Patent Application
20250201633
With enhanced performance requirement for more compact, miniaturized, and high-density semiconductor packages, considerable attention has been given to heat dissipation problem of the semiconductor packages. Efficient removal of heat from semiconductor packages has become a crucial issue to offer promising performance without functional failure. While heat fluxes in semiconductor packages have been significantly increased, thermal design margins are continuing to decline for heat dissipation, which has led to an increase in demand for advanced thermal conductive materials. Several ceramic-based materials can greatly meet these requirements, while offering dielectric isolation to prevent electrical short. However, currently there is lack of an accurate and reliable thermal characterization technique for monitoring and analyzing thermal characteristics of the thermal conductive materials during manufacturing.
Aspects of the present disclosure are best understood from the following detailed description when read with the accompanying figures. It is noted that, in accordance with the standard practice in the industry, various features are not drawn to scale. In fact, the dimensions of the various features may be arbitrarily increased or reduced for clarity of discussion.
The following disclosure provides many different embodiments or examples, for implementing different features of the provided subject matter. Specific examples of components and arrangements are described below to simplify the present disclosure. These are, of course, merely examples and are not intended to be limiting. For example, the formation of a first feature over or on a second feature in the description that follows may include embodiments in which the first and second features are formed in direct contact, and may also include embodiments in which additional features may be formed between the first and second features, such that the first and second features may not be in direct contact. In addition, the present disclosure may repeat reference numerals and/or letters in the various examples. This repetition is for the purpose of simplicity and clarity and does not in itself dictate a relationship between the various embodiments and/or configurations discussed.
Further, spatially relative terms, such as “beneath,” “below,” “lower,” “above,” “upper” and the like, may be used herein for ease of description to describe one element or feature's relationship to another element(s) or feature(s) as illustrated in the figures. The spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. The apparatus may be otherwise oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein may likewise be interpreted accordingly.
The present disclosure provides a solution for characterizing a thermal conductive layer during manufacturing of a semiconductor package.
The XRD apparatus 10 includes an X-ray emission module 100 configured to emit X-ray beams to a workpiece 102. The incident X-ray beams may interact with (e.g., be scattered by) crystal structure in the workpiece 102, and a detector 104 is positioned to collect diffracted X-ray beams. As a result of destructive diffraction and constructive diffraction, some of the diffracted X-ray beams cancel each other out, while others of the diffracted X-ray beams are added to create a new beam with higher amplitude. In this way, a diffraction pattern as a fingerprint of the crystal structure in the workpiece 102 can be sensed by the detector 104, and the crystal structure can be analyzed by extracting information from the diffraction pattern. According to some embodiments, the X-ray beams may pass through a set of incident optics 106 before striking the workpiece 102. Similarly, in some embodiments, the diffracted X-ray beams may pass through receiving optics 108 before being collected by the detector 104.
The workpiece 102 may be an intermediate structure during manufacturing of a semiconductor package, and a thermal conductive layer is exposed at an X-ray receiving surface of the workpiece 102. Rather than being a bulk structure, the thermal conductive layer is in a form of thin film. As an example (but not limited to), a thickness of the thermal conductive layer may be no greater than 500 nm. In order to target such shallow region of the workpiece 102, penetration depth by which the X-ray beams enter the workpiece 102 should be limited. Therefore, an incident beam angle ω defined between paths of the incident X-ray beams and the X-ray receiving surface of the workpiece 102 is fixed at a very small angle, such as 0.1° to 5°. By keeping a low incident beam angle ω, the XRD is also referred to as grazing angle XRD (GIXRD).
While the X-ray emission module 100 may be fixed, the detector 104 may move along a circle around the workpiece 102, and position of the detector 104 may be represented by an angle 2θ, which is defined between the incident X-ray beams and the diffracted X-ray beams. Also, the detector 104 records counts of X-ray at each value of the angle 2θ. In this way, a plot of X-ray intensity on Y-axis versus the angle 2θ on X-axis can be obtained. This plot, which may also be referred to as a diffractogram, reveals information about the crystal structure of the thermal conductive layer exposed at the X-ray receiving surface of the workpiece 102. As will be described in greater details, thermal properties of the thermal conductive layer can be predicted by extracting and analyzing the crystal information from the diffractogram.
The thermal conductive layer in the workpiece 102 is formed of a thermal conductive material. As compared to other candidates, certain ceramic based thermal conductive materials can provide great thermal conductivity, while offering dielectric isolation to prevent electrical short. AlN is one of these ceramic based thermal conductive materials that further shows great compatibility to back-end-of-line (BEOL) process, and the thermal characterization method will be described according to an example of which the target thermal conductive layer is formed of AlN.
Among other crystalline planes, (002) and (103) crystalline planes exist in AlN crystal structure (i.e., Wurtzite structure). In single crystalline AlN, the (002) plane is normal to direction, and the GIXRD cannot detect the (002) plane. That is, a peak corresponding to the (002) plane would be absent in the diffractogram. However, crystal grains in polycrystalline AlN may have a preferential orientation, and may be particularly longer in one direction. This results that a normal direction of the (002) plane in polycrystalline AlN may be tilted from the [001] direction, and a peak corresponding to the (002) plane may be shown at the angle 2θ of about 66° in the diffractogram. On the other hand, a normal direction of the (103) crystalline plane is tilted from the [001] direction under various conditions. A peak corresponding to the (103) crystalline plane may be shown at the angle 2θ of about 36° in the diffractogram. Hereinafter, the peak corresponding to the (002) crystalline plane will be simply referred to as a (002) peak, and the peak corresponding to the (103) crystalline plane will be simply referred to as a (103) peak.
Based on results of a series of preliminary experiments, anisotropy of thermal conductivity of the AlN thermal conductive layer is found to be highly dependent on an intensity ratio of the (103) peak with respect to the (002) peak. The anisotropy of thermal conductivity is defined as a ratio of cross-plane (vertical) thermal conductivity over in-plane (lateral) thermal conductivity, and is simply referred to as thermal anisotropy hereinafter. When the intensity ratio of the (103) peak with respect to the (002) peak is high, the (002) peak is rather weak, and the normal direction of the (002) crystalline plane is less tilted from the [001] direction. It indicates that crystal grains in the AlN thermal conductive layer are more aligned with the [001] direction, which may result in more efficient thermal conduction along a vertical direction, and therefore the AlN thermal conductive layer is resulted with a higher thermal anisotropy. On the other hand, when the intensity ratio of the (103) peak with respect to the (002) peak is low, the (002) peak is rather strong, and the normal direction of the (002) crystalline plane is more tilted from the direction. In this case, the crystal grains in the AlN thermal conductive layer are less aligned with the [001] direction, which may result in less efficient thermal conduction along the vertical direction, and therefore the AlN thermal conductive layer is resulted with a lower thermal anisotropy.
Instead of being roughly estimated, the thermal anisotropy of the AlN thermal conductive layer can be accurately predicted. Specifically, according to the results of the preliminary experiments, the cross-plane thermal conductivity of the AlN thermal conductive layer is in positive correlation with the intensity ratio of the (103) peak with respect to the (002) peak when this intensity ratio is greater than 1, while being a function of grain size of the AlN crystal grains along the (002) crystalline plane when this intensity ratio is less than 1. In addition, the thermal anisotropy of the AlN thermal conductive layer varies according to the cross-plane thermal conductivity of the AlN thermal conductive layer. That is, the cross-plane thermal conductivity of the AlN thermal conductive layer can be found as the intensity ratio of the (103) peak with respect to the (002) peak (and the grain size of the AlN crystal grains along the (002) crystalline plane) is/are determined, and then the thermal anisotropy of the AlN thermal conductive layer can be accurately predicted by the determined cross-plane thermal conductivity.
During the preliminary experiments, multiple AlN thin film samples are prepared. The samples respectively include an AlN layer similar to that in the workpiece 102 (which may be an intermediate structure of a semiconductor package, as described above), but the samples may not have an integrated circuit lying below each AlN layer, and are formed by using different process parameters so as to be different from one another in terms of alignment of crystal grains. By applying the described GIXRD characterization to these samples, a diffractogram for each sample can be obtained. Accordingly, the intensity ratio of the (103) peak with respect to the (002) peak for each sample can be extracted from the corresponding diffractogram. In addition, for each sample, grain size of the AlN crystal grains along the (002) crystalline plane can be calculated by using full width at maximum height (FWMH) of the (002) peak. Further, cross-plane thermal conductivity for each sample can be directly measured. As an example (but not limited to), an electrothermal method may be used for measuring the cross-plane thermal conductivity for each sample. By having the extracted crystal information and the measured cross-plane thermal conductivity, correlations in between can be found. To be more specific, for those samples with the intensity ratio (the ratio of the intensity of the (103) peak with respect to the intensity of the (002) peak) greater than 1, the described positive correlation between the intensity ratio and the cross-plane thermal conductivity can be determined. Also, for those samples with the intensity ratio (of the (103) peak with respect to the (002) peak) less than 1, the function describing how the grain size relates to the cross-plane thermal conductivity can be obtained.
Moreover, the preliminary experiments further include directly measuring in-plane thermal conductivity for each sample. Therefore, how the thermal anisotropy (which is defined as a ratio of cross-plane thermal conductivity over in-plane thermal conductivity) varies according to the cross-plane thermal conductivity can be identified. As an example (but not limited to), an electrothermal method or a thermal reflectance method may be used for measuring the in-plane thermal conductivity for each sample.
As shown in
These results of the preliminary experiments can be used for accurately predicting the cross-plane thermal conductivity and the thermal anisotropy of the AlN thermal conductive layer in the workpiece 102 (which may be an intermediate structure of a semiconductor package, as described), without directly measuring them. Specifically, based on the results shown in
According to some embodiments, the preliminary experiments further include measuring a thermal boundary conductance across an interface between the AlN layer and an underlying substrate for each sample. Result shows that the thermal boundary conductance is also related to the intensity ratio of the (103) peak with respect to the (002) peak for all of the samples.
As shown in
Accordingly, a thermal boundary conductance across an interface between the AlN thermal conductive layer and an underlying layer in the workpiece 102 can be predicted by using the intensity ratio (of the (103) peak with respect to the (002) peak) extracted from the diffractogram of the workpiece 102 and the preliminary work shown in
As described, according to the preliminary experiments, correlations between crystal information and thermal properties of the AlN samples can be determined. Once having these correlations, thermal properties of an AlN thermal conductive layer in an under-manufacturing workpiece can be predicted by performing the GIXRD characterization on the workpiece and extracting crystal information from the resulted diffractogram, without directly measuring the thermal properties of the workpiece. Since the GIXRD characterization is non-destructive, the workpiece can still be subjected to rest process steps after completing the GIXRD characterization. Therefore, this provides an in-line solution for monitoring thermal properties of a thermal conductive layer during manufacturing.
According to other embodiments, the preliminary experiments further include performing compositional analysis on the AlN samples. In this way, not only the described correlations between the thermal properties and the crystal information of AlN layers can be established, but also correlation between the thermal properties and elemental compositions of the AlN layers. For instance (but not limited to), according to results of the preliminary experiments, AlN layers with oxygen content below 2% and/or aluminum to nitrogen ratio greater than 0.9 and less than 1 may have thermal anisotropy greater than 4 and thermal boundary conductance over 200 MW/m2/K. In these embodiments, thermal properties of an AlN thermal conductive layer in an under-manufacturing workpiece can be further confirmed by performing a compositional analysis.
It should be appreciated that the disclosed thermal characterization method should not be limited to only targeting the AlN thermal conductive layer. Similar correlations between thermal properties and crystal information (and correlations between the thermal properties and elemental compositions) can be found for other crystalline thermal conductive materials by the same approach, thus thermal properties of a thermal conductive layer formed by any of these crystalline thermal conductive materials can be predicted by applying the GIXRD characterization and extracting the crystal information from the resulted diffractogram, and may be confirmed by further performing a compositional analysis. As examples (but not limited to), the available crystalline thermal conductive materials may include hexagonal boron nitride (h-BN), cubic boron nitride (c-BN), diamond, boron phosphide (BP), beryllium oxide (BeO) and aluminum oxide (Al2O3).
The manufacturing process includes an in-line thermal characterization targeting a thermal conductive layer in the semiconductor package. As shown in
Referring to
Referring to
Referring to
As shown in
Subsequently, at a step S308, the crystal information obtained in the step S306 can be used for predicting thermal properties of the thermal conductive layer 412, based on the results concluded from the preliminary experiments. That is, the thermal properties of the thermal conductive layer 412 are not directly measured, but are predicted by translating the result of the GIXRD characterization according to the correlations found in the preliminary experiments.
If the predicted thermal properties of the thermal conductive layer 412 fail to meet predetermined specification, then a cycle of the steps S302, S304, S306, S308 may be repeated again, and process parameters for forming the thermal conductive layer 412 at the step S304 are adjusted. The repetition may continue, until the thermal conductive layer 412 have expected thermal properties.
When the predicted thermal properties of the thermal conductive layer 412 meet the predetermined specification, then the manufacturing process may proceed to a step S310, at which the current wafer structure is further processed. As an example shown in
In certain embodiments, fabrication of a device wafer has been completed at the current stage. Subsequently, a singulation process may be performed, and a resulted semiconductor die 420 may be resulted. Although not shown, the semiconductor die 420 may be further packaged for forming a semiconductor package.
Formation of the die stacking structure 500a may involve bonding two of the semiconductor dies 420 shown in
As similar to the thermal conductive layer 412 at a front side of the first semiconductor die 420-1, the additional thermal conductive layer 502 at a back side of the second semiconductor die 420-2 may be formed of a ceramic based thermal conductive material, such as AlN, hexagonal boron nitride (h-BN), cubic boron nitride (c-BN), diamond, boron phosphide (BP), beryllium oxide (BeO) or aluminum oxide (Al2O3). Optionally, thermal properties of the additional thermal conductive layer 502 in the second semiconductor die 420-2 can be monitored by the described thermal characterization method right after deposition.
Moreover, the die stacking structure 500a may be further processed to form a semiconductor package. At least, the die stacking structure 500a may be encapsulated by a molding compound, and integrated circuits in the semiconductor dies 420 may be routed to package inputs/outputs (I/Os) exposed at a side (e.g., a bottom side) of the semiconductor package.
The die stacking structure 500b is similar to the die stacking structure 500a described with reference to
According to some embodiments, the bonding layers 508 are formed of a dielectric material different from the thermal conductive material for forming the thermal conductive layers 412. For instance, while the thermal conductive layers 412 are formed of AlN, the bonding layers 508 may be formed of silicon oxide.
As similar to the die stacking structure 500a described with reference to
In further embodiments, more wafer processing steps may be performed on the wafer structure shown in
The semiconductor die 600 may be obtained by further processing the wafer structure shown in
Referring to
The die stacking structure 700b is similar to the die stacking structure 700a described with reference to
According to some embodiments, the bonding layers 708 are formed of a dielectric material different from the thermal conductive material for forming the thermal conductive layers 412, 602. For instance, while the thermal conductive layers 412, 602 are formed of AlN, the bonding layers 708 may be formed of silicon oxide.
Although not explicitly illustrated, either the die stacking structure 700a shown in
In the afore-described embodiments, the thermal conductive layer 412 is functioned as a metallization layer and/or a bonding layer for engaging with another semiconductor die. In other embodiments, the thermal conductive layer 412 is attached to a heat dissipation module.
The semiconductor package 800 may include a semiconductor die 802 singulated from the wafer structure shown in
As above, an in-line thermal characterization method is provided. The thermal characterization method is used for instantly monitoring thermal properties of a thermal conductive layer in a semiconductor structure during manufacturing of the semiconductor structure, and for tuning process parameters for forming the thermal conductive layer. Specifically, preliminary experiments are performed before manufacturing of the semiconductor structure, to establish correlations between crystal information and the thermal properties of thin film samples made of the same material as the thermal conductive layer. During manufacturing of the semiconductor structure, a GIXRD analysis is performed on an intermediate structure with the thermal conductive layer exposed at its surface, for determining crystal information of the thermal conductive layer. Based on the correlations concluded from the preliminary experiments, the thermal properties of the as-deposited thermal conductive layer can be accurately predicted by using the determined crystal information. Since the GIXRD analysis is non-destructive and a destructive sampling is not required for the GIXRD analysis, the tested intermediate structure can be further processed for completing fabrication of the semiconductor structure, if the thermal conductive layer therein is ensured to have expected thermal properties. As the semiconductor structure has the qualified thermal conductive layer, efficient heat dissipation of the semiconductor structure may be achieved. On the other hand, if the identified thermal properties of the thermal conductive layer fall below expectation, process parameters for forming the thermal conductive layer may be adjusted.
Even though the semiconductor structure is illustrated as a wafer structure and the thermal conductive layer is described as being formed along a front side or a back side of the wafer structure, it can be appreciated the semiconductor structure may otherwise be a package structure with a semiconductor die and a surrounding encapsulant, and the thermal conductive layer may cover both of the encapsulant and the semiconductor die. The disclosed thermal characterization can be applied for various semiconductor structures each with one or more crystalline thermal conductive layers formed therein.
In an aspect of the present disclosure, a thermal characterization method is provided. The thermal characterization method is used for investigating a thermal conductive layer in a semiconductor structure, and comprises: conducting preliminary experiments before manufacturing of the semiconductor structure, to establish correlations between crystal properties and thermal properties of thin film samples formed of a crystalline material identical with a crystalline material for forming the thermal conductive layer; performing a grazing angle X-ray diffraction (GIXRD) characterization on the thermal conductive layer during manufacturing of the semiconductor structure, and extracting crystal properties of the thermal conducive layer from a resulted diffractogram; and using the correlations established by the preliminary experiments to find thermal properties corresponding to the extracted crystal properties of the thermal conductive layer.
In another aspect of the present disclosure, a method for manufacturing a semiconductor package with a thermal conductive layer is provided. The method comprises: conducting preliminary experiments, to establish correlations between crystal properties and thermal properties of thin film samples formed of a crystalline material identical with a crystalline material for forming the thermal conductive layer; depositing the thermal conductive layer over a substrate; subjecting an intermediate structure with the thermal conductive layer exposed at its surface to a grazing angle X-ray diffraction (GIXRD) characterization, and extracting crystal properties of the thermal conducive layer from a resulted diffractogram; using the correlations established by the preliminary experiments to find thermal properties corresponding to the extracted crystal properties of the thermal conductive layer; and further processing the intermediate structure to complete formation of the semiconductor package.
In yet another aspect of the present disclosure, a method for manufacturing a semiconductor package with a thermal conductive layer is provided. The method comprises: conducting preliminary experiments, to establish correlations between crystal properties and thermal properties of thin film samples formed of a crystalline material identical with a crystalline material for forming the thermal conductive layer; depositing the thermal conductive layer over a substrate; subjecting an intermediate structure with the thermal conductive layer exposed at its surface to a grazing angle X-ray diffraction (GIXRD) characterization, and extracting crystal properties of the thermal conducive layer from a resulted diffractogram; using the correlations established by the preliminary experiments to find thermal properties corresponding to the extracted crystal properties of the thermal conductive layer; determining if at least one of the thermal properties of the thermal conductive layer is greater than a predetermined expectation level; when the at least one of the thermal properties of the thermal conductive layer is greater than the predetermined expectation level, then further processing the intermediate structure to complete formation of the semiconductor package; and if the at least one of the thermal properties of the thermal conductive layer is lower than the predetermined expectation level, then performing a next cycle from depositing another thermal conductive layer on another substrate by using adjusted deposition parameters.
The foregoing outlines features of several embodiments so that those skilled in the art may better understand the aspects of the present disclosure. Those skilled in the art should appreciate that they may readily use the present disclosure as a basis for designing or modifying other processes and structures for carrying out the same purposes and/or achieving the same advantages of the embodiments introduced herein. Those skilled in the art should also realize that such equivalent constructions do not depart from the spirit and scope of the present disclosure, and that they may make various changes, substitutions, and alterations herein without departing from the spirit and scope of the present disclosure.
This application claims the priority benefit of U.S. provisional application Ser. No. 63/609,353, filed on Dec. 13, 2023. The entirety of the above-mentioned patent application is hereby incorporated by reference herein and made a part of this specification.
| Number | Date | Country | |
|---|---|---|---|
| 63609353 | Dec 2023 | US |
