Garnet structure ferrite material with low sintering temperature and high saturation magnetization for co-sintering with silver or gold metallization and manufacturing process of the ferrite material.
A ferrite material with a garnet structure, using lithium carbonate as a flux and tin substitution, addresses the challenge of high sintering temperatures and metallization incompatibilities, enabling cost-effective and durable microwave circulators and isolators with high saturation magnetization.
Patent Information
- Application Number
- FR2021011861
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-11-09
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2041-11-09
AI Technical Summary
Existing technologies face challenges in producing microwave circulators and isolators with reduced cost and size without degrading performance, particularly due to complex assembly processes and the incompatibility of sintering temperatures with gold or silver metallization.
A ferrite material with a garnet structure is developed, utilizing lithium carbonate as a flux to lower the sintering temperature to below 920°C and combine it with high saturation magnetization, achieved by substituting iron ions with tin ions, enabling co-sintering with silver or gold.
The solution allows for simplified production by reducing assembly steps and avoiding the use of glue, while maintaining high saturation magnetization, thus achieving cost-effective and durable microwave circulators and isolators.
Smart Images

Figure 00000014_0000 
Figure 00000014_0001 
Figure 00000014_0002
Abstract
Description
Title of the invention: Ferrite material with garnet structure with low sintering temperature and high saturation magnetization for co-sintering with silver or gold metallization and method of manufacturing the ferrite material.
[0001] The field of the invention is that of microwave circulators and isolators. These non-reciprocal devices are used in the case of simultaneous transmission / reception or in isolator mode, in order to protect the source from impedance mismatches.
[0002] More specifically, the present invention relates to new materials and their production for integration into microwave circulators / isolators.
[0003] [Fig. 1] illustrates the diagram of a three-plate circulator consisting of 4 main elements: - ferrite 10, which causes the non-reciprocity of the circulator under the effect of a magnetic field. It can advantageously be a garnet for applications between 1 and 12 GHz; - the conductive lines 11, which may be made of gold or silver; - ground plans 12; - the permanent magnet 13.
[0004] A challenge when producing such devices, particularly those used in active scanning electronic antennas, is to reduce their cost and size without degrading their performance.
[0005] For this, these devices can advantageously be produced by co-sintering. This technique allows: - to reduce the number of steps when producing assemblies (in fact, the different elements constituting a circulator are assembled before overall sintering) and thus to simplify the production of a circulator using triplate or microstrip line technology; - to avoid the use of glue which affects the performance of the devices as well as their lifespan.
[0006] Figures 2a and 2b illustrate diagrams of the strip-plate and microstrip line technologies respectively. [Fig.2a] shows a magnet 13, a central ferrite 10, a Y junction 11 and a ground plane 12. [Fig.2b] shows a ground plane 12, a ferrite 10 and a Y junction 11.
[0007] The process of producing these devices is long and costly and the mechanical assembly of the different elements is complex. In order to simplify this process, adapting the different materials to a co-sintering process can be a solution.
[0008] In order to co-sinter with gold or silver, it is essential to adapt the central material, the ferrite, responsible for the non-reciprocal effect of the circulator when it is subjected to a magnetic field.
[0009] A ferrite material particularly well suited to the applications targeted in the present invention and in particular for applications in the frequency domain [1GHz, 12GHz] is a ferrimagnetic yttrium iron garnet (YIG).
[0010] The metals considered for the production of co-sintered metal tracks are generally gold and silver, the melting temperatures of these metals being respectively 1064 and 962 °C, it is therefore essential to adapt the sintering temperature of the ferrite. Yttrium iron garnet (YIG) has a high sintering temperature of 1450-1500 °C. It cannot therefore be considered to co-sinter it with gold or silver.
[0011] Studies published in patent applications EPI829061Al, EP2622613A1, the thesis of L. Pinier (New ferrimagnetic garnets with low sintering temperature for microwave applications. Doctoral thesis of the Ecole Polytechnique, 2006) and the thesis of L. Qassym (Study and development of garnet structure ferrites with low sintering temperature for integration in microwave circulators. Doctoral thesis of the University of Paris Saclay, 2017) have shown that it is possible to sinter YIG-derived garnet ferrites below their initial sintering temperatures. More precisely: - Cu-substituted YIG has a sintering temperature lowered to 1050°C;
[0012] - the Cu and Bi substituted YIG has a sintering temperature lowered to 950°C;
[0013] - the Cu, Bi, V substituted YIG has a sintering temperature lowered to 900°C.
[0014] These results are summarized in the following table: Substitution Sintering temperature (°C) Publication Co-sintering compatibility Comment Au Ag Yttrium iron garnet (YIG) 1450 - 1500 No No Sintering temperature too high YIG substituted Cu 1050 EP1829061A1 EP2622613A1 Th. L. Pinier No No YIG substituted Cu and Bi 950 Yes No YIG substituted Cu, Bi and V <900 Thesis L. Qassy m Yes No Chemical compatibility problem
[0015] These first results show that even with sintering temperatures sufficiently far from the melting temperature of silver, co-sintering with silver remains complicated. Document FR 3105217, disclosing the results of a study carried out by the Applicant, shows that it is possible to obtain materials with a low sintering temperature allowing co-sintering with silver using lithium carbonate as a flux. On the basis of the formulation presented above, a lithium carbonate or oxide is added after calcination (garnet phase already formed) to act as a flux. This addition makes it possible to reduce the sintering temperature to around 900°C and the co-sintering tests show compatibility of this material with silver.
[0016] The addition of lithium carbonate as a flux therefore makes it possible to lower the sintering temperature.
[0017] Furthermore, for certain applications, such as higher frequency applications for example, it is necessary to have saturation magnetizations greater than 180 mT.
[0018] From conventional garnets, a solution to increase the saturation magnetization of YIG is the substitution of the Fe3+ ions located in the octahedral site by 4+ ions, such as Zr4+, Ti4+ or Sn4+, disclosed in the following articles: - Article by G. Winkler, P. Hansen and P. Holts: Variation of the magnetic material parameters and lattice constants of polycrystalline yttrium-iron garnet by incorporation of non-magnetic ions (1972); - Article by Q. Xu et al., Effect of Sn-substitution on the micro structure and magnetic properties of Bi-CVG ferrite with low temperature sintering (2011); - Article by Y. Machida, Y. Nakayama, H. Saji, T. Yamadaya, and M. Asuma, Magnetic Properties and Resonance Linewidths of Zr- and Ti-Substituted Ca-V Garnet, IEEE Transactions on Magnetics, September 1972.
[0019] These materials have a lower permittivity.
[0020] Furthermore, the sintering temperatures presented in these articles are not low enough: 1450°C for the first, 1075°C for the second and 1270-1350°C for the third. These materials with substitution of iron by zirconium, titanium and tin are therefore not compatible for co-sintering, neither with silver nor with gold. These solutions go against the initial objective of having a sintering temperature lower than 900°C.
[0021] On a similar principle, the Applicant carried out tests with other substitutions of iron by titanium and hafnium. The results were not conclusive. The sintering temperature remains too high and incompatible with co-sintering with silver.
[0022] Another solution to obtain sufficiently high magnetizations and low sintering temperatures is the substitution of the Fe3+ ion by the Zr4+ ion associated with bismuth / copper substitutions (reference: Thesis Lilia Qassym. Study and development of garnet structure ferrites at low sintering temperature for integration in microwave circulators. Doctoral thesis from the University of Paris Saclay, 2017). This gives a material with Ms (saturation magnetization) between 180 and 200 mT sintering at 950 °C. This substitution appears to be the most promising route. Such a material then has a sufficiently high saturation magnetization. On the other hand, its sintering temperature of 950 °C remains too high to allow co-sintering with silver. Two routes to reduce the sintering temperature can be pursued:
[0023] - the use of vanadium as a substitute; or
[0024] - the use of lithium carbonate as a flux.
[0025] Due to the high chemical affinity of vanadium oxide and silver oxide at low temperature, the use of vanadium for co-sintering with silver metallization has been prohibited. The remaining solution is therefore the use of lithium carbonate on the zirconium-substituted material. A study has been conducted highlighting the ineffectiveness of using lithium carbonate: while its addition as a flux after chamotte does indeed result in sufficient densification below 950 °C, it also leads to the appearance of a parasitic phase disturbing the dielectric (permittivity and dielectric losses) and magnetic properties of the material with a significant decrease in saturation magnetization.
[0026] Indeed, based on the teachings of the prior art, an addition of lithium carbonate has been made to the zirconium-substituted compositions which meet the criterion of sufficiently high saturation magnetization. Unfortunately, the insertion of lithium into the zirconium-substituted mesh is not optimal and, if the density is sufficient at low sintering temperature (at least 95% densification at 900°C), its insertion leads to a significant increase in losses and especially a reduction in saturation magnetization.
[0027] The table below shows the saturation magnetization of two materials (0.3 and 0.5 zirconium substituted) for an addition of 0, 0.64% and 1.28% by weight of lithium carbonate. It appears that the more lithium carbonate is added, the more the saturation magnetization decreases. % by mass Li 2 CO 3 Zr = 0.3 Zr = 0.5 0 190 mT 183 mT 0.64 179 mT 170 mT 1.28 168 mT 155 mT
[0028] Thus, it has been demonstrated that new Bi-Cu substituted garnets can be co-sintered with gold but also silver, with a view to reducing costs, thanks to the use of lithium carbonate flux. But these new materials have a saturation magnetization of around 170 mT. However, it has been shown that for higher frequency applications (Ku band for example), it is essential to increase the magnetization of our garnets. A solution from previous studies makes it possible to obtain saturation magnetizations greater than 180 mT but this material is not compatible with the use of lithium carbonate as a flux (good densification at low sintering temperature but appearance of losses and reduction of the saturation magnetization due to the appearance of parasitic phases).
[0029] In other words, known solutions from the prior art make it possible to reduce the co-sintering temperature of the material. Other known solutions make it possible to increase the saturation magnetization of the material. But these solutions are not compatible with each other. The known prior art therefore does not offer a solution for obtaining a ferrite material with a garnet structure at a low sintering temperature and high saturation magnetization for co-sintering with a silver or gold metallization.
[0030] In this context, the Applicant proposes another solution using a new ferrite material comprising lithium and sintering at less than 920°C while having a saturation magnetization greater than 180 mT. This new ferrite material can be advantageously obtained from the addition of lithium salt as a flux to calcined (or chamotted) ferrite powder known as “raw” before sintering. The flux phase, having a low melting temperature, helps with sintering.
[0031] The lithium salt used can advantageously be, for example, lithium carbonate, its melting temperature being 723°C.
[0032] More specifically, the present invention relates to a ferrite material with a garnet structure characterized in that it corresponds to the following chemical formula:
[0033] (YaTRbBib FeeAldlneCafCugSnhViCOjSikOn+y) + x (Li2CO3 or Li2O or LiF or LiCl) with x, in mass %, between 0.1 and 2; with TR: a rare earth or a combination of rare earths not including element Y;
[0034] -1 < y < 1 ; 3 (a+b+b'+c+d+e) + 2 (f+g+j) + 4 (h+k) + 5 i = 24±2y 0 < a < 2.5; 0 <b< 1,5 ; 0.5 < b' <3 ; 4 <c<5; 0 <d< 1,5 ; 0 < e < 0.8 ; 0 <f< 1; 0 < g < 0.05 ; 0 <h< 1 ;
[0035] 0 < i < 0.8 ; 0 <j<0,5; 0 < k < 0.5.
[0036] The invention also relates to a component or device comprising the ferrite material according to the invention and metallizations advantageously made of silver. The component or device may be a circulator, an isolator, a switch or a phase shifter.
[0037] The invention also relates to a method for manufacturing a ferrite material according to the invention, said method comprising: - weighing of raw materials such as oxides or carbonates; - initial grinding of said raw materials; - a calcination step (also called chamotte) of said raw materials ground at a temperature between 750°C and 850°C to obtain a powder of raw ferrite material; - a second grinding of said raw ferrite material powder; - the addition of a lithium salt flux which may be Li2CO3, Li2O, LiF or LiCl, to said raw ferrite material powder.
[0038] The ferrite material resulting from the raw ferrite material powder and the lithium salt flux can advantageously be sintered at a temperature between 850°C and 920°C.
[0039] According to variants of the process of the invention, the raw material being Y2>O3Cao, 22Bi o,67Cuo,o5Fe4,7Sno,30i2, or Y1,98Cao,3Bio,67Cuo,o5Fe4,65Sn0> 3 oOj2 y, between 0.2 and 0.75% by mass of Li2CO3 flux is added, the sintering temperature on the resulting assembly being of the order of 850°C to 920°C.
[0040] The invention also relates to a method for manufacturing a component or device comprising several elements including a ferrite material according to the invention, characterized in that: - said ferrite material is prepared according to the method of the invention; - a sintering operation is carried out during which said ferrite is secured to at least one of said other elements of said component or device, or to a precursor (gold or silver metal, low sintering temperature dielectric, low sintering temperature ferrite, etc.) of said element by a co-sintering process.
[0041] In the component or device, at least one of the other elements is a layer of metal and advantageously silver.
[0042] The invention will be better understood and other advantages will appear on reading the description which follows, given without limitation and thanks to the appended figures among which:
[0043] [Fig-1] illustrates the diagram of a triplate circulator according to the known art;
[0044] [Fig.2a] illustrates a triplate circulator according to the known art which can be produced by co-sintering;
[0045] [Fig.2b] illustrates a microstrip line circulator according to the known art which can be produced by co-sintering;
[0046] [Fig.3] illustrates dielectric loss and permittivity test results of a ferrite material of the invention as a function of the frequency expressed in Hertz;
[0047] [Fig.4] illustrates co-sintering test results of a ferrite of the invention with a silver deposit, after co-sintering at 900°C;
[0048] [Fig.5] represents a flowchart of the steps of the manufacturing process of a ferrite material according to the invention.
[0049] Generally speaking, the ferrite material of the invention corresponds to the following chemical formula:
[0050] (YaTRbBib FecAldIneCafCugSnhViCOjSikOi2±Y)+ x (Li2CO3 or Li2O or LiF or LiCl) with x, in mass %, between 0.1 and 2; with TR: a rare earth or a combination of rare earths not including element Y;
[0051] -1 < y < 1 ; 3 (a+b+b'+c+d+e) + 2 (f+g+j) + 4 (h+k) + 5 i = 24±2y; 0 < a < 2.5 ; 0 <b< 1,5 ; 0.5 < b' < 3 ; 4 <c<5; 0 <d< 1,5 ; 0 < e < 0.8 ; 0 <f< 1; 0 < g < 0.05 ; 0 <h< 1 ;
[0052] 0 < i < 0.8 ; o < j < 0.5 ; 0 < k < 0.5.
[0053] By this formula, it is necessary to understand that we take x (between 0.1 and 2) grams of Li2CO3 (or Li2O or LiF or LiCl) for 100g of ferrite of formulation Ya TRbBib'FecAldlneCafCugSnhViCojSikOn+y.
[0054] Rare earths can be used to modify saturation magnetization and improve power handling.
[0055] The use of copper and bismuth makes it possible to reduce the sintering temperature.,
[0056] Bismuth is used in order to increase the permittivity and lower the sintering temperature.
[0057] Calcium can be used to counterbalance excess charges due to the introduction of zirconium, tin or titanium. It can help stabilize the mesh in the case of using a high level of bismuth for example.
[0058] Substituting iron with aluminum can allow a reduction in line widths and a decrease in saturation magnetization.
[0059] Preferably, the added lithium salt has a mass proportion of between 0.15 and 1.5%, and even more preferably between 0.15 and 1.2%, for example 0.75%. This mass proportion depends on the ferrite material. Adding lithium salt can lower the sintering temperature. However, lithium present in the material can, beyond a certain quantity, cause material pollution. The preferred ranges and the example of 0.75% cited allow for a good compromise in terms of sintering temperature and the properties of the resulting material.
[0060] As will become clear below, the substitution of iron by tin makes it possible to adapt the saturation magnetization.
[0061] The ferrite materials of the present invention are produced according to a conventional ceramic synthesis process. The different stages, represented by a flowchart in [Fig.5], are:
[0062] - weighing of raw materials (step 100); - grinding the previously obtained mixture (step 110); - a calcination step (step 120), also called chamotte, the aim of which is to synthesize the garnet phase in powder form and obtain a powder of so-called raw ferrite material; - a second grinding step (step 130); - the introduction of a lithium-based flux addition (step 140): the flux addition takes place on the calcined phase (garnet phase formed by heat treatment). It is added to the calcined phase using a mixer (paddle mixer, attritor, jar, etc.).
[0063] After this addition of flux, conventional shaping (step 150) and sintering (step 160) steps can be carried out. Sintering at a higher temperature of the The purpose of reground chamotte powder is to densify the ceramic while giving it the desired shape.
[0064] The first and second grinding are advantageously carried out in a humid environment.
[0065] The sintering of the reground powder is advantageously carried out in air or under oxygen.
[0066] The grinding operations are advantageously carried out with a ball mill and / or by attrition.
[0067] The method for manufacturing a ferrite material according to the invention may optionally comprise an additional calcination step. However, it may be noted that advantageously, the manufacturing method according to the invention may comprise only a single calcination step.
[0068] Indeed, the materials of the invention are obtained by a conventional ceramic synthesis process. But the number of steps and therefore the simplicity of the synthesis depends on the material. The insertion of zirconium oxide into the garnet structure is complicated and two calcination steps are necessary to obtain a pure phase. Some impurities may even remain after these two calcinations and be responsible for magnetic and dielectric losses. In the case of the use of tin oxide in the material of the invention, on the other hand, a single calcination may be sufficient to obtain the completely pure phase.
[0069] The performances obtained are described below in the context of examples with tin-substituted ferrite materials comprising the use of lithium carbonate as flux and with tin-substituted ferrite materials not comprising the addition of lithium-based flux.
[0070] In order to highlight the effects of the use of tin and lithium carbonate, different compositions were studied: — Ferrite A: Y2jo3Caoj25Bio,67Cuojo5Fe4j7Snoj3Oi2 - Ferrite B: Y^osCao^Bio.evCuo.osFe^vSno.sO^ + 0.64% by mass Li2CO3 — Ferrite C: Ylj93Caoj35Bioj67Cuojo5Fe4>6Snoj40i2 - Ferrite D: Yij93Caoj35Bio,67Cuojo5Fe4j6Snoj40i2 + 0.64% by mass Li2CO3 — Ferrite E: Ylj83Caoj45Bioj67Cuojo5Fe4j5Snoj50i2 - Ferrite F: Yij83Caoj45Bioj67Cuojo5Fe4j5Snoj50i2 + 0.64% by mass Li2CO3
[0071] The first characterization is the verification of the formation of the phase by X-ray diffraction. This measurement confirms good formation of the phase.
[0072] Initially, the shrinkages of the materials were observed for sintering at 900°C. In order to have a reasonable density, the expected shrinkage value is 16-17%. Material % shrinkage Densities (g / cm3) Ferrite A 0 - Ferrite B 17.1 5.32 Ferrite C 0 - Ferrite D 16.1 5.25 Ferrite E 0 - Ferrite F 16.8 5.21
[0073] The impact of lithium carbonate on densification is clear: the shrinkage, non-existent in its absence, increases to more than 16% when lithium carbonate is added as a flux to the phase already formed. Lithium carbonate therefore allows for an improvement in densification. It remains to be verified that the saturation magnetization is not impacted as in the case of substitutions of iron by zirconium, titanium or hafnium.
[0074] The saturation magnetizations of Sn-substituted materials and those with the addition of lithium carbonate are presented in the table below. Material Saturation magnetization (mT) Ferrite B 192 Ferrite D 178 Ferrite F 169
[0075] These results show that the ferrite B material substituted for tin comprising the use of lithium carbonate according to the invention has a saturation magnetization greater than 190 mT.
[0076] [Fig. 3] illustrates test results of the dielectric losses (tan ô) and permittivity (e) of a ferrite material of the invention as a function of the frequency expressed in Hertz. The dielectric losses were measured. The results are presented for ferrite material B, which has a saturation magnetization greater than 190 mT.
[0077] Curve C1 represents the permittivity with a sintering temperature of 900 °C. Curve C2 represents the permittivity with a sintering temperature of 900 °C and annealing at 925 °C. Curve C3 represents the dielectric losses with a sintering temperature of 900 °C. Curve C4 represents the dielectric losses with a sintering temperature of 900 °C and annealing at 925 °C. We see that the permittivity remains high, around 21 and the losses are on the other hand a little high at low frequency (tan ô of the order of 5.102). A second annealing was therefore carried out, it allows a reduction in relaxation: the permittivity remains the same, but the losses decrease (tan ô of the order of 3.102).
[0078] Dielectric losses were also measured at 9.2 GHz (waveguide measurement). The results are as follows: e = 21 and tan ô = 7.104.
[0079] These values are entirely satisfactory for the application: the permittivity remains high for a garnet and the losses are very low.
[0080] [Fig.4] illustrates co-sintering test results of a ferrite of the invention with a silver deposit, after co-sintering at 900°C. Co-sintering tests confirm the compatibility of the material according to the invention with silver. A silver layer was deposited using a screen printing screen and then the assembly was co-sintered at 900°C. The images shown in [Fig.4] present the results obtained.
[0081] These co-sintering tests are entirely conclusive. Indeed, after sintering the ferrite pellet and the metallization at 900°C, the quality of the silver layer is not altered and no diffusion is observed.
[0082] The direct application of the present invention is the production of circulators and insulators using LTCC (Low Temperature Cofired Ceramics) technology or other additive technology. The present invention makes it possible to simplify complex assemblies thanks to co-sintering: the number of production steps is reduced. In addition, it is possible to avoid the use of glues which degrade the properties and make it a less durable technology.
[0083] It is also recalled that it was not possible, until now, to co-sinter a garnet structure ferrite with high saturation magnetization (for applications in Ku band circulators for example) with silver (problem of sintering temperature or compatibility with silver). The present invention makes it possible to co-sinter silver with a garnet structure ferrite having a saturation magnetization greater than 180 mT.
[0084] Thus, the invention is based on the substitution of the iron ions of the garnet phase by tin ions combined with the use of lithium carbonate or lithium oxide as flux. While the substitutions known from the prior art indicated that the increase in saturation magnetization is to the detriment of the sintering temperature, and vice versa, the two characteristics combined together according to the invention make it possible to obtain a ferrite of garnet structure compatible for co-sintering with silver and having a sufficiently high saturation magnetization.
Claims
Claims
1. Ferrite material of garnet structure characterized in that it corresponds to the following chemical formula: (¥aTRbBib FecAldIneCafCugSnhViCOjSikOi2±Y)+ x (Li2CO3 or Li2O or LiF or LiCl) with x, in mass %, between 0.1 and 2; with TR: a rare earth or a combination of rare earths not including the element Y; -1 < y < 1; 3 (a+b+b'+c+d+e) + 2 (f+g+j) + 4 (h+k) + 5 i = 24±2y; 0 < a < 2.5; 0 <b< 1,5 ; 0,5 < b’ < 3 ; 4<c<5; 0<d< 1,5 ; 0 < e < 0,8 ; 0<f< 1; 0 < g < 0,05 ; 0<h< 1 ; 0 < i < 0,8 ; 0<j<0,5; 0 < k < 0,5,
2. Device comprising the ferrite material according to claim 1 and metallizations, the metallizations preferably being silver.
3. Circulator comprising the ferrite material according to claim 1 and metallizations, the metallizations preferably being silver.
4. Insulator comprising the ferrite material according to claim 1 and metallizations, the metallizations preferably being silver.
5. A method of manufacturing a ferrite material according to claim 1, said method comprising: - a step of weighing (step 100) raw materials of oxide or carbonate type; - a first grinding (step 110) of said raw materials; - a step of calcining (step 120) said raw materials ground at a temperature between 750°C and 800°C to obtain the raw ferrite material; - a second grinding (step 130) of said raw ferrite material powder; - a step of adding (step 140) a lithium salt flux from among Li2CO3, Li2O, LiF or LiCl, to said raw ferrite material powder.
6. Method for manufacturing a device comprising several elements including a ferrite material according to claim 1, characterized in that: - said ferrite material is prepared by the method according to the claims - a sintering operation is carried out during which said ferrite material is secured to at least one of said other elements of said device, or to a precursor of said element by a co-sintering process.
7. Method of manufacturing a device, according to claim 6, characterized in that at least one of the other elements is a layer of metal.
8. A method of manufacturing a device, according to claim 7, characterized in that the metal is silver.
