CN114181629B - A kind of polishing liquid and its preparation method and application - Google Patents

Abstract

The invention discloses a polishing liquid, a preparation method and application thereof. The polishing liquid includes _CeO2 suspension and a surfactant, the surfactant is selected from ionic surfactants or nonionic surfactants. Carry out chemical mechanical polishing to SiO ₂ and Si ₃ N ₄ blank wafers with the polishing liquid of the present invention, the removal rate of silicon nitride can be lower than

The surface roughness of the polished SiO ₂ and Si ₃ N ₄ blank wafers are all below 1nm, and the number of surface scratches on the SiO ₂ film layer and Si ₃ N ₄ film layer is less than or equal to the surface scratches when no active agent is added Quantity, the maximum depth of scratches on the surface are also all less than the maximum depth of scratches when no activator is added, and are reduced by about 200pm and 250pm respectively, indicating that the wafer can have good surface quality when polished with the polishing solution of the present invention.

Description

A kind of polishing liquid and its preparation method and application

technical field

The invention relates to the technical field of polishing liquid, in particular to a polishing liquid and its preparation method and application.

Background technique

Chemical-mechanical polishing (CMP) is the only technology that can simultaneously achieve global planarization and local planarization in VLSI manufacturing, among which shallow trench isolation chemical-mechanical polishing (STICMP) is an important step in integrated circuit manufacturing. process.

STI is a structure used to form isolation devices. The formation process of the STI structure is as follows: first deposit a thin layer (about 10nm) of Tetra-ethyl-ortho-silicate (Tetra-ethyl-ortho-silicate, TEOS)-based silicon dioxide, and then deposit a layer of silicon nitride film with a thickness of about 35nm on it. This layer of TEOS silicon dioxide is to buffer the stress between the silicon nitride film and the silicon wafer to reduce or avoid Damage to silicon wafers. Then etching is performed on the surface of the silicon wafer to form an isolation trench; and then the surface of the silicon wafer is coated with silicon dioxide by high density plasma (High Density Plasma, HDP) to fill the trench. After the groove is fully filled with HDP SiO ₂ , the HDP SiO ₂ on the surface of the silicon wafer outside the groove is removed by CMP with a polishing solution, and the polishing is stopped on the surface of the silicon nitride film layer. Then use high-temperature phosphoric acid to etch the silicon nitride film layer to remove the silicon nitride, and HDP SiO ₂ is retained because it is not etched by phosphoric acid, so an STI structure is formed on the surface of the silicon wafer. It can be seen that STI CMP requires the polishing solution to have high removal selectivity between SiO ₂ and Si ₃ N ₄ to ensure that the HDP SiO ₂ layer outside the trench is completely removed and the Si ₃ N ₄ layer is retained.

The main component of the polishing fluid used in the STI CMP process is CeO2 suspension, and the cerium _oxide (CeO2 ₎ particles (20nm to 1μm in diameter ₎ in the suspension have high removal selectivity to _SiO2 and _Si3N4 . _The usual process of preparing CeO2 suspension is to disperse the prepared _CeO2 particles into water to form _CeO2 suspension. However, because _{CeO2 particles} usually agglomerate under the action of capillary force in the air, they stick together to form agglomerates, and it is difficult to directly disperse the agglomerates in water. Obvious sedimentation and poor stability, used for polishing will not only reduce the removal rate of _SiO2 , but also increase surface scratches, resulting in many surface defects of _SiO2 layer after polishing (see Lei H, Lu H, Luo J et al. "Preparation ofα-alumina-g-polyacrylamide composite abrasive and chemicalmechanical polishing behavior"[J].Thin Solid Films,2008,516(10):3005–3008.), the surface quality after polishing is poor.

而仅含有CeO₂的抛光液对氮化硅的去除速率约为

不能满足CMP的要求。最常见的是向CeO₂悬浮液中加入表面活性剂以抑制Si₃N₄的去除。然而，由于CeO₂悬浮液的稳定性差，由其制备得到的抛光液稳定性也较差，用于CMP时，抛光后的晶圆片表面质量差。In practical applications, the removal rate of silicon nitride by polishing fluid is required to be lower than

However, the removal rate of silicon nitride by the polishing solution containing only _CeO2 is about

Cannot meet the requirements of CMP. Most commonly, surfactants are added to the _CeO2 suspension to inhibit the removal _of Si3N4 _. However, due to the poor stability of the _CeO2 suspension, the polishing fluid prepared from it is also poor in stability, and when used for CMP, the surface quality of the polished wafer is poor.

In addition, the polishing liquid composed of surfactant and CeO2 suspension is usually sensitive to the change of pH value, that is, the removal rate of silicon _dioxide and silicon nitride can only be guaranteed at a certain specific pH value, and the pH value occurs A small change will have a great impact on the removal rate of SiO ₂ and Si ₃ N ₄ , the removal selectivity ratio and/or the number of scratches on the wafer surface, roughness, etc., which limits the scope of application of the polishing fluid.

Contents of the invention

The object of the present invention is to address the technical defects in the prior art. In the first aspect, it provides a polishing solution for CMP with improved stability. This polishing liquid comprises surfactant and CeO Suspension, described surfactant is selected from ionic surfactant or nonionic surfactant, and ionic surfactant is preferably piperazine or ₂ -methylpiperazine, nonionic Type surfactants are preferably PEG or PVP (preferably PVP with a molecular weight not less than 50,000, such as PVP-K30 and/or PVP-K90).

_The CeO2 concentration is 0.015wt.%-0.15wt.% (preferably 0.015wt.%-0.075wt%), the concentration of the surfactant is 0.015wt%-0.075wt.% (preferably 0.015wt.% -0.045wt.%); Optionally, CeO ₂ is equal to the concentration of surfactant.

The particle size of the CeO ₂ particles is 30-150nm (preferably 30-120nm or 55-150nm, more preferably 30-90nm or 55-60nm); optionally, there is no obvious precipitation after standing for 7 days.

(优选

)；可选的，pH值为4或12时，所述抛光液对二氧化硅和氮化硅的去除选择比为20-30。When the pH value is 4 or the pH value is 7-12 (preferably 10-12), the removal rate of silicon nitride by the polishing solution is

(preferred

); Optionally, when the pH value is 4 or 12, the removal selectivity ratio of the polishing solution to silicon dioxide and silicon nitride is 20-30.

After polishing the silicon dioxide blank wafer and the silicon nitride blank wafer respectively with the said polishing liquid having a pH value of 4-12 (preferably 8-12), the number of scratches on the surface of the silicon dioxide blank wafer ≤11 (preferably ≤8), and the number of scratches on the surface of the blank silicon nitride wafer is ≤4 (preferably ≤2).

Second aspect, the present invention provides a kind of preparation method of above _- mentioned polishing fluid, CeO granule, surfactant and water are added in ball mill and carry out wet milling pulverization, the liquid obtained after pulverization is described polishing fluid, during pulverization, in ball mill The diameter of the grinding ball is 0.5-2mm, preferably 0.5-1.5mm.

The mass ratio of the balls in the ball mill to the CeO2 particles during pulverization is _2-8 , preferably 5-7.

_The mass ratio of water to CeO2 particles during pulverization is 2-32, preferably 8-32.

The crushing time is 6-12h, preferably 10-12h.

In a third aspect, the present invention provides the application of the above polishing liquid or the polishing liquid prepared by the above method in the polishing liquid for chemical mechanical polishing, especially the application of the polishing liquid for STI CMP.

_The CeO2 suspension contained in the polishing liquid of the present invention has good stability, and the suspension is obtained by using a ball mill to wet _- mill CeO2 particles, that is, the mixture of _CeO2 particles and water is placed in a ball mill for pulverization, and the ball mill is adjusted simultaneously. The parameters such as ball diameter, ball-abrasive mass ratio, water-abrasive mass ratio, ball milling time and other parameters in the process make the above parameters play a synergistic effect _on the crushing of CeO2 particles, and the _CeO2 particles agglomerated to the micron level are crushed to the electron microscope Nanoparticles with an average particle size of about 100nm can significantly improve the agglomeration and hardening of CeO2 particles, and the obtained _CeO2 suspension has good stability, and there is still no obvious settlement after standing for ₇ days. The surfactant contained in the polishing liquid of the present invention is ionic or nonionic, and the addition of the surfactant will not reduce the stability of the ceria suspension, and there is still no obvious settlement after standing for 7 days.

抛光后氧化硅和氮化硅空白晶圆片的二氧化硅膜层和氮化硅膜层表面划痕最大深度分别比CeO₂悬浮液的最大划痕深度降低约200pm和250pm，抛光后的空白晶圆片具有良好的表面质量。Carry out chemical mechanical polishing to SiO ₂ and Si ₃ N ₄ blank wafers with the polishing liquid of the present invention, the removal rate of silicon nitride can be lower than

After polishing, the maximum depth of scratches on the silicon dioxide film layer and silicon nitride film layer surface of silicon oxide and silicon nitride blank wafers is about 200pm and 250pm lower _than the maximum scratch depth of CeO2 suspension, and the blank wafer after polishing Wafers have good surface quality.

甚至还可在pH值为4-12内，达到此去除速率；与仅使用CeO₂悬浮液抛光相比，在pH值为4或10-12时可显著提高氧化硅/氮化硅的去除选择比，在pH为4或6、10-12时可显著降低空白晶圆片上氧化硅膜层的表面划痕数量，在pH为6-7时可显著降低空白晶圆片上氮化硅膜层的表面划痕数量；即：pH在上述范围内变化时，本发明抛光液在二氧化硅、氮化硅的去除速率，氧化硅/氮化硅的去除选择比和/或晶圆片表面的划痕数量等方面的抛光效果不受影响，甚至能够提高抛光效果，拓宽了抛光液的适用范围。In addition, the sensitivity of the polishing liquid of the present invention to the pH value is reduced, that is, the pH value can be guaranteed in a large range, and even the removal rate of the polishing liquid to silicon dioxide and silicon nitride, silicon oxide/nitridation can be improved. The removal selectivity ratio of silicon can reduce the number of scratches on the surface of silicon oxide and silicon nitride film layers on blank wafers, such as: when the polishing solution of the present invention has a pH value of 10-12, the removal rate of silicon nitride can be guaranteed to be low At

This removal rate can even be achieved in the pH range 4-12 _; significantly improved silicon oxide/silicon nitride removal selectivity at pH 4 or 10-12 compared to polishing with CeO2 suspension only Compared, when the pH is 4 or 6, 10-12, the number of surface scratches on the silicon oxide film layer on the blank wafer can be significantly reduced, and the number of scratches on the silicon nitride film layer on the blank wafer can be significantly reduced when the pH is 6-7. Surface scratch quantity; That is: when pH changes in the above-mentioned range, the removal rate of polishing solution of the present invention in silicon dioxide, silicon nitride, the removal selectivity ratio of silicon oxide/silicon nitride and/or the scratch on wafer surface The polishing effect in aspects such as the number of scratches is not affected, and even the polishing effect can be improved, which broadens the scope of application of the polishing liquid.

Description of drawings

Fig. ₁ shows the CeO before and after ball milling The morphology photo of nanoparticles under the scanning electron microscope;

Fig. ₂ shows the CeO in the suspension of embodiment 4 and 7 The X-ray diffraction (XRD) figure of particle;

Fig. 3 shows that embodiment ₄ and comparative example 1-3 CeO in the CeO in the suspension liquid Particle size _- light intensity distribution curve;

Fig. 4 shows that embodiment 1-3, embodiment 6-7 and comparative example ₄ CeO in the CeO in the suspension liquid Particle size _- light intensity distribution curve;

Fig. 5 shows the CeO of embodiment ₄ and comparative example 1-3 Stability photo of the suspension;

Fig. 6 is shown as embodiment 1-3, embodiment 6-7 and the CeO of comparative example _4Stability photo of suspension;

Fig. ₇ is shown as embodiment 7,9 and the CeO of comparative example 6-7 Stability photo of suspension;

The removal rate curve graph of the cerium oxide suspension of embodiment 5 of Fig. 8 different concentrations to silicon oxide and silicon nitride;

The histogram of the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ by the cerium oxide suspension of embodiment 5 with different concentrations in Fig. 9 ;

The surface roughness histogram of the cerium oxide suspension of embodiment 5 with different concentrations of Fig. 10 polishing SiO ₂ and Si ₃ N ₄ blank wafers;

Fig. 11 is the 2D surface topography photo after polishing SiO ₂ and Si ₃ N ₄ blank wafers with the cerium oxide suspension of Example 5 with a concentration of 0.15wt.%;

Fig. 12 uses the cerium oxide suspension of embodiment 5 of different concentrations to polish SiO ₂ and Si ₃ N _The film layer thickness profile after blank wafer;

Figure 13 shows the stability photos containing different molecular weight PVP polishing fluids;

Shown in Figure 14 is to contain the stability photo of six kinds of different ionic surfactant polishing liquids;

Figure 15 shows the stability photos containing five different nonionic active agent polishing fluids;

Fig. 16 shows that CeO in the polishing liquid of embodiment 2-1 and _2-2 The topography photo of nanoparticles under the scanning electron microscope;

Fig. 17 is shown as comparative example 2-5～2-7 CeO in the polishing solution ₂ The morphology photo of the nanoparticles under the scanning electron microscope;

Fig. 18 is a histogram showing the removal rate of silicon nitride and the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ by the polishing solutions of Examples 2-1 and 2-2 at different pH;

Figure 19 shows the histogram of the number of scratches on the surface of SiO ₂ and Si ₃ N ₄ blank wafers polished by the polishing fluids of Examples 2-1 and 2-2 at different pHs;

Fig. 20 is a histogram showing the removal rate of silicon nitride and the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ by the polishing solutions of Examples 2-3 and 2-4 at different pH;

Fig. 21 is a histogram showing the number of scratches on the surface of SiO ₂ and Si ₃ N ₄ blank wafers polished by the polishing solutions of Examples 2-3 and 2-4 at different pHs.

Detailed ways

Through research and exploration, the present invention finds that the combined use of certain ball milling parameters (such as ball diameter, ball-abrasive mass ratio, water-abrasive mass ratio, and ball milling time ₎ can significantly improve the pulverization effect of CeO particles, thereby improving CeO ₂ dispersibility of particles in water, and then prepare CeO ₂ suspension with improved stability. The CeO ₂ suspension prepared by this method is used in the polishing of STI CMP, which can achieve high SiO ₂ and Si ₃ N ₄ removal selectivity, and the polished SiO ₂ and Si ₃ N ₄ blank wafers have rough surfaces Low density, good wafer uniformity, and less surface scratches.

_A method for preparing a CeO2 suspension, specifically:

Commercially available _{CeO2nanoparticles} (purchased from Zhongke Leiming (Beijing) Technology Co., Ltd., the purchased particle size specification is 200nm) are placed in the ball mill jar of a ball mill (QM-3SP4 planetary ball mill), and water is added for ball milling; Wherein, the inner diameter of the ball mill jar is 70mm, the rotating speed of the ball mill is 400r/min, the rotation direction of the ball mill jar and the sun disk of the ball mill is opposite, the ratio of the speed of the ball mill jar to the sun disk speed is 2, and the ball milling time (that is, the crushing time when pulverizing with a ball mill) ) is 6-12h, water-abrasive mass ratio (i.e.: water and CeO _The mass ratio of nanoparticle, hereinafter referred to as water-material ratio) is 2-32, grinding ball-abrasive mass ratio (i.e.: the grinding ball in the ball mill tank The mass ratio of the total mass to the CeO2 nanoparticles ₍ hereinafter referred to as the ball-to-material ratio) is 2-8, and the diameter of the grinding ball is 0.5-2mm. The number of grinding balls is determined by the two factors of ball-material ratio and grinding ball diameter, generally 10-100. After ball milling, the liquid in the ball mill tank is _CeO2 suspension.

Subsequently, the prepared CeO ₂ suspension can be mixed with deionized water according to specific needs to obtain CeO ₂ suspensions with different concentrations.

Among ball milling parameters:

Ball milling time: It is generally believed that with the increase of ball milling time, the particle size of the abrasive will become smaller. However, it has been reported in the literature that the longer the grinding time, the larger the particle size of the abrasive (see AN Streletskii and THCourtney, "Kinetic, chemical and mechanical factors affecting mechanical alloying of Ni-bcctransition metal mixtures," Mater. Sci. Eng. A, vol. 282, no. 1–2, pp. 213–222, 2000.). In the method of the present invention, the ball milling time is 6-12h, preferably 10-12h. Experiments show that in the range of 6-12h, the longer the ball milling time is, the better the crushing effect of the abrasive is; The collision of the abrasive (CeO ₂ nanoparticles) is not sufficient, and the pulverization effect is poor; if the ball milling time is too long, the CeO ₂ nanoparticles will undergo phase transition and produce new substances (see Yadav TP, Srivastava O N. "Synthesis of nanocrystalline cerium oxide by high energy ballmilling”[J]. Ceramics International, Elsevier Ltd and Techna Group Srl, 2012, 38(7):5783–5789. When the ball milling time is 20 hours, the number of XRD diffraction peaks of CeO ₂ nanoparticles changes).

Grinding ball-abrasive mass ratio: In the method of the present invention, the grinding ball-abrasive mass ratio is 2-8, preferably 5-7, and the ball-material ratio is too small, which means that the number of grinding balls is also small, and the ratio of grinding balls and abrasives Insufficient collision between ball mills leads to poor crushing effect of the ball mill, and when the ball-to-material ratio is greater than 8 (as in Comparative Example 5 in Table 2), the ball-mill crushing effect no longer changes significantly with the change of the ball-to-material ratio, and the ball-to-material ratio is too large. increase cost.

Water-abrasive mass ratio: In the method of the present invention, the water-abrasive mass ratio is 2-32, preferably 8-32, and the water-material ratio is too small, and the abrasive can not be completely submerged in water during the ball milling process, so that the unimmersed abrasive Dry milling and pulverization by direct contact with the grinding balls (similar to putting the CeO2 particles prepared by the hydrothermal method in a ball mill for dry milling and pulverization as mentioned in the background art ₎ will aggravate the hardening of the abrasive. When the water-to-material ratio is too large, the mass of abrasives per unit volume decreases, the frequency of collisions between abrasives and balls decreases, and the crushing effect of ball milling is reduced.

Grinding ball diameter: In the method of the present invention, the grinding ball diameter is 0.5-2mm, preferably 0.5-1.5mm. It is found in the experiment that the grinding ball diameter is too large to aggravate the compaction of the abrasive, even more serious than the compaction before ball milling.

The present invention is by adjusting these several parameters of ball diameter in step (1), ball-abrasive mass ratio, water-abrasive mass ratio, ball milling time, and makes these parameters synergistically improve CeO ₂ The comminution effect of particle, thereby improves _The dispersibility of CeO2 particles in water, and then the preparation of _CeO2 suspension with improved stability and more uniform particle size distribution of _CeO2 particles.

On this basis, the present invention provides a polishing liquid, which comprises 0.015-0.15wt.% of the above-mentioned CeO ₂ suspension and 0.015-0.15wt.% of surfactant in terms of mass percentage. Wherein the surfactant is selected from ionic active agents or non-ionic active agents, preferably piperazine, 2-methylpiperazine, polyvinylpyrrolidone (Polyvinylpyrrolidone, PVP) or polyethylene glycol (Polyethylene Glycol, PEG). PVP can be selected with a molecular weight of not less than 50,000, such as PVP-K30 with a molecular weight of about 58,000 or PVP-K90 with a molecular weight of about 1,300,000.

The method for preparing this polishing solution is to place commercially available _{CeO2nanoparticles} (purchased from Zhongke Leiming (Beijing) Technology Co., Ltd., with an initial particle diameter of about 200nm) and surfactants in a ball mill (QM-3SP4 planetary ball mill) In the ball mill jar, add water and carry out ball milling; All parameters of ball milling process are the same as above _- mentioned preparation CeO in the method for suspension.

The content of the present invention will be described in more detail below in conjunction with specific examples, and the present invention will be further elaborated, but these examples are by no means limiting the present invention.

Embodiment one: _CeO Preparation of suspension

CeO2 suspension was prepared according to the following process, ball milling time, water _- to-material ratio, ball-to-material ratio, and ball diameter were shown in Table ₁ , and commercially available CeO2 nanoparticles (purchased from Zhongke Leiming (Beijing) Technology Co., Ltd., The purchased particle size specification is 200nm) is placed in the ball mill tank of the ball mill (QM-3SP4 planetary ball mill), and water is added for ball milling; wherein, the inner diameter of the ball mill tank is 70mm, the ball mill speed is 400r/min, the ball mill tank and the sun disc of the ball mill The direction of rotation is opposite, and the ratio of the rotating speed of the ball milling tank to the rotating speed of the sun disk is 2; after the ball milling, the liquid in the ball milling tank is CeO ₂ suspension.

Table ₁ CeO suspensions of Examples 1-12

Simultaneously prepare a series of comparative examples according to the CeO suspension, only the ball milling time, water _- to-material ratio, ball-to-material ratio, and ball diameter are shown in Table 2.

Table ₂ CeO Suspensions of Comparative Examples 1-7

Embodiment two: the preparation of polishing liquid

The specific process of preparing the polishing solution is as follows _: commercially available CeO2 nanoparticles (purchased from Zhongke Leiming (Beijing) Technology Co., Ltd., the purchased particle size specification is 200nm) are placed in the ball mill tank of the ball mill (QM-3SP4 planetary ball mill) , add water and carry out ball milling; wherein, the inner diameter of the ball mill tank is 70mm, the speed of the ball mill is 400r/min, the rotation direction of the ball mill tank and the sun disc of the ball mill is opposite, the ratio of the speed of the ball mill tank to the sun disc speed is 2, the ball mill time,, Water-to-material ratio, ball-to-material ratio, and ball diameter are the same as in Example 4 in Example 1. After ball milling, the liquid in the ball mill tank is the polishing liquid. A series of polishing liquids were prepared according to this process. The selection of surfactants in the polishing liquids, the final concentrations of surfactants and CeO2 are shown in Table ₃ .

The polishing liquid of table 3 embodiment

Prepare a series of polishing liquids of comparative examples according to the above process at the same time, only the types of surfactants are shown in Table 4, wherein PVP-K15, PVP-K17, and PVP-K25 represent molecular weights of about 8000, 10000, and 24000 respectively. Left and right PVP.

The polishing liquid of table 4 comparative example

Experiment 1. SEM characterization

Get respectively the CeO2 suspension that embodiment 1-12 obtains with the pipette gun of ₂ μ L ₍ CeO2 concentration is 0.015wt.%), drop on the clean silicon chip, after naturally air-drying, scan electron microscope (Hitachi, the model is SU8220 ) to observe the morphology of the cerium oxide particles on the silicon wafer under the working voltage of 5kV.

With commercially available _{CeO2nanoparticles} (that is, _{CeO2nanoparticles} before ball milling, purchased from Zhongke Leiming (Beijing) Technology Co., Ltd., the particle size specification purchased is 200nm) as a comparison before ball milling, the commercially available _CeO2nm Particles are dispersed in water ( _{CeO2concentration} is 0.015wt.%), ultrasonic 5min to obtain a suspension, drop on a clean silicon wafer, after natural air-drying, use a scanning electron microscope to observe the shape of cerium oxide particles on a silicon wafer under a working voltage of 5kV appearance.

Taking the result of Example 4 as an example, the result is shown in FIG. 1 .

The morphology of CeO ₂ nanoparticles before ball milling (that is, commercially available CeO ₂ nanoparticles) is shown in panel (a) of Fig. 1 . The figure shows that although commercially available CeO ₂ nanoparticles have a particle size of 200nm, they have compacted and agglomerated under the capillary action of air. The agglomerated particles have random and irregular shapes and a length of about 11 μm.

The morphology of the CeO ₂ nanoparticles obtained in Example 4 is shown in panel (b) of FIG. 1 . _The figure shows that the compaction and agglomeration of CeO2 nanoparticles in the suspension after ball milling have been significantly improved, the average particle size is only about 100nm, and the particle size uniformity of the particles has also been significantly improved.

Experiment 2. X-ray Diffraction Characterization of CeO ₂ Crystal Structure in CeO ₂ Suspension

Since long-time ball milling may promote the generation of new material phases, in order to verify whether the crystal structure of cerium oxide changes during ball milling, a scintillation detector (Rigaku Smart Lab) equipped with a graphite monochromator was used to suspend Examples 1-12 The crystal structure of the cerium oxide particles in the solution was characterized by X-ray diffraction (XRD). Specifically _: take 5mL of the CeO2 suspensions of Examples 1-12 in test tubes with a pipette gun, and after natural air drying, take the powder for X-ray diffraction (XRD) characterization. Taking Examples 4 and 7 as examples, the results are shown in Figure 2.

Table ₅ Standard XRD data of CeO2

Table ₅ lists the XRD data of the cubic fluorite structured CeO2 standards. Fig. 2 shows that in the suspension that embodiment 4 and 7 obtains, 9 diffraction peaks all appear in cerium oxide particle, and the quantity of diffraction peak and corresponding angle numerical value (being peak position) all keep consistent; Each of Fig. 2 The corresponding angle of each diffraction peak is compared with the CeO ₂ standard substance comparison, and the angle value difference of each peak is only 0.01 °-0.02 °, and it can be considered that the cerium dioxide particles in the suspension prepared by the method of the present invention still maintain cubic Fluorite structure, no new material phases are produced.

Experiment 3. Polydispersity Index (PD.I)

The cerium oxide suspension obtained in Examples 1-12 and Comparative Examples 1-7 was diluted to the same concentration (3.125wt.%), ultrasonicated for 5 minutes, and vibrated in a thermal vortex mixer at a speed of 1600r/min for 5 minutes, Obtain a sample of the suspension. The Z-average particle size of the cerium oxide particles in the suspension sample was measured by a Malvern laser particle size analyzer (Nano-ZS90, Malvern), and the polydispersity index (PD.I) was calculated.

The results show that the PD.I of the suspensions of the examples and the comparative examples are between 0.08-0.7, meeting the application range of the dynamic light scattering method, so they can be used in subsequent experiments to characterize the particle size-light intensity distribution by the dynamic light scattering method.

Experiment 4. Particle size-light intensity distribution of CeO ₂ suspension

What the dynamic light scattering method measures is the hydrodynamic particle size of the particle, that is, the equivalent spherical diameter that diffuses at the same speed as the sample. The particle size-light intensity distribution curve is obtained by using a multi-exponential analysis model, which indicates the relative contribution rate of particles with different particle sizes to the scattered light intensity.

The particle size _- light intensity distribution curves of the CeO suspensions of Examples 1-12 and Comparative Examples 1-7 were obtained by dynamic light scattering. This experiment is to test the scattered light intensity of each particle in the suspension. The scattered light intensity corresponds to the particle size of the particle. Taking the particle size as the abscissa, the scattered light intensity under the particle size accounts for the percentage of the scattered light intensity of all particle sizes. The particle size-light intensity distribution curve obtained by the coordinates can reflect the proportion of this particle size in all particle sizes. The larger the proportion, the more particles of this particle size in the suspension, that is, the particle size distribution span of the particles in the suspension The smaller the particle, the more uniform it is.

(1), the influence of the diameter of the grinding ball

Taking Example 4, Comparative Example 1, Comparative Example 2, and Comparative Example 3 with grinding ball diameters of 1 mm, 3 mm, 5 mm, and 10 mm as examples, the particle size-light intensity distribution curves are shown in Figure 3 .

The curve in Fig. ₃ represents the relative contribution rate of the CeO2 particle size to the scattered light intensity. Peak 1 is the scattered light intensity peak corresponding to the particle that makes the largest relative contribution to the scattered light intensity. Embodiment 4, comparative example 1, comparative example 2, the peak value of the cerium dioxide suspension of comparative example 3 and its corresponding light intensity ratio are shown in Table 6, and the second row represents CeO in table ₆ Suspension in particle size-light The particle size corresponding to the strongest peak (peak 1) in the intensity distribution curve, and the third column shows the percentage of the peak intensity of peak 1 in the curve to the overall peak intensity.

Table 6 The particle size corresponding to the strongest peak 1 of CeO ₂ suspension and the peak intensity percentage of peak 1

Grinding ball diameter/mm Particle size corresponding to peak 1/nm Proportion of peak intensity/% 1 198.8 94.5 3 292.5 96.7 5 365.2 94.2 10 394.4 96.9

The result of table 6 and Fig. 3 shows, embodiment 4, comparative example 1, comparative example 2, the abrasive grain (being the cerium oxide particle as abrasive material, be called for short abrasive grain) of peak 1 corresponding size of the CeO suspension of comparative example ₃ The proportion of peak intensity is greater than 94%. It can be approximately considered that peak 1 can represent the light intensity distribution peak of the suspension as a whole, that is, the particle size corresponding to peak 1 can be considered to be the most important particle size in the entire suspension. It can be seen from Figure 3 that as the diameter of the ball increases, the particle size value corresponding to peak 1 increases. Not only that, when the diameter of the grinding ball is greater than 1mm, such as 3mm, 5mm and 10mm in diameter, the peak shape of peak ₁ becomes wider, indicating that the range of particle size distribution of CeO2 particles in the suspension is wider, and the range of particle size distribution is wider It shows that the poor uniformity of the particles will lead to poor stability of the suspension, which will affect the polishing removal rate and surface quality.

(2) The influence of water to material ratio

Be respectively 2,4,8,12,16 and 40 embodiment 1, embodiment 2, embodiment 3, embodiment 6, embodiment 7, comparative example 4 with water-material ratio as example, its particle diameter-light intensity The distribution curve is shown in Figure 4. See Table 7 for the peak 1 values and corresponding light intensity ratios of the ceria suspensions in Example 1, Example 2, Example 3, Example 6, Example 7, and Comparative Example 4.

Table 7 The particle size corresponding to the strongest peak 1 of CeO ₂ suspension and the peak intensity percentage of peak 1

Ratio of water to material Particle size corresponding to peak 1/nm Proportion of peak intensity/% 2 163.4 97.3 4 169.5 95.2 8 186.5 95.9 12 196.7 95.1 16 198.8 94.5 40 276.7 96.6

Table 7 and Fig. 4 show, embodiment 1, embodiment 2, embodiment 3, embodiment 6, embodiment 7, comparative example 4 The ratio of the peak intensity of abrasive particles corresponding to the size of peak ₁ of the CeO2 suspension is greater than 94 %, it can be approximately considered that peak 1 can represent the overall light intensity distribution peak of the suspension, that is, the particle size corresponding to peak 1 can be considered to be the most important particle size in the entire suspension. The particle size value corresponding to peak 1 increases with the increase of water-material ratio; at the same time, with the increase of water-material ratio, the peak shape of peak 1 becomes wider. Not only that, when the water-material ratio is 40 (> 32), the peak shape of peak ₁ becomes wider, indicating that the range of particle size distribution of CeO2 particles in the suspension is wider, and the wider range of particle size distribution indicates that the uniformity of particles becomes larger. If it is poor, it will lead to poor stability of the suspension, which will affect the polishing effect and surface quality.

Experiment 5. Stability of CeO ₂ suspension

The suspensions obtained in Examples 1-12 and Comparative Examples 1-7 were left to stand for a period of time, and pictures were taken intermittently to record the stability of the suspension. Take embodiment 4, comparative example 1, comparative example 2 and comparative example 3 as example, the result is as shown in Figure 5; With embodiment 1, embodiment 2, embodiment 3, embodiment 6, embodiment 7, comparative example 4 As an example, the results are shown in Figure 6; taking Example 7, Example 9, Comparative Example 6, and Comparative Example 7 as examples, the results are shown in Figure 7.

Figure 5 shows that the color of the freshly prepared suspension is uniform milky white. After standing for several days, obvious particle sedimentation can be observed in the suspension (comparative example 1-3) of peak 1 peak type in the particle size-light intensity distribution curve; that is: the dispersibility, The stability is poor, and the suspension of Comparative Example 1-3 is obviously stratified after standing for 7 days. This result indicates that too large ball diameter leads to poor stability of the suspension.

Figure 6 shows that the color of the freshly prepared suspension is uniform milky white. After standing for 7 days, in the particle size-light intensity distribution curve, peak 1 peak type wider suspension (comparative example 4) can observe obvious particle sedimentation, and stratification phenomenon is obvious; Namely: the dispersibility of comparative example 4 suspension , Poor stability. The results indicate that too large a water-to-material ratio leads to poor suspension stability.

Figure 7 shows that the color of the freshly prepared suspension is uniform milky white. After standing for 7 days, obvious particle settlement can be observed in the suspensions of Comparative Example 6 and Comparative Example 7, and the layering phenomenon is obvious; while the thickness of the suspension layers of Example 7 is almost zero, the suspension of Example 9 Obvious delamination is basically not observed; that is: the dispersibility and stability of the suspensions of Comparative Example 6 and Comparative Example 7 are poor; while the dispersibility and stability of the suspensions of Example 7 and Example 9 are good. This result indicates that too short ball milling time will lead to poor stability of the suspension.

Experiment 6. _The removal rate and removal selectivity ratio of CeO2 suspension in chemical mechanical polishing

作为SiO₂和Si₃N₄的去除速率(

例如，SiO₂的去除速率＝(抛光前SiO₂膜层的厚度-抛光后SiO₂膜层的厚度)/抛光时间)。SiO₂空白晶圆片的结构是以硅作底层，在底层的表面上沉积一层氧化硅膜；Si₃N₄空白晶圆片同理，是以硅作底层，在底层的表面上沉积一层氮化硅膜，以下实验测试或观察的均是氧化硅膜层和氮化硅膜层的厚度或形貌。以实施例5的悬浮液样品为例，以市售含氧化铈的抛光液(购自ASAHI GLASS CO.,LTD.)作对照，结果如图8所示。Dilute the cerium _oxide suspension of the example with water to CeO2 concentrations of 0.15wt.%, 0.03wt.%, 0.015wt.%, 0.0075wt.%, respectively, as a suspension sample, and then used to polish _SiO2 and Si ₃ N ₄ blank wafers. During the polishing process, the suspension sample was transported to the polishing pad of a chemical mechanical polishing machine (purchased from Huahai Qingke Co., Ltd., model Universal 150) under the stirring of a magnetic stirring bar (to maintain the stability of the suspension), and used The suspension samples were polished on _{2 -} inch _SiO2 and Si3N4 blank wafers, respectively. The parameters of the polishing process refer to the standard of STI CMP in actual production. The flow rate of the suspension sample is 150mL/min, the polishing pad is DH3002-T80D30-S20M3S1, the polishing pressure on SiO ₂ and Si ₃ N ₄ blank wafers is about 3 psi, and each experiment is repeated 3 times. The surface morphology of SiO ₂ and Si ₃ N ₄ blank wafers before and after polishing was characterized by atomic force microscopy (AFM, DimensionICON of Bruker). Thickness difference of SiO ₂ film layer and Si ₃ N ₄ film layer before and after polishing with unit polishing time

As the removal rate of SiO ₂ and Si ₃ N ₄ (

For example, the removal rate of SiO ₂ = (thickness of SiO ₂ film before polishing - thickness of SiO ₂ film after polishing)/polishing time). The structure of the SiO ₂ blank wafer is based on silicon as the bottom layer, and a layer of silicon oxide film is deposited on the surface of the bottom layer; the same is true for the Si ₃ N ₄ blank wafer, which uses silicon as the bottom layer and deposits a layer of silicon oxide film on the surface of the bottom layer. Silicon nitride film, the thickness or morphology of the silicon oxide film layer and silicon nitride film layer are all tested or observed in the following experiments. Taking the suspension sample in Example 5 as an example, a commercially available polishing solution containing cerium oxide (purchased from ASAHI GLASS CO., LTD.) was used as a control, and the results are shown in FIG. 8 .

是市售氧化铈(0.25wt.％)悬浮液(约为

)的7倍，浓度为0.03wt.％的氧化铈悬浮液对氧化硅的去除速率约为

是市售氧化铈(0.25wt.％)悬浮液(约为

)的4倍。实际生产要求抛光液对SiO₂的去除速率达到

左右即可，可见本发明的氧化铈悬浮液在0.03wt.％以上的浓度就可达到该去除速率，在保证去除速率的同时节约了成本。Figure 8 shows that the greater the concentration of cerium oxide particles, the more cerium oxide particles in contact with SiO ₂ and Si ₃ N ₄ blank wafers per unit time, and the higher the polishing removal rate. The concentration of 0.15wt.% cerium oxide suspension to silicon oxide removal rate is about

is a commercially available cerium oxide (0.25wt.%) suspension (approx.

) 7 times that of cerium oxide suspension with a concentration of 0.03wt.%, the removal rate of silicon oxide is about

is a commercially available cerium oxide (0.25wt.%) suspension (approx.

) 4 times. The actual production requires that the removal rate of _SiO2 by the polishing liquid reaches

It can be seen that the cerium oxide suspension of the present invention can achieve the removal rate at a concentration above 0.03wt.%, which saves costs while ensuring the removal rate.

According to the removal rate obtained in Figure 8, the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ can be calculated, that is, the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ = removal rate of SiO ₂ / removal rate of Si ₃ N ₄ , The results are shown in Figure 9.

Figure 9 shows that as the concentration of cerium oxide decreases, the removal rates of SiO ₂ and Si ₃ N ₄ decrease correspondingly, and the removal rate of SiO ₂ decreases faster, so the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ Also decreased to some extent, but still more than 11, equivalent to 14 of the removal selectivity ratio of commercially available cerium oxide suspension; When the concentration of cerium oxide was 0.03wt.%, the removal selectivity ratio could reach 14, comparable to that of commercially available cerium oxide suspension (oxidized cerium concentration is 0.25wt.%) equivalent; when cerium oxide concentration is 0.15wt.%, removal selectivity is 16, is better than commercially available cerium oxide suspension (cerium oxide concentration is 0.25wt.%) Remove the selection ratio.

Experiment 7. The surface roughness of the film layer after polishing the blank wafer with CeO ₂ suspension

Use AFM equipment to observe the surface roughness of SiO ₂ and Si ₃ N ₄ blank wafers in the range of 10×10 μm after polishing in Experiment 7, and the results are shown in Figure 10.

Figure 10 shows that after polishing with the cerium oxide suspension of different concentrations of Example 5, the roughness of the silicon dioxide film layer surface is all lower than 0.3nm, which is far lower than the silicon dioxide film layer after many reports of cerium oxide suspension polishing Surface roughness (about 1nm, see Lee S H, Lu Z, Babu S V. et al. "Chemical mechanical polishing of thermaloxide films using silica particles coated with ceria" [J]. Journal of Materials Research, 2002, 17 (10 ): 2744-2749.), show that after polishing with the cerium oxide suspension of the present invention, the surface of the silicon dioxide film layer of the blank wafer is smooth and smooth, with few scratches, few defects, and high surface quality.

Experiment 8. Thickness uniformity of the film layer after polishing a blank wafer with CeO ₂ suspension

Use the optical film thickness measuring instrument Filmetrics 50 (F50) to measure the thickness of the silicon oxide film layer and the silicon nitride film layer on the surface of the blank wafer after experiment 7 polishing, specifically: measure 81 points symmetrically along the center of the blank wafer The thickness of the silicon oxide film layer or silicon nitride film layer, the results are shown in Figure 11. At the same time, a distribution map of the thickness of the silicon oxide film or silicon nitride film on the surface of the blank wafer after polishing is generated, and the result is shown in Figure 12. In this figure, the darker the color, the greater the thickness.

Regardless of whether it is before or after polishing, there are individual points on the surface of each blank wafer, such as particle residue, a small amount of doping during coating, and uneven polishing. These individual points will have inaccurate situations such as higher thickness after polishing than before polishing, measured goodness of fit (GOF) less than 0.99, over-polishing in local areas, etc. All such data points will be eliminated to ensure the accuracy of data measurement Above 99%.

Figure 11(a) is the SiO ₂ film surface of the SiO ₂ blank wafer, the lines in this figure are scratches, and there are few visible scratches; (b) shows the Si ₃ N ₄ blank wafer The surface of the Si ₃ N ₄ film layer of the wafer, there are basically no lines in the figure, indicating that the surface of the Si ₃ N ₄ film layer has no scratches, few defects, and good surface quality. In a word, after being polished by the suspension of the present invention, the surface of the SiO ₂ film layer and the Si ₃ N ₄ film layer of the blank wafer is relatively smooth and flat, with few scratches and defects, and good surface quality.

Figure 12 shows that after polishing with different concentrations of CeO ₂ suspensions, the color distribution on the surface of SiO ₂ and Si ₃ N ₄ blank wafers is relatively uniform, which means that the polished SiO ₂ film layer and Si ₃ N ₄ The thickness of the film on the surface of the film layer is relatively uniform. Among them, when the concentration of cerium oxide is 0.15wt.%, the visual field is basically blue and the depth of blue is basically the same, which shows that there is a silicon oxide film on the polished SiO ₂ blank wafer, and the Si ₃ N ₄ blank There is a silicon nitride film layer on the wafer, and the thickness distribution of the SiO ₂ film layer and the Si ₃ N ₄ film layer is quite uniform. In contrast, when the CeO ₂ concentration drops below 0.015wt.%, the center and edge thickness difference (i.e. color difference) of SiO ₂ film increases, and the thickness of Si ₃ N ₄ film is from the edge to the center, presenting The phenomenon of "thin-thick-thin" (the color difference appears "shallow-deep-shallow"), that is, the uniformity of the suspension with a CeO2 concentration above 0.03wt.% used to polish a blank wafer is better _than that at a concentration of 0.015 Suspensions below wt.%.

Experiment 9. Stability experiment of polishing fluid

(one)

Select the polishing liquid of Examples 2-1 to 2-4 in Table 3 of Example Two to stand still, and observe the state of the polishing liquid on the 0th day, the 1st day, the 3rd day, and the 7th day when standing respectively; The same method as in Example 2-1 was used to prepare the polishing solutions of Comparative Examples 2-1 to 2-7 in Table 4 of Example 2, except that the selected surfactants were different, and the same experiment was carried out. Wherein, the surfactants in the polishing liquids of Comparative Examples 2-1 to 2-4 are ionic surfactants, which are respectively SDS (sodium dodecyl sulfate), piperazine-2-carboxylic acid dihydrochloride, hexametaphosphoric acid Sodium and glycine hydrochloric acid; The surfactant in the polishing liquid of comparative example 2-5～2-7 is nonionic surfactant, is respectively PAM (polyacrylamide), PVA-124 (polyvinyl alcohol PVA-124), PPG (polypropylene glycol). The photographs of Example 2-1, Example 2-2, Comparative Example 2-1, Comparative Example 2-2, Comparative Example 2-3, and Comparative Example 2-4 after the polishing solution is left standing are shown in Figure 14, corresponding to Figure 14 respectively. B, F, A, C, D, E in; embodiment 2-3, embodiment 2-4, comparative example 2-5, comparative example 2-6, comparative example 2-7 state after the polishing liquid is left standstill The photos are shown in Figure 15, corresponding to C, A, B, D, and E in the figure.

Figure 14 shows the states of the polishing fluids containing six different ionic surfactants after standing, that is, the polishing fluids containing different ionic surfactants are respectively on the 0th day, the 1st day, and the 3rd day after standing still. Different degrees of delamination occurred on the 7th day and the 7th day. According to the clearer and more transparent upper liquid, the larger the thickness of the layer, the faster the cerium oxide particles settled, the worse the dispersibility of the cerium oxide particles, and the poorer the stability of the polishing solution. Standard, these six kinds of polishing fluids are sorted by stability from low to high as: Comparative Example 2-2 (piperazine-2-carboxylic acid dihydrochloric acid) < Comparative Example 2-1 (SDS) < Comparative Example 2-4 ( Glycine hydrochloride)<Comparative example 2-3 (sodium hexametaphosphate)<Example 2-1 (piperazine)<Example 2-1 (2-methylpiperazine). This result shows that the polishing fluids of Example 2-1 and Example 2-2 have the best stability, and there is no obvious sedimentation after standing for 7 days. However, the polishing fluids containing the other four ionic surfactants all had significant sedimentation after standing for a few minutes or within 1 day.

Figure 15 shows the states of the polishing fluids containing five different nonionic surfactants after standing, that is, the polishing fluids containing different nonionic surfactants were left standing on the 0th day, the 1st day, and the 1st day respectively. Different degrees of delamination appeared on the 3rd and 7th days. According to the clearer and more transparent upper layer liquid, the greater the thickness of the layer, the faster the cerium oxide particles settle, the worse the dispersibility of the cerium oxide particles, and the worse the stability of the polishing solution. One standard, these five kinds of polishing fluids are sorted by stability from low to high as: Comparative Example 2-5 (PAM) <Comparative Example 2-7 (PPG) <Comparative Example 2-6 (PVA-124) <Example 2-4 (PEG)<Example 2-3 (PVP-K30). This result shows that the polishing fluids of Examples 2-4 and 2-3 have the best stability, and there is no obvious sedimentation after standing for 7 days. However, the polishing solutions containing the other three non-ionic surfactants all showed significant sedimentation after standing for a few minutes or within 1 day.

(two)

Select the polishing liquid of embodiment 2-3, 2-5 in the table 3 of embodiment two and the comparative example 2-8～2-10 in the table 4 to carry out above-mentioned standing experiment, these five kinds of polishing liquids all select PVP as surface for use Active agent, the only difference is the molecular weight of PVP, the corresponding molecular weight are: 58000 (PVP-K30), 1300000 (PVP-K90), 8000 (PVP-K15), 10000 (PVP-K17) and 24000 (PVP-K25 ). Observe the states of these five polishing solutions on the 0th day, the 1st day, the 3rd day, and the 7th day respectively. The results are shown in Figure 13, corresponding to D, E, A, B, and C in the figure.

Figure 13 shows the states of the polishing fluids containing five different molecular weight PVPs after standing, and the polishing fluids containing different molecular weights of PVP appeared in different degrees on the 0th day, the 1st day, the 3rd day, and the 7th day. The layering, and as the molecular weight decreases (from 1,300,000 to 8,000), the clearer and more transparent the upper layer liquid, the larger the thickness of the layer, indicating that the faster the cerium oxide particles settle, the worse the dispersibility of the cerium oxide particles, and the stability of the polishing liquid Worse, when the molecular weight of PVP was 58000 (i.e. PVP-K30), the thickness of the polishing fluid stratified after standing for 7 days was almost zero; No obvious delamination was observed in Tianhou polishing fluid. The result shows that when the molecular weight of PVP is ≥50000, the stability of the formed polishing liquid is good.

Experiment 10. Morphology of Polishing Fluid

(one)

Take respectively the polishing liquids of Examples 2-1～2-4 and Comparative Examples 2-5～2-7 left standing for 7 days in Experiment 9, and use a scanning electron microscope (Hitachi, model SU8220) to observe the 5kV working conditions on a silicon wafer. The morphology of the polishing solution under the voltage (the specific process is the same as Experiment 1), the results are shown in Figures 16 and 17.

The morphology of the abrasive grains (that is, cerium oxide particles as abrasive, referred to as abrasive grains) in the polishing liquid of embodiment 2-1 (piperazine) and embodiment 2-2 (2-methylpiperazine) is shown in Fig. 16 respectively. (a) and (b), the morphology of abrasive grains in the CeO suspension without surfactant is shown in (c) in ₁₆ . Figure 16 shows that the particle diameters of the abrasive particles in the polishing fluid of Example 2-1 (piperazine) and example 2-2 (2-methylpiperazine) are close, being 30-90nm and 50-120nm respectively, which are slightly smaller than CeO ₂ The particle diameter (90-150nm) of the abrasive grains in the suspension shows that the gap between the abrasive grains in the polishing liquid of the present invention is larger than when no surfactant is added (i.e. CeO ₂ suspensions), indicating that the two ionic types The surfactant can disperse the cerium oxide particles, thereby improving the stability of the polishing liquid.

The morphology of abrasive grains in the polishing fluid of Example 2-3 (PVP-K30) and Example 2-4 (PEG) is shown in (d) and (e) panels in Figure 17, respectively, and comparative examples 2-5-2 The morphology of the abrasive particles in the -7 polishing fluid is shown in (a), (b) and (c) of Figure 17, respectively. Figure 17 shows that the particle diameter of a single abrasive grain in the polishing liquid of Example 2-3 (PVP-K30) is 55-60nm, the surface of the abrasive grain is completely covered by PVP-K30, there is a significant separation between the individual abrasive grains, and the abrasive grains Good dispersion. This is because PVP-K30 blocks the abrasive particles from the air by covering them, avoiding the influence of capillary action on the abrasive particles. In Example 2-4 (PEG), the particle size of a single abrasive grain in the polishing liquid is 60-150nm, and the surface of the abrasive grain is partially covered by PEG. Although some abrasive grains are still agglomerated due to capillary action, the dispersion of the abrasive grains is still limited. improve. Compared with the CeO ₂ suspension without surfactant, since no surfactant was added, the particle surface was not covered by any substance, and the particle size of a single abrasive particle was 90-150nm, and the abrasive particles were agglomerated due to capillary action. It can be seen that adding PVP _- K30 or PEG to the polishing solution can improve the dispersion of CeO2 particles.

Figure 17 also shows that although the PAM in the polishing fluid of Comparative Example 2-5 (PAM) isolates the ceria particles from each other, the surface of the ceria is wrapped by a gelatinous viscous substance, and the viscous substance aggravates the ceria particles. The bonding between them resulted in a compaction of about 3 μm, indicating that the dispersion of abrasive grains in the polishing solution was poor. In the polishing liquid of comparative example 2-6 (PVA-124), the cerium oxide particles are wrapped and separated by strip-shaped PVA-124, and the length of the hardened object is about 8 μm, which shows that the dispersion of abrasive grains in the polishing liquid is also poor. When the polishing liquid of Comparative Example 2-7 (PPG) was left to stand for 7 days, almost all CeO ₂ particles settled to the bottom, and no CeO ₂ particles were seen in the upper liquid, only about 11 μm flakes, indicating that the abrasive grains were in the polishing liquid very poor dispersion.

In addition, the experimental results also show that in the polishing liquid comparative examples 2-5 (PAM), 2-6 (PVA-124) and comparative examples 2-7 (PPG) containing the same concentration (3wt.%) surfactant, The greater the molecular weight of the surfactant (respectively 2 million to 14 million, 105000, 400 to 2000), the greater the viscosity of the formed polishing liquid, the lowest is about 50mPa·s, the polishing liquid is gel _- like, and the CeO2 particles are "Fixed" in the jelly, there is no electrostatic force between each other, which may be the reason for the poor dispersion of abrasive particles in these kinds of polishing fluids. However, at the same concentration (3wt.%) of the polishing fluid Example 2-3 (PVP-K30) and Example 2-4 (PEG) containing PVP-K30 and PEG, the molecular weights are 58000 and 9000-12500 respectively, forming The maximum viscosity of the polishing liquid is about 10mPa·s, the polishing liquid is in the form of a solution, and there is no phenomenon of gelation; and the greater the molecular weight and concentration of PVP, the better the dispersion effect on CeO ₂ particles, which verifies The dispersion mechanism of PVP to CeO ₂ is understood: when the molecular weight of PVP increases gradually, the number of monomers of PVP increases, and the "C=O" bond that can form hydrogen bonds with CeO ₂ also increases, so that the CeO ₂ particles are better dispersed Come. Similarly, when the concentration of PVP increases, the number of "C=O" bonds that form hydrogen bonds with CeO ₂ also increases, thereby improving the dispersion effect of PVP on CeO ₂ . For PEG, the oxygen atoms on the PEG "CO" bond and the oxygen atoms in ceria can also be connected by hydrogen bonds, and finally form the structure of "ceria-PEG-ceria", which may be this structure It has a better dispersion effect on ceria, although its dispersion effect is not as good as that of PVP-K30.

(two)

After standing still for the 7th day, the surface and bottom liquids of the polishing liquids in Examples 2-1 to 2-4 were taken respectively, as in Experiment 1, and after natural air drying, the morphology of abrasive grains under a working voltage of 5kV was observed with a scanning electron microscope.

The result shows that embodiment 2-1 (piperazine) and embodiment 2-2 (2-methylpiperazine) polishing liquid are positioned at the particle diameter of the single abrasive particle of surface layer and bottom all below 500nm, the single abrasive particle of surface layer and bottom The difference in particle diameter all remains within 50-300nm; The particle diameter of the individual abrasive grains that embodiment 2-3 (PVP-K30) and embodiment 2-4 (PEG) polishing liquid is positioned at surface layer and bottom also is below 500nm, surface layer and The single abrasive particle diameter difference at the bottom is kept within 0-210nm, especially the single abrasive particle diameter difference between the surface layer and the bottom of the polishing liquid in Example 2-3 (PVP-K30) is only 30-110nm, and the abrasive particle diameter The difference is small, indicating that after standing for 7 days, the abrasive particles in the polishing solution of the present invention are still evenly distributed, and basically no agglomeration and hardening have occurred.

Experiment 11: The removal rate and removal selectivity ratio of polishing fluid to silicon dioxide and silicon nitride

作为SiO₂或Si₃N₄的去除速率(

例如，Si₃N₄的去除速率＝(抛光前Si₃N₄膜层的厚度-抛光后Si₃N₄膜层的厚度)/抛光时间)。根据上述方法得到的去除速率，计算出SiO₂/Si₃N₄的去除选择比，即SiO₂/Si₃N₄的去除选择比＝SiO₂的去除速率/Si₃N₄的去除速率。Same as Experiment 6, the SiO ₂ blank wafer and the Si ₃ N ₄ blank wafer were respectively polished with the polishing solutions of Examples 2-1 to 2-4 with different pH values. During the polishing process, the polishing liquid was transported to the polishing pad of a chemical mechanical polishing machine (purchased from Huahai Qingke Co., Ltd., model Universal150) under the stirring of the magnetic stirring bar (to maintain the stability of the polishing liquid), and the The 2-inch SiO ₂ and Si ₃ N ₄ blank wafers were polished separately. The parameters of the polishing process refer to the standard of STICMP in actual production. The flow rate of the polishing liquid is 150mL/min, the polishing pad is DH3002-T80D30-S20M3S1, the polishing pressure for SiO ₂ and Si ₃ N ₄ blank wafers is 3psi, and each experiment is repeated 3 times. The surface morphology of SiO ₂ blank wafers and Si ₃ N ₄ blank wafers before and after polishing was characterized by atomic force microscopy (AFM, DimensionICON of Bruker). Thickness difference of SiO ₂ film layer or Si ₃ N ₄ film layer before and after polishing with unit polishing time

As the removal rate of SiO ₂ or Si ₃ N ₄ (

For example, the removal rate of Si ₃ N ₄ = (thickness of Si ₃ N ₄ film before polishing - thickness of Si ₃ N ₄ film after polishing)/polishing time). According to the removal rate obtained by the above method, the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ is calculated, that is, the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ = removal rate of SiO ₂ /removal rate of Si ₃ N ₄ .

The results of Si ₃ N ₄ removal rate and removal selectivity ratio of embodiment 2-1 (piperazine) and embodiment 2-2 (2-methylpiperazine) polishing solution are shown in Figure 18; embodiment 2-4 (PEG) The results of Si ₃ N ₄ removal rate and removal selectivity ratio of the polishing solution of Example 2-3 (PVP-K30) are shown in FIG. 20 .

以下，显著低于CeO₂悬浮液，表明哌嗪的加入在酸性和碱性条件下(pH＝4或8-12)均可显著抑制氮化硅的去除。在pH值为4、6、10、12时，实施例2-2(2-甲基哌嗪)的抛光液对Si₃N₄的去除速率均在

以下，显著低于CeO₂悬浮液，表明2-甲基哌嗪的加入在酸性、弱酸性和碱性条件下(pH＝4-6或8-12)均可显著抑制氮化硅的去除。The results in Figure 18 show that when the pH values are 4, 10, and 12, the removal rates of Si ₃ N ₄ in the polishing solution of Example 2-1 (piperazine) are all within

Below, significantly lower _than that of the CeO2 suspension, indicating that the addition of piperazine can significantly inhibit the removal of SiN under both acidic and basic conditions (pH = 4 or 8–12). When the pH value is 4, 6, 10, 12, the removal rate of the polishing solution of embodiment 2-2 (2-methylpiperazine) to Si ₃ N ₄ is in

Below, significantly lower than that of the CeO2 suspension, indicating that the addition of ₂ -methylpiperazine can significantly inhibit the removal of silicon nitride under acidic, weakly acidic and basic conditions (pH = 4–6 or 8–12).

以下，显著低于CeO₂悬浮液，表明PVP-K30的加入在酸性、中性和碱性的条件下(pH＝4-12)均可显著抑制氮化硅的去除，在酸性和碱性条件下(pH＝4或8-12)可显著提高SiO₂/Si₃N₄的去除选择比，进而提高抛光液的抛光效果。在pH值为6、10、12时，实施例2-4(PEG)的抛光液对Si₃N₄的去除速率均在

以下，显著低于CeO₂悬浮液，该抛光液在pH值为7-12时的SiO₂/Si₃N₄的去除选择比显著高于CeO₂悬浮液，表明PEG的加入在碱性条件下(pH＝8-12)可显著抑制氮化硅的去除，从而显著提高SiO₂/Si₃N₄的去除选择比，进而提高抛光液的抛光效果。The results in Figure 20 show that when the pH values are 4, 6, 7, 10, and 12, the removal rates of Si ₃ N ₄ by the polishing solution of Example 2-3 (PVP-K30) are all in the

Below, significantly lower _than the CeO2 suspension, indicating that the addition of PVP-K30 can significantly inhibit the removal of silicon nitride under acidic, neutral and alkaline conditions (pH = 4-12), and in acidic and alkaline conditions Under low pH (pH=4 or 8-12), the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ can be significantly increased, thereby improving the polishing effect of the polishing solution. When the pH value is 6, 10, 12, the removal rate of Si ₃ N ₄ by the polishing solution of embodiment 2-4 (PEG) is all in

Below, significantly lower _than the CeO2 suspension, the removal selectivity ratio of _SiO2 / _{Si3N4 of} the polishing solution at pH 7-12 was significantly higher _than that of the CeO2 suspension, indicating that the addition of PEG was in alkaline condition (pH=8-12) can significantly inhibit the removal of silicon nitride, thereby significantly increasing the removal selectivity ratio of SiO ₂ /Si ₃ N ₄ , and further improving the polishing effect of the polishing liquid.

Experiment 12: The surface quality of the film layer after polishing the blank wafer with polishing liquid

Same as Experiment 7, use AFM equipment to observe the number of scratches on the silicon oxide film and silicon nitride film in the range of 10×10 μm on the surface of the blank wafer after polishing in Experiment 11, and the results are shown in Figures 19 and 21.

Figure 19 shows that when the pH values are 6, 7, 10, and 12, the number of scratches on the silicon oxide film layer after polishing with the polishing solution of Example 2-1 (piperazine) is no more than 10, and the silicon nitride film The number of scratches in the layer is no more than 4, and at least the number of scratches in the silicon oxide film layer and/or silicon nitride film layer is significantly lower _than that of the CeO2 suspension, indicating that the addition of piperazine is effective in weakly acidic, neutral and Under alkaline conditions (pH=6-12), the surface defects of SiO ₂ or Si ₃ N ₄ blank wafers after polishing can be significantly reduced, thereby significantly improving the surface of SiO ₂ or Si ₃ N ₄ blank wafers after polishing Quality, thereby improving the polishing effect of the polishing liquid. When the pH value was 6, 7, 10, and 12, the number of scratches on the silicon oxide film layer after polishing with the polishing solution of Example 2-2 (2-methylpiperazine) was no more than 11, and the silicon nitride film The number of scratches in the layer is no more than 4, and at least the number of scratches in the silicon oxide film layer and/or silicon nitride film layer is significantly lower than that of the CeO2 suspension, indicating that the addition of ₂ -methylpiperazine is effective in weak acidity. , neutral and alkaline conditions (pH=6-12) can significantly reduce the surface defects of polished SiO ₂ or Si ₃ N ₄ blank wafers, thereby significantly improving the polished SiO ₂ or Si ₃ N ₄ blank wafers The surface quality of the wafer can be improved, thereby improving the polishing effect of the polishing solution.

Figure 21 shows that when the pH value is 4, 6, 7, 10, 12, the number of scratches on the silicon oxide film layer after polishing with the polishing solution of Example 2-3 (PVP-K30) is no more than 11, nitrogen The number of scratches in the silicon oxide film layer is no more than 4, at least in the number of scratches in the silicon oxide film layer and/or silicon nitride film layer is significantly lower than that of CeO ₂ suspensions, indicating that the addition of PVP-K30 is effective in acidic, Under neutral and alkaline conditions (pH=4-12), the surface defects of SiO ₂ or Si ₃ N ₄ blank wafers after polishing can be significantly reduced, thereby significantly improving the quality of SiO ₂ or Si ₃ N ₄ blank wafers after polishing. The surface quality of the chip can be improved, thereby improving the polishing effect of the polishing liquid. When the pH value was 4, 6, 7, 10, 12, the number of scratches on the silicon oxide film layer after polishing with embodiment 2-4 (PEG) polishing solution was no more than 11, and the scratches on the silicon nitride film layer The number of scratches is no more than 3, at least in the number of scratches in the silicon oxide film layer and/or silicon nitride film layer is significantly lower _than that of the CeO2 suspension, indicating that the addition of PEG is under acidic, neutral and alkaline conditions ( pH=4-12) can significantly reduce the surface defects of SiO ₂ or Si ₃ N ₄ blank wafers after polishing, thereby significantly improving the surface quality of SiO ₂ or Si ₃ N ₄ blank wafers after polishing, and then improve the polishing Liquid polishing effect.

In addition, when pH value is 4,6,7,10,12, with the SiO after embodiment _2-3 (PVP-K30) and embodiment 2-4 (PEG) polishing liquid polishing The surface roughness of the film layer all Maintained at 0.20-0.32nm, it shows that these two polishing solutions can meet the requirement of SiO ₂ film surface roughness less than 1nm and smooth surface of picometer level under acidic, neutral and alkaline conditions. When pH value is 4,6,10, the polishing liquid of embodiment 2-3 (PVP-K30) and embodiment 2-4 (PEG) are compared with cerium oxide suspension, after polishing SiO _2The surface roughness of film layer The degree is significantly reduced, indicating that the addition of PVP-K30 and PEG can significantly improve the surface quality of the polished _SiO2 blank wafer, and then improve the polishing effect of the polishing solution.

When the pH value was 4, 6, 7, 10, 12, the surface roughness of the Si ₃ N ₄ film layer after polishing with embodiment 2-3 (PVP-K30) and embodiment 2-4 (PEG) polishing liquid was uniform. Maintained at 0.08-0.16nm (respectively 0.10-0.16nm and 0.08-0.19nm), indicating that the two polishing solutions can meet the requirement of surface roughness less than 1nm under acidic, neutral and alkaline conditions. Especially when the pH value=4, the surface roughness of the Si ₃ N ₄ film layer after polishing with the polishing liquid of embodiment 2-3 (PVP-K30) and embodiment 2-4 (PEG) is the lowest, is respectively 0.11 nm and 0.08nm, significantly lower than the ceria suspension, about 200pm and 250pm lower than the maximum scratch depth polished with the CeO2 suspension, indicating that the addition _of PVP _- K30 and PEG can significantly reduce the _Si3N4 void crystals after polishing The surface defects of the wafer, thereby significantly improving the surface quality of the Si ₃ N ₄ blank wafer after polishing, and then improving the polishing effect of the polishing solution.

The above is only a preferred embodiment of the present invention, it should be pointed out that, for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, these improvements and Retouching should also be considered as part of the invention.

Claims (18)

1. a polishing liquid, is characterized in that, described polishing liquid is made up of CeO suspension and tensio _- active agent; Wherein, the _CeO2 suspension is composed of _CeO2 particles and water, the water _- material ratio of the CeO2 suspension is 2-32, and the _CeO2 suspension is prepared by ball milling for 6-12 hours, and the surfactant is PEG; Wherein, the concentration of the CeO2 is 0.015wt.%- _0.075wt %, and the concentration of the surfactant is 0.015wt.%-0.045wt.%; Wherein, the particle size of the CeO2 particles is _30-150nm ; After the polishing liquid polishes the silicon dioxide blank wafer and the silicon nitride blank wafer respectively, the number of scratches on the surface of the silicon dioxide blank wafer is ≤ 11, and the number of scratches on the surface of the silicon nitride blank wafer is less than or equal to 11. When the number of traces is ≤ 3, the pH value of the polishing solution is 4-12. 2. The polishing liquid according to claim ₁ , characterized in that, the CeO Concentration is equal to the concentration of surfactant. 3. The polishing liquid according to claim ₁ , wherein the CeO2 particles have a particle diameter of 30-120nm. 4. The polishing liquid according to claim ₃ , wherein the CeO2 particles have a particle diameter of 30-90nm. 5. The polishing liquid according to claim 4, characterized in that, the CeO suspension was left standing for ₇ days without significant precipitation. 6. according to the arbitrary described polishing liquid of claim 1-2, it is characterized in that, the removal rate of described polishing liquid to silicon nitride is

pH 4 or pH 7-12. 7. according to the arbitrary described polishing liquid of claim 1-2, it is characterized in that, the removal rate of described polishing liquid to silicon nitride is

The pH is 10-12. 8. The polishing liquid according to any one of claims 1-2, characterized in that, the removal selectivity ratio of the polishing liquid to silicon dioxide and silicon nitride is 20-30, and the pH value is 4 or 12. 9. according to the described polishing fluid of any one of claim 1-2, it is characterized in that, after described polishing fluid is polished silicon dioxide blank wafer and silicon nitride blank wafer respectively, silicon dioxide blank wafer When the number of scratches on the wafer surface is ≤8, and the number of scratches on the surface of a silicon nitride blank wafer is ≤2, the pH value is 8-12. 10. according to the preparation method of the described polishing liquid of any one of claim _1-2 , it is characterized in that, CeO granule, tensio-active agent and water are added in the ball mill and carry out wet grinding pulverization, the liquid obtained after pulverization is the described Polishing liquid, the diameter of the grinding ball in the ball mill is 0.5-2mm when crushing. 11. The preparation method according to claim 10, wherein the diameter of the balls in the ball mill is 0.5-1.5mm during pulverization. 12. according to the described preparation method of claim 10, it is characterized in that, the grinding ball in the ball mill and CeO2 mass ratio of particle is _2-8 when pulverizing. 13. according to the described preparation method of claim 12, it is characterized in that, when pulverizing, the ball in the ball mill and CeO _The mass ratio of particle is 5-7. 14. The preparation method according to claim ₁₀ , characterized in that the mass ratio of water to CeO2 particles is 2-32 during pulverization. 15. The preparation method according to claim ₁₄ , characterized in that the mass ratio of water to CeO2 particles is 8-32 during pulverization. 16. The preparation method according to claim 10, characterized in that, wherein the pulverization time is 6-12h. 17. The preparation method according to claim 16, wherein the pulverization time is 10-12h. 18. The application of the polishing fluid according to any one of claims 1-5 in the polishing fluid for chemical mechanical polishing.

Images