after adding an air medium, the impedance of the two ports is changed, and the equivalent capacitances C1 to C4 between the ground pads and the signal pads are reduced, thereby increasing the resonant frequency of the GSG-thru and GSG-open test structures. Therefore, for the GSG-thru test structure, before the resonant frequency appears, its reflection coefficient S11 is reduced, and its transmission coefficient S21 in the high-frequency band is increased; while for the GSG-open test structure, its transmission coefficient S21 in the high-frequency band is reduced.

[0070] Embodiment 2

[0071] In other possible embodiments, as shown in the attached Figure 15 figures, the difference from Embodiment 1 is that the depth of the groove 50 is greater than the thickness of the dielectric layer 30 and penetrates into the semiconductor substrate layer 10. In theory, the groove 50 can penetrate into the semiconductor substrate layer 10, but the disadvantage of this embodiment is that when the etching depth increases, the size of the groove 50 will increase. When the spacing between the pad and the transmission line is limited, the etching technology requirements are very high, and the processing difficulty increases.

[0072] Embodiment 3

[0073] As shown in the attached Figure 16 figures, the difference from Embodiment 1 is that the groove 50 is filled with a metal material 70. In the implementation of the present invention, the filled metal material is gold. The spacing between the metal cylinders is 28 μm, the cylinder radius is 10 μm, and the cylinder depth is 1.6 μm. As shown in Figure 17 、 18 figures, compared with Embodiment 1, this embodiment can obtain similar beneficial effects. Although filling the metal material will increase a certain cost, in theory, after filling the metal material, beneficial effects of improving heat dissipation can also be obtained, which is beneficial to improving the measurement power of the GSG pad test structure.

[0074] Figure 17 (a) is the S-parameter curve graph of S(1,1) of the GSG-thru test structure (lower blue curve) of this embodiment and the existing comparative example (upper black curve). For the GSG-thru test structure, compared with the existing comparative example, in this embodiment, its reflection coefficient S11 is reduced within the 100 GHz frequency range, especially around 52G - 74 GHz, the reflection coefficient S11 is less than -20 dB, and within the frequency band range of 76G - 100 GHz, its reflection coefficient S11 is reduced to below -10 dB. Figure 17(b) is the S-parameter curve of S(2,1) of the GSG-thru test structure (upper blue curve) of this embodiment and the existing comparative example (lower black curve). Compared with the existing comparative example, in the frequency band range of 76G to 100GHz, the transmission coefficient S21 of this embodiment has increased by 0.5 to 26.4dB, and its S21 is greater than -1dB in the frequency range below 94GHz. This means that in the frequency band of 76G to 94GHz of the present invention, the GSG-thru test structure can also effectively improve the transmission of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0075] Figure 18 (a) is the S-parameter curve of S(1,1) of the GSG-open test structure (upper blue curve) of this embodiment and the existing comparative example (lower black curve); Figure 18 (b) is the S-parameter curve of S(2,1) of the GSG-open test structure (blue curve) of this embodiment and the existing comparative example (black curve); for the GSG-open test structure, its S11 increases significantly in the frequency band range of 85GHz to 100GHz in this embodiment, and its S11 is greater than -1dB in the frequency range below 97GHz, while its S21 is basically smaller than that of the existing comparative example within 60GHz to 100GHz, and S21 is less than -80dB around 60GHz. This means that in the frequency band of 76G to 97GHz of the present invention, the GSG-open test structure can effectively reduce the leakage of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0076] Based on the above results of the GSG-thru test structure and the GSG-open test structure, the embodiment of the present invention can increase the measurable frequency of the GSG pad test structure from 76GHz to about 94GHz.

[0077] Embodiment 4

[0078] The difference from Embodiment 1 is that, as shown in the appendix Figures 19 - 21 , the groove 50 is formed by holes with zero spacing to form a ditch. In the implementation of the present invention, the length of the ditch is 142μm, the width is 30μm, and the depth is 7.81μm. Figure 25 、 Figure 26 are respectively the S-parameter curves of S(1,1) and S(2,1) of the GSG-thru and GSG-open test structures and the existing comparative example; compared with Embodiment 1, the highest measurable frequency of this embodiment is slightly lower than that of Embodiment 1. Although this embodiment can present beneficial effects similar to those of Embodiment 1, the depth of the groove reaches 7.81μm and has penetrated into the semiconductor substrate layer 10, which is slightly more difficult and costly in the implementation process.

[0079] Figure 25 (a) is the S-parameter curve graph of S(1,1) of the GSG-thru test structure (lower pink curve) of this embodiment and the existing comparative example (upper black curve). For the GSG-thru test structure, compared with the existing comparative example, the reflection coefficient S11 of this embodiment decreases within the 100 GHz frequency range. Especially around 48G - 72GHz, the reflection coefficient S11 is less than -20 dB, and within the frequency band range of 80G - 100GHz, its reflection coefficient S11 is reduced to below -10 dB. Figure 25 (b) is the S-parameter curve graph of S(2,1) of the GSG-thru test structure (upper pink curve) of this embodiment and the existing comparative example (lower black curve). Compared with the existing comparative example, within the frequency band range of 70 - 99GHz, the transmission coefficient S21 of this embodiment is larger, and its S21 is greater than -1 dB within the frequency range below 95GHz. This means that in the frequency band of 76G - 95GHz of the present invention, the GSG-thru test structure can also effectively improve the radio frequency signal transmission from the first signal pad S1 to the second signal pad S2.

[0080] Figure 26 (a) is the S-parameter curve graph of S(1,1) of the GSG-open test structure (upper pink curve) of this embodiment and the existing comparative example (lower black curve); Figure 26 (b) is the S-parameter curve graph of S(2,1) of the GSG-open test structure (pink curve) of this embodiment and the existing comparative example (black curve); For the GSG-open test structure, its S11 increases significantly within the frequency band range of 90GHz - 98GHz of this embodiment, and its S11 is greater than -1 dB within the frequency range below 94GHz, while its S21 is basically smaller than that of the existing comparative example within 55GHz - 94GHz, and S21 is about -80 dB at 62GHz. This means that in the frequency band of 76G - 94GHz of the present invention, the GSG-open test structure can effectively reduce the leakage of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0081] Based on the results of the above GSG-thru test structure and GSG-open test structure, the embodiment of the present invention can increase the measurable frequency of the GSG pad test structure from 76GHz to about 94GHz.

[0082] Example 5

[0083] As shown in the appendix Figures 22 - 24As shown, the difference in Embodiment 1 is that the first to fourth ground pads G1 to G4 in Embodiment 1 are rectangular, forming a 2×2 array, while in the embodiment of the present invention, the first to fourth ground pads G1 to G4 near the center of the array respectively form cut corners, that is, bevel angles of 90°, so that the distance between the first to fourth ground pads G1 to G4 and the metal connection lines is increased.

[0084] In the embodiment of the present invention, the angle of the cut corner is 45 degrees. From the projection view, the length of the side of the first ground pad G1 corresponding to the first signal pad S1 is similar to the length of the side of the first signal pad S1; the distance between the ground pad and the signal pad is 36 μm, the distance between the holes is 28 μm, the radius of the round hole is 10 μm, and the depth of the round hole is 1.6 μm.

[0085] Figure 27 (a) is the S-parameter curve graph of S(1,1) of the GSG-thru test structure (lower red curve) of this embodiment and the existing comparative example (upper black curve). For the GSG-thru test structure, compared with the existing comparative example, the reflection coefficient S11 of this embodiment decreases within the frequency range of 100 GHz, especially around 46 - 70 GHz, the reflection coefficient S11 is less than -20 dB, and within the frequency band range of 76G - 100 GHz, its reflection coefficient S11 is basically reduced to below -10 dB. Figure 25 (b) is the S-parameter curve graph of S(2,1) of the GSG-thru test structure (upper red curve) of this embodiment and the existing comparative example (lower black curve). Compared with the existing comparative example, within the frequency band range of 76 - 100 GHz, the transmission coefficient S21 of this embodiment is larger, and its S21 is greater than -1 dB within the frequency range below 98 GHz, which means that in the frequency band of 76G - 98 GHz of the present invention, the GSG-thru test structure can also effectively improve the transmission of the radio frequency signal from the first signal pad S1 to the second signal pad S2.

[0086] Figure 28 (a) is the S-parameter curve graph of S(1,1) of the GSG-open test structure (upper red curve) of this embodiment and the existing comparative example (lower black curve); Figure 26(b) is the S-parameter curve of S(2,1) of the GSG-open test structure (red curve) of this embodiment and the existing comparative example (black curve); for the GSG-open test structure, in the frequency range of 85 GHz to 100 GHz, S11 of this embodiment increases significantly, and below 100 GHz, S11 is greater than -1 dB. And its S21 is basically smaller than that of the existing comparative example within 60 GHz to 96 GHz. At 60 GHz, S21 is about -70 dB, which means that in the frequency band of 76G to 100 GHz of the present invention, the GSG-open test structure can effectively reduce the leakage of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0087] Based on the results of the above GSG-thru test structure and GSG-open test structure, the embodiment of the present invention can increase the measurable frequency of the GSG pad test structure from 76 GHz to about 98 GHz.

[0088] Embodiment 6

[0089] The difference from Embodiment 5 is that the groove 50 is filled with a metal material. In the implementation of the present invention, the filled metal material is gold. The distance between the metal cylinders is 28 μm, the radius of the cylinder is 10 μm, and the depth of the cylinder is 1.6 μm. Compared with Embodiment 5, this embodiment can obtain similar beneficial effects. Although filling the metal material will increase a certain cost, theoretically, beneficial effects of improved heat dissipation can also be obtained after filling the metal material, which is beneficial to improving the measurement power of the GSG pad test structure.

[0090] Figure 29 (a) is the S-parameter curve of S(1,1) of the GSG-thru test structure (lower blue curve) of this embodiment and the existing comparative example (upper black curve). For the GSG-thru test structure, compared with the existing comparative example, in the 100 GHz frequency range, the reflection coefficient S11 of this embodiment is mostly small. Especially around 46G to 70 GHz, the reflection coefficient S11 is less than -20 dB, and in the frequency band of 80G to 100 GHz, its reflection coefficient S11 is basically reduced to below -10 dB. Figure 29 (b) is the S-parameter curve of S(2,1) of the GSG-thru test structure (upper blue curve) of this embodiment and the existing comparative example (lower black curve). Compared with the existing comparative example, in the frequency band of 76G to 100 GHz, the transmission coefficient S21 of this embodiment is larger, and below 97 GHz, S21 is greater than -1 dB, which means that in the range of 76G to 97 GHz of the present invention, the GSG-thru test structure can also effectively improve the transmission of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0091] Figure 30 (a) is the S-parameter curve of S(1,1) of the GSG-open test structure (upper blue curve) of this embodiment and the existing comparative example (lower black curve); Figure 30 (b) is the S-parameter curve of S(2,1) of the GSG-open test structure (blue curve) of this embodiment and the existing comparative example (black curve); for the GSG-open test structure, in the frequency range of 85 GHz to 100 GHz, its S11 increases significantly in this embodiment, and its S11 is greater than -1 dB in the frequency range below 99 GHz, while its S21 is basically smaller than that of the existing comparative example within 60 GHz to 96 GHz, and S21 is about -68 dB at 72 GHz, which means that in the frequency band of 76G to 99 GHz of the present invention, the GSG-open test structure can effectively reduce the leakage of radio frequency signals from the first signal pad S1 to the second signal pad S2. Combining the results of the above GSG-thru test structure and GSG-open test structure, the embodiment of the present invention can increase the measurable frequency of the GSG pad test structure from 76 GHz to about 97 GHz. It should be noted that in this embodiment, the groove is completely filled with metal. Of course, the groove can also be partially filled with metal, or a part of the groove is filled with metal and the other part of the groove is filled with a cavity.

[0092] Example 7

[0093] The difference from Example 4 is that the groove 50 forms a ditch by holes with zero spacing, as shown in the appendix Figures 31 - 33 As shown, in the implementation of the present invention, the length of the ditch is 102 μm, the width is 30 μm, and the depth is 7.81 μm. Compared with Example 5, the highest measurable frequency of this embodiment is slightly lower than that of Example 5. Although this embodiment can present beneficial effects similar to those of Example 5, the depth of the groove reaches 7.81 μm, which has penetrated into the semiconductor substrate layer 10, and the difficulty is slightly greater and the cost will be higher during the implementation process.

[0094] Figure 34 (a) is the S-parameter curve of S(1,1) of the GSG-thru test structure (lower pink curve) of this embodiment and the existing comparative example (upper black curve). For the GSG-thru test structure, compared with the existing comparative example, in the 100 GHz frequency range, the reflection coefficient S11 of this embodiment is mostly small, especially around 48G to 72 GHz, the reflection coefficient S11 is less than -20 dB, and in the frequency band of 76G to 100 GHz, its reflection coefficient S11 is reduced to below -10 dB. Figure 25(b) is the S-parameter curve of S(2,1) of the GSG-thru test structure (upper pink curve) of this embodiment and the existing comparative example (lower black curve). Compared with the existing comparative example, in the frequency band range of 76G to 100GHz, the transmission coefficient S21 of this embodiment is larger, and its S21 is greater than -1dB in the frequency range below 97GHz. This means that in the frequency band of 76G to 97GHz of the present invention, the GSG-thru test structure can also effectively improve the transmission of radio frequency signals from the first signal pad S1 to the second signal pad S2.

[0095] Figure 35 (a) is the S-parameter curve of S(1,1) of the GSG-open test structure (upper pink curve) of this embodiment and the existing comparative example (lower black curve); Figure 35 (b) is the S-parameter curve of S(2,1) of the GSG-open test structure (pink curve) of this embodiment and the existing comparative example (black curve); for the GSG-open test structure, its S11 increases significantly in the frequency band range of 87GHz to 98GHz, and its S11 is greater than -1dB in the frequency range below 95GHz, while its S21 is basically smaller than that of the existing comparative example in the range of 60GHz to 96GHz, and S21 is about -68dB at 72GHz. This means that in the frequency band of 76G to 95GHz of the present invention, the GSG-open test structure can effectively reduce the leakage of radio frequency signals from the first signal pad S1 to the second signal pad S2. Combining the results of the above GSG-thru test structure and GSG-open test structure, the embodiment of the present invention can increase the measurable frequency of the GSG pad test structure from 76GHz to about 95GHz.

[0096] It should be noted that the shape of the projection of the groove 50 on the semiconductor substrate layer 10 is circular, and the size can be changed arbitrarily (as shown in the appendix Figure 37 ), or it can be several ellipses or several rectangles as shown in the appendix Figure 36 . The formed holes are not limited to cylindrical holes, conical holes, etc. It can also be a single groove to form a ditch. For example, the shape of the hole can be circular, elliptical, conical, cuboid, etc. In the embodiment of the present invention, the hole pitch is 28μm, but in fact, the minimum hole pitch can be zero (forming a ditch), and only one unit of the hole can also be used. The above only needs to ensure that the grooves between the ground pads / signal pads of the GSG pad test structure, GSG-thru test structure, and GSG-open test structure are consistent.

[0097] It should be noted that in the embodiments provided by the present invention, in the radio frequency device GSG pad test structure (ground-signal ground test structure), radio frequency device GSG-thru test structure, and radio frequency device GSG-open test structure, grooves 50 are provided in the dielectric layer 30 of four regions, namely, the region between the first ground pad G1 and the first signal pad S1, the region between the second ground pad G2 and the first signal pad S1, the region between the third ground pad G3 and the second signal pad S2, and the region between the fourth ground pad G4 and the second signal pad S2. In other embodiments of the present invention, grooves 50 may also be provided in the dielectric layer 30 of at least one of the regions between the first ground pad G1 and the first signal pad S1, the region between the second ground pad G2 and the first signal pad S1, the region between the third ground pad G3 and the second signal pad S2, and the region between the fourth ground pad G4 and the second signal pad S2. It may be any one of these regions, or any two regions, or any three regions. The present invention will not elaborate here.

[0098] Correspondingly, the present invention also provides a semiconductor wafer 100, which includes a number of semiconductor radio frequency chips and a radio frequency test structure of one of the above three types. As shown in the Figure 38 accompanying drawings, the semiconductor wafer 100 includes a number of position regions. Each position region includes a number of semiconductor devices and a PCM region 200 (Process Control Monitor). The PCM region is set in the upper left of the position region. The semiconductor device test structure (GSG pad test structure) and the device under test of the present invention are set in the PCM region 200 (Process Control Monitor) of the semiconductor wafer 100 to monitor the process control through the electrical parameters of the devices during the manufacturing process. This is also a means to reflect the quality of each wafer product. It can accurately and timely reflect whether the products on the line are abnormal. At the same time, as an outgoing inspection standard, only the devices with the GSG pad test structure in the PCM region can be sent to the next outgoing inspection station after passing the characteristic test.

[0099] It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of other elements or steps not listed in the claim. In a unit claim listing several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc. does not denote any order. These words may be interpreted as names.

[0100] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made to these embodiments by those skilled in the art once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present invention.

[0101] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

[0102] The above are only the preferred embodiments of the present invention, and all equivalent changes and modifications made in accordance with the scope of the patent application of the present invention shall fall within the scope covered by the present invention.

Claims

1. A radio frequency device test structure, characterized in that: It includes a semiconductor base layer (10), and first to fourth ground pads (G1 to G4), a first signal pad (S1), and a second signal pad (S2) disposed on the semiconductor base layer (10). The first ground pad (G1), the first signal pad (S1), and the second ground pad (G2) form port one; the third ground pad (G3), the second signal pad (S2), and the fourth ground pad (G4) form port two; the first to fourth ground pads (G1 to G4) are connected; a device under test (DUT) is disposed between the first signal pad (S1) and the second signal pad (S2). The input end of the first signal pad (S1) and the DUT is connected by a wiring metal, and the output end of the second signal pad (S2) and the DUT is connected by a wiring metal; a dielectric layer (30) is provided between the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2); a groove (50) is provided in the dielectric layer (30) in at least one of the regions between the first ground pad (G1) and the first signal pad (S1), between the second ground pad (G2) and the first signal pad (S1), between the third ground pad (G3) and the second signal pad (S2), and between the fourth ground pad (G4) and the second signal pad (S2).

2. The radio frequency device test structure according to claim 1, wherein: The projections of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) form a 2×2 array, and chamfer angles (90) are respectively formed at positions of the first to fourth ground pads (G1 to G4) close to the center of the array.

3. The radio frequency device test structure according to claim 1, wherein: The dielectric layer is single-layer or multi-layer; the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively composed of single-layer or multi-layer metal. The thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2) are respectively greater than the thickness of the dielectric layer, and the thicknesses of the wiring metals are respectively less than or equal to the thicknesses of the first to fourth ground pads (G1 to G4), the first signal pad (S1), and the second signal pad (S2).

4. The radio frequency device test structure according to claim 1, wherein: It further includes a back gold layer (20). The projected areas of the first to fourth ground pads (G1 to G4) on the semiconductor base layer (10) are all greater than the projected areas of the first signal pad (S1) or the second signal pad (S2) on the semiconductor base layer 10. The first to fourth ground pads (G1 to G4) are respectively connected to the back gold layer (20) through metal back holes (40); the groove is filled with air or a metal material.

5. The RF device test structure according to claim 1, wherein: The number of grooves (50) provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is one or more; the number of grooves (50) provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is one or more; the number of grooves (50) provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is one or more; the number of grooves (50) provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is one or more; the projected pattern of the grooves on the semiconductor substrate layer (10) is circular, rectangular or elliptical; the depth of the grooves is greater than or equal to the thickness of the dielectric layer.

6. A radio frequency device test structure, characterized in that: It includes a semiconductor substrate layer and the first to fourth ground pads (G1 - G4), the first signal pad (S1), and the second signal pad (S2) provided on the semiconductor substrate layer (10). The first ground pad (G1), the first signal pad (S1), and the second ground pad (G2) form port one; the third ground pad (G3), the second signal pad (S2), and the fourth ground pad (G4) form port two; the first to fourth ground pads (G1 - G4) are connected; the first signal pad (S1) and the second signal pad (S2) are connected by a wiring metal; at least one region of the dielectric layer (30) in the region between the first ground pad (G1) and the first signal pad (S1), the region between the second ground pad (G2) and the first signal pad (S1), the region between the third ground pad (G3) and the second signal pad (S2), and the region between the fourth ground pad (G4) and the second signal pad (S2) is provided with grooves (50).

7. The radio frequency device test structure according to claim 6, wherein: The projections of the first to fourth ground pads (G1 - G4) on the semiconductor substrate layer (10) form a 2*2 array, and chamfer angles (90) are respectively formed at positions of the first to fourth ground pads (G1 - G4) close to the center of the array.

8. The radio frequency device test structure according to claim 6, wherein: The dielectric layer is single-layer or multi-layer; the first to fourth ground pads (G1 - G4), the first signal pad (S1), and the second signal pad (S2) are respectively composed of single-layer or multi-layer metal. The thicknesses of the first to fourth ground pads (G1 - G4), the first signal pad (S1), and the second signal pad (S2) are respectively greater than the thickness of the dielectric layer, and the thickness of the wiring metal is respectively less than or equal to the thicknesses of the first to fourth ground pads (G1 - G4), the first signal pad (S1), and the second signal pad (S2).

9. The RF device test structure according to claim 6, wherein: It further includes a back gold layer (20). The projected areas of the first to fourth ground pads (G1 - G4) on the semiconductor substrate layer (10) are all greater than the projected areas of the first signal pad (S1) or the second signal pad (S2) on the semiconductor substrate 10. The first to fourth ground pads (G1 - G4) are respectively connected to the back gold layer (20) through metal back holes (40); the grooves are filled with air or filled with a metal material.

10. The radio frequency device test structure according to claim 1, wherein: The number of grooves (50) provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is single or multiple; the projected pattern of the grooves on the semiconductor substrate layer (10) is circular, rectangular or oval; the depth of the grooves is greater than or equal to the thickness of the dielectric layer.

11. A radio frequency device test structure, characterized in that: It includes a semiconductor substrate layer and the first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2) provided on the semiconductor substrate layer (10). The first ground pad (G1), the first signal pad (S1), and the second ground pad (G2) constitute port one; the third ground pad (G3), the second signal pad (S2), and the fourth ground pad (G4) constitute port two; the first to fourth ground pads (G1-G4) are connected; a dielectric layer (30) is provided between the first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2); at least one region of the dielectric layer (30) in the region between the first ground pad (G1) and the first signal pad (S1), the region between the second ground pad (G2) and the first signal pad (S1), the region between the third ground pad (G3) and the second signal pad (S2), and the region between the fourth ground pad (G4) and the second signal pad (S2) is provided with grooves (50).

12. The radio frequency device test structure according to claim 6, characterized in that: The projections of the first to fourth ground pads (G1-G4) on the semiconductor substrate layer (10) form a 2*2 array, and chamfer angles (90) are respectively formed at positions of the first to fourth ground pads (G1-G4) close to the center of the array.

13. The RF device test structure according to claim 6, characterized in that:; The first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2) are respectively composed of single-layer or multi-layer metal. The thicknesses of the first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2) are respectively greater than the thickness of the dielectric layer, and the thicknesses of the wiring metals are respectively less than or equal to the thicknesses of the first to fourth ground pads (G1-G4), the first signal pad (S1), and the second signal pad (S2).

14. The radio frequency device test structure according to claim 6, wherein: It further includes a back gold layer (20). The projected areas of the first to fourth ground pads (G1-G4) on the semiconductor substrate layer (10) are all greater than the projected areas of the first signal pad (S1) or the second signal pad (S2) on the semiconductor substrate 10. The first to fourth ground pads (G1-G4) are respectively connected to the back gold layer (20) through metal back holes (40); the grooves are filled with air or filled with metal materials.

15. The RF device test structure according to claim 1, wherein: The number of grooves (50) provided in the dielectric layer (30) between the first ground pad (G1) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the second ground pad (G2) and the first signal pad (S1) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the third ground pad (G3) and the second signal pad (S2) is single or multiple; the number of grooves (50) provided in the dielectric layer (30) between the fourth ground pad (G4) and the second signal pad (S2) is single or multiple; the projected pattern of the grooves on the semiconductor substrate layer (10) is circular, rectangular or elliptical; the depth of the grooves is greater than or equal to the thickness of the dielectric layer.

16. A semiconductor wafer, characterized in that, Comprising a plurality of semiconductor radio frequency chips and the radio frequency test structure according to any one of claims 1-15.

Images