US20060266652A1

US20060266652A1 - Method and system for on-line controlling of solder bump deposition

Info

Publication number: US20060266652A1
Application number: US11/347,454
Authority: US
Inventors: Andreas Netz; Niels Rackwitz; Joern Schnapke; Frank Kuechenmeister; Norbert Schroeder
Original assignee: Individual
Current assignee: Advanced Micro Devices Inc
Priority date: 2005-05-31
Filing date: 2006-02-03
Publication date: 2006-11-30
Also published as: DE102005024910A1

Abstract

By evaluating a dynamic failure signal, such as a voltage signal and/or a current signal, obtained during an electroplating operation for forming solder bumps, an inline control system with a responsiveness on a substrate basis may be established. Thus, the electroplating tool may be controlled on a single wafer basis to improve process uniformity and also significantly reduce yield loss.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention
The present invention relates to a process flow for forming a contact layer including bumps of a contact material, such as solder, which is used to provide contact areas for directly attaching an appropriately formed package or carrier substrate to a die carrying an integrated circuit.
2. Description of the Related Art
In manufacturing integrated circuits, it is usually necessary to package a chip and provide leads and terminals for connecting the chip circuitry with the periphery. In some packaging techniques, chips, chip packages or other appropriate units may be connected by means of balls of solder or any other conductive material, formed from so-called solder bumps or bumps that are formed on a corresponding layer, which will be referred to herein as a contact layer, of at least one of the units, for instance on a dielectric passivation layer of the microelectronic chip. In order to connect the microelectronic chip with the corresponding carrier, the surfaces of the two respective units to be connected, i.e., a microelectronic chip comprising, for instance, a plurality of integrated circuits, and a corresponding package, have formed thereon adequate pad arrangements to electrically connect the two units after reflowing the bumps provided at least on one of the units, for instance on the microelectronic chip. In other techniques, bumps may have to be formed that are to be connected to corresponding wires, or the bumps may be brought into contact with corresponding pad areas of another substrate acting as a heat sink. Consequently, it may be necessary to form a large number of bumps that may be distributed over the entire chip area, thereby providing, for example, the I/O capability required for modern microelectronic chips that usually include complex circuitry, such as microprocessors, storage circuits and the like and/or include a plurality of integrated circuits forming a complete complex circuit system.
In order to provide hundreds or thousands of mechanically well-fastened bumps on corresponding pads, the attachment procedure of the bumps requires a careful design, since the entire device may be rendered useless upon failure of only one of the bumps. For this reason, one or more carefully chosen layers are generally placed between the bumps and the underlying substrate or wafer including the pad arrangement. In addition to the important role these interfacial layers, herein also referred to as underbump metallization layers, may play in endowing a sufficient mechanical adhesion of the bump to the underlying pad and the surrounding passivation material, the underbump metallization has to meet further requirements with respect to diffusion characteristics and current conductivity. Regarding the former issue, the underbump metallization layers have to provide an adequate diffusion barrier to prevent the solder material or bump material, frequently a mixture of lead (Pb) and tin (Sn), from attacking the chip's underlying metallization layers and thereby destroying or negatively affecting their functionality.
Moreover, migration of bump material, such as lead, to other sensitive device areas, for instance into the dielectric, where a radioactive decay of lead may also significantly affect the device performance, has to be effectively suppressed by the underbump metallization. Regarding current conductivity, the underbump metallization, which serves as an interconnect between the bump and the underlying metallization layer of the chip, has to exhibit a thickness and a specific resistance that does not inappropriately increase the overall resistance of the metallization pad/bump system.
In addition, the underbump metallization will serve as a current distribution layer during electroplating of the bump material. Electroplating is presently the preferred deposition technique for solder material, since physical vapor deposition of solder bump material, which is also used in the art, requires a complex mask technology in order to avoid any misalignments due to thermal expansion of the mask while it is contacted by the hot metal vapors. Moreover, it is extremely difficult to remove the metal mask after completion of the deposition process without damaging the solder pads, particularly when large wafers are processed or the pitch between adjacent solder pads decreases.
Although a mask is also used in the electroplating deposition method, this technique differs from the evaporation method in that the mask is created using photolithography to thereby avoid the above-identified problems caused by physical vapor deposition techniques. However, electroplating requires a continuous and highly uniform current distribution layer adhered to the substrate that is mainly insulative, except for the pads on which the bumps have to be formed. Thus, the underbump metallization also has to meet strictly set constraints with respect to a uniform current distribution, as any non-uniformities during the plating process may affect the final configuration of the bumps and, after reflowing the bumps, of the resulting solder balls in terms of, for instance, height non-uniformities, which may in turn translate into fluctuations of the finally obtained electric connections and the mechanical integrity thereof. Since the height of the bumps is determined by the local deposition rate during the electroplating process, which is per se a highly complex process, any process non-uniformities resulting from irregularities of the plating tool or any components thereof may also directly cause corresponding non-uniformities during the final assembly process. Moreover, since the formation of the bumps is one of the final steps that is performed on a substrate basis, any variations of the plating process or even loss of substrates due to tool failures immensely contributes to increased production costs and reduced yield.
Consequently, the metal deposition based on a patterned photoresist is a key process step with respect to reliability, yield and production cost, wherein a plurality of process-specific issues, such as the handling of multiple materials exposed on the substrate surface, the influence of pattern density at the substrate, die and feature scale, have to be taken into consideration to obtain a highly uniform metal deposition. Particularly, the factors, such as thickness uniformity, deposition rate and, if an alloy is to be used as the solder or bump material, the control of the alloy composition, is also an important criterion, as both the deposition rate and the alloy composition may strongly be affected by the mass transfer in the electroplating tool.
In view of the above-described situation, a need exists for an enhanced technique that may avoid or at least reduce the effects of one or more of the problems identified above.

SUMMARY OF THE INVENTION

The following presents a simplified summary of the invention in order to provide a basic understanding of some aspects of the invention. This summary is not an exhaustive overview of the invention. It is not intended to identify key or critical elements of the invention or to delineate the scope of the invention. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is discussed later.
Generally, the present invention is directed to a technique that enables the detection of a failure status of an electroplating tool on the basis of a dynamic status signal, thereby providing the potential for rapidly detecting and thus reacting on deviations from standard situations. For this purpose, a failure data collection (FDC) technique is provided that may operate on a single substrate basis, thereby significantly enhancing process control and production yield.
According to one illustrative embodiment of the present invention, a method comprises forming a plurality of bumps on a first substrate by an electrochemical deposition process in an electroplating tool. A dynamic status signal is obtained during the processing of the first substrate, wherein the dynamic status signal represents a dynamic behavior of at least one tool parameter during the electrochemical deposition process. Furthermore, a current tool status is estimated on the basis of the dynamic status signal and the electroplating tool is released for forming a plurality of bumps on at least one subsequent substrate on the basis of the estimated tool status.
In accordance with another illustrative embodiment of the present invention, a method comprises processing a first substrate in an electroplating tool that operates on a single substrate basis. Furthermore, a failure status of the electroplating tool is monitored on the basis of a dynamic status signal that is obtained during the processing of the first substrate. Finally, the failure status is compared with at least one reference status prior to the processing of an additional substrate.
According to yet another illustrative embodiment of the present invention, a system comprises an electroplating tool having an anode assembly and a failure detection unit that is connected to the electroplating tool. The failure detection unit is configured to indicate a failure status of the electroplating tool for each process run on the basis of a dynamic status signal.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention may be understood by reference to the following description taken in conjunction with the accompanying drawings, in which like reference numerals identify like elements, and in which:
FIG. 1 a schematically shows a cross-sectional view of a substrate that receives a plurality of solder bumps during a specific manufacturing stage;
FIG. 1 b schematically shows a system for forming bumps on a substrate, such as the substrate of FIG. 1 a, by electroplating, wherein the system comprises a failure detection unit according to illustrative embodiments of the present invention;
FIG. 1 c schematically represents a graph depicting a voltage signal as an example for a dynamic signal according to an illustrative embodiment; and
FIG. 1 d schematically illustrates the failure detection unit in more detail in accordance with still further illustrative embodiments of the present invention.
While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and are herein described in detail. It should be understood, however, that the description herein of specific embodiments is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.

DETAILED DESCRIPTION OF THE INVENTION

Illustrative embodiments of the invention are described below. In the interest of clarity, not all features of an actual implementation are described in this specification. It will of course be appreciated that in the development of any such actual embodiment, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which will vary from one implementation to another. Moreover, it will be appreciated that such a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure.
The present invention will now be described with reference to the attached figures. Various structures, systems and devices are schematically depicted in the drawings for purposes of explanation only and so as to not obscure the present invention with details that are well known to those skilled in the art. Nevertheless, the attached drawings are included to describe and explain illustrative examples of the present invention. The words and phrases used herein should be understood and interpreted to have a meaning consistent with the understanding of those words and phrases by those skilled in the relevant art. No special definition of a term or phrase, i.e., a definition that is different from the ordinary and customary meaning as understood by those skilled in the art, is intended to be implied by consistent usage of the term or phrase herein. To the extent that a term or phrase is intended to have a special meaning, i.e., a meaning other than that understood by skilled artisans, such a special definition will be expressly set forth in the specification in a definitional manner that directly and unequivocally provides the special definition for the term or phrase.
The present invention is generally based on the concept that a failure data collection approach enables the implementation of a control strategy for electroplating material, and, in particular embodiments, for electroplating bumps such as solder bumps, wherein any misprocessing of substrates is significantly reduced. Moreover, by using an appropriate failure data collection approach, an in-line control and associated therewith an automatic failure detection in the plating process may be accomplished. For this purpose, an appropriate dynamic failure signal, which is indicative of the currently prevailing tool and process status, may be monitored and may be processed to establish the current failure status of the tool, wherein the further operating mode and/or the release of the electroplating tool for the next substrate may be based on the estimated failure state. With reference to the accompanying drawings, further illustrative embodiments of the present invention will now be described in more detail.
FIG. 1 a schematically shows a substrate 101 in cross-sectional view, on which a plurality of semiconductor devices may be formed which are to receive a contact layer 102 for providing electrical, thermal and mechanical connection to a carrier substrate (not shown). The substrate 101 may comprise a layer 103 having formed therein a plurality of microstructural features, such as circuit elements of integrated circuits, which, for convenience, are not shown. Moreover, the layer 103 may comprise a plurality of contact pads 104, at least some of which may be in electric contact to any lower-lying circuit elements. Above the layer 103 there is formed a dielectric layer 105 made of any appropriate material, wherein respective openings are formed in the dielectric layer 105 to allow electric contact to at least a portion of the contact pads 104. Formed on the dielectric layer 105 and the contact pads 104 is an underbump metallization layer 106, which may typically be comprised of a plurality of specific layers in order to provide the required functionality, such as diffusion blocking, adhesion to the contact pads 104, thermal mechanical characteristics, such as thermal expansion, the current distribution during a subsequent electroplating process, appropriately seeding and initializing the electroplating process, and the like. The substrate 101 further comprises a mask 107, such as a resist mask, which is patterned in accordance with design requirements for forming bumps 108, such as solder bumps with a specified size and especially with a well-defined height, as the height of the bumps 108 may significantly determine the reliability and characteristics of contacts to be formed with corresponding contact areas of the carrier substrate. In some embodiments, the bumps 108 may be formed by a composition of two or more different metals. In this case, the composition of the bumps 108 may also represent an important device feature, since the bumps 108 may have to consistently melt at a specified temperature during the reflow process for forming solder balls or for directly contacting respective contact areas of a carrier substrate. Thus, as previously pointed out, a precise process control for forming the bumps 108 is required, in particular as in sophisticated applications, the pitch between neighboring bumps 108 may continuously be reduced, while the overall number of bumps 108 per substrate 101 may steadily increase due to the increasing requirements with respect to I/O capabilities of sophisticated semiconductor devices.
FIG. 1 b schematically illustrates a system 100 for processing substrates which are to receive metal bumps, such as the bumps 108 of the substrate 101, by an electroplating process. The system 100 comprises an electroplating tool 110 having a reactor bowl 111, which is configured to receive and hold an appropriate electrolyte 112. The reactor bowl 111 further comprises an electrode assembly 113, which may also be referred to as an anode assembly, since during an actual deposition process the electrode assembly 113 provides an averaged positive electric field, irrespective of whether temporarily a negative voltage is applied to the electrode assembly 113 during any intermediate time periods. Thus, it should be understood that the notion “anode assembly” is to be considered in this general sense. The reactor bowl 111 may further be configured to receive and hold in place a substrate, such as the substrate 101, wherein any appropriate means for electrically contacting the substrate 101 to thereby make the substrate 101 a counter electrode, are not shown. The system 100 may further comprise a control unit 120, which is configured to control the operation of the electroplating tool 110, by for instance coordinating any loading or unloading activity for conveying the substrate 101 into the reactor bowl 111 after a previously processed substrate has been removed from the bowl 111. Moreover, the control unit 120 may be configured to control the composition and the flow of the electrolyte 112 within the reactor bowl 111 by controlling appropriate supply tanks and supply lines (not shown) for replenishing any bath components of the electrolyte 112. Moreover, the relative position of the substrate 101 with respect to the electrode assembly 113 for establishing a required distance between the electrode assembly 113 and the counter electrode, i.e., the surface of substrate 101, and the movement of the substrate 101 during the actual deposition process may also be controlled by the control unit 120 by means of corresponding drive assemblies (not shown).
Furthermore, the control unit 120 may be configured to initiate a current flow from the electrode assembly 113 through the electrolyte 112 and to the substrate 101, thereby depositing metal on surface portions exposed by the resist mask 107. For this purpose, the system 100 may comprise or may be connected to a controllable power supply 121, which is configured to supply power to the electrode assembly 113 in accordance with specified process requirements. For example, the controllable power supply 121 may comprise a controllable current source, which is designed such that a controllable amount of current may be supplied to the electrode assembly 113. Based on the adjusted amount of current and the effective deposition time, the amount of metal deposited on the substrate 101 may be efficiently controlled for a given bath composition of the electrolyte 112. The controllable power supply 121 may be configured to be operated in a plurality of operating modes, such as a pulsed operating mode in which a sequence of current pulses may be applied, wherein, in intermediate periods, substantially no current or only a small current, or even an inverse current, may be generated. It should be noted that in case of an operating mode with reverse current pulses, the electrode assembly 113 temporarily acts as a cathode. The controllable power supply 121 may also be operable in a substantially continuous constant current mode wherein, however, the magnitude of this “constant” current may be varied over time. In other embodiments, the controllable power supply 121 may additionally or alternatively have implemented therein a constant voltage operating mode, in which a constant voltage is supplied to the electrode assembly 113, wherein, similarly as in the constant current operating mode, the constant voltage may be supplied in pulses or in a substantially continuous fashion, wherein the plurality of the constant voltage pulses may even be reversed during certain deposition phases. In illustrative embodiments, the controllable power supply 121 is configured to individually operate each of a plurality of anode segments 113 a, 113 b, thereby providing the potential for controlling the deposition profile on the substrate 101, since the deposition rate may locally be varied by supplying an appropriately controlled amount of current to each of the anode segments 113 a, 113 b.
The system 100 further comprises a failure detection unit 130 which is at least operatively coupled to the electroplating tool 110 for receiving a dynamic status signal 131 a, 131 b from the electroplating tool 110. The dynamic status signal 131 a, 131 b may represent a signal that is sensitive to a change of at least one tool parameter to allow an estimation of the at least one tool parameter during the processing of a single substrate, such as the substrate 101, and therefore the signal 131 a, 131 b may be considered as a dynamic signal.
As is previously explained, the electrochemical deposition process in the tool 110 is a highly complex process, wherein typically subtle changes of the process output, such as uniformity of the bumps 108, the material composition thereof, thickness variation from substrate to substrate, and the like, may be detected upon processing of a plurality of substrates and corresponding measurement results may be used for process control with a significant delay. Consequently, in some embodiments, the control unit may have implemented therein sophisticated APC (advanced process control) strategies to provide a certain predictability in establishing appropriate manipulated variables, such as values for the current supplied to the assembly 113, the replenishing of bath components, and the like, on a run-to-run basis, even for a significant delay of the measurement results of previously processed substrates. Contrary to the dynamic signal 131 a, 131 b, such measurement results will be referred to as post-process data, which are considered as being non-dynamic in the sense that this data may not be used for qualifying a single process run of a substrate that is currently being processed. Similarly, measurement results obtained from one or more of the substrates 101 prior to being processed in the electroplating tool 110 may also be supplied to the control unit 120, indicated as pre-process data, so as to determine appropriate values of the manipulated variables of the process recipe under consideration. For instance, measurement results relating to the underbump metallization layer 106 may be used to determine appropriate values for the manipulated variables. For example, measurement data may indicate a reduced layer thickness of the underbump metallization layer 106 so that a reduced deposition rate may be expected due to the increased resistance of the layer 106, which may require a higher current and/or an increased deposition time. Also, in this case, the pre-process data may not be considered as dynamic process information, since this data may not per se allow extraction of any information about the actual deposition process for the substrate under consideration.
In one illustrative embodiment, the dynamic signal 131 a, 131 b may represent a voltage determined between the electrode assembly 113 and an appropriate second point in the electric circuit formed by the controllable power supply 121, the electrode assembly 113, the electrolyte 112 and the substrate 101. In one embodiment, the voltage between the electrode assembly 113 and the substrate 101, which acts as a counter electrode, may represent the signal 131 a, 131 b, which is determined by an appropriate voltage detector 132. It should be appreciated, however, that any other “measurement point” within the electric circuit may be selected, such as an auxiliary electrode (not shown), which may be provided within the reactor bowl 111. In some illustrative embodiments, when the electrode assembly 113 comprises the plurality of anode segments 113 a, 113 b, the detector 132 may comprise corresponding detector segments 132 a, 132 b, to individually establish the respective dynamic signals 131 a, 131 b, thereby increasing the amount of information that may be extracted with respect to the dynamic behavior within the reactor bowl 111. In this respect, it is to be noted that the provision of the two anode segments 131 a, 131 b is of illustrative nature only and in other embodiments the electrode assembly 113 may comprise any appropriate number of anode segments, which may be provided in the form of concentric ring electrodes, in the form of interleaved anode segments, or any other appropriate arrangement. Moreover, it should be appreciated that in other embodiments the voltage detector 132 may not necessarily be directly connected to the electrode assembly 113, although this arrangement is advantageous, since the dynamic signals 131 a, 131 b provided by the detector 132, as shown in FIG. 1 b, represent the voltage drop from the electrode assembly 113 across the electrolyte 112 to the substrate surface of the substrate 101. Consequently, these voltage signals are based on the “dynamic” behavior of the reactor bowl 111 during deposition while substantially rejecting other influences, such as any voltage drops of external components, such as external connectors, the controllable power supply 121, and the like.
In other embodiments it may be considered advantageous to provide, in addition or alternatively to the electrode assembly 113 as one measurement node, different measurement nodes, such as a specifically positioned auxiliary electrode, when a certain location within the reactor bowl 111 has been identified as being highly sensitive to any changes of a specified tool parameter. For example, one or more auxiliary electrodes, which may be positioned between the electrode assembly 113 and the substrate 101, may be operated continuously or intermittently with an appropriate current to “probe” the interior of the reactor bowl 111 with a higher spatial “resolution,” at least in the vertical direction, compared to the arrangement in which the electrode assembly 113 and the substrate 101 represent the measurement nodes. In such an arrangement, the auxiliary electrodes may be operated with an extremely low current so as to not significantly influence the overall deposition behavior or, in other cases, may be used so as to provide, in combination with the electrode assembly 113, a desired deposition profile.
In still other embodiments, the dynamic signal 131 a, 131 b may be supplied by a current measurement detector, which may be implemented in the controllable power supply 121 or in any other external device (not shown), wherein optionally the voltage detector 132 may provide corresponding voltage signals so as to provide measurement readings of the actual voltage at the electrode assembly 113, when the controllable power supply 121 is operated in a constant voltage mode. Similarly, the failure detection unit 130 may be configured to receive corresponding current signals from the controllable power supply 121, even when it is operated in the constant current mode so that the unit 130 may have the currently valid current values, which may be varied by the control unit 120 on the basis of the post-process data and the pre-process data, as is previously discussed.
In other embodiments, however, the control unit 120 may be configured to operate the electroplating tool 110 on the basis of a predetermined process recipe without any model predictive control strategy so that constant current values are used for each run. In this case, the failure detection unit 130 may not receive any current signals from the controllable power supply 121 and a change in the dynamic signal 131 a, 131 b may directly indicate a change in one or more tool parameters of the tool 110. The failure detection unit 130 is further configured to estimate the status of the electroplating tool 110, at least with respect to a failure status, on the basis of the dynamic signal 131 a, 131 b. In estimating at least the failure status of the tool 110, it is to be understood that the failure detection unit 130 is adapted to recognize at least an invalid tool status of the tool 110 on the basis of the signal 131 a, 131 b, by comparing the signal 131 a, 131 b with appropriately defined reference data, wherein the comparison is performed such that at least the estimation of the tool status is completed prior to the processing of a subsequent substrate in the tool 110. In other illustrative embodiments, the failure detection unit 130 may be configured to perform a more detailed status analysis on the basis of the signal 131 a, 131 b, as will be described in more detail with reference to FIG. 1 c.
During operation of the system 100, a plurality of substrates 101 are sequentially to be processed in the electroplating tool 110 in accordance with a specified process recipe to form the bumps 108 (FIG. 1 a) with a desired material composition and predefined co-planarity. One of the plurality of substrates 101 is loaded into the reactor bowl 111, wherein an appropriate substrate holder (not shown) receives the substrate 101 and positions it at a specific operating position, which may have been defined in a previous calibration procedure or which may be determined on the basis of manual, semi-automatic and automatic initialization procedures. It should be appreciated that the present invention is not restricted to any type of electroplating reactor and the operating position of the substrate 101 as shown in FIG. 1 b is of illustrative nature only. Thus, any other type of electroplating reactor, including reactors with a vertically arranged electrode assembly and substrate, may be used in combination with the present invention.
Thereafter, the control unit 120 may establish a current flow between the electrode assembly 113 and the surface of the substrate 101 in accordance with the specified process recipe wherein, as previously explained, a variety of process recipes such as constant current mode, constant voltage mode, pulsed operation, continuous operation and any combinations of these different modes, may have been established so as to obtain, for a valid tool status of the electroplating tool 110, a highly uniform formation of the bumps 108. Prior, during or after the creation of a current flow through the electrolyte 112, the dynamic signal 131 a, 131 b may be obtained by the detector 132 and may be supplied to the failure detection unit 130. As previously explained, the dynamic signal 131 a, 131 b may be generated by any appropriate mechanism that allows extraction of information on the presently prevailing status of the tool 110. For example, if dedicated auxiliary electrodes are present within the reactor bowl 111, these electrodes may be operated to establish the signals 131 a, 131 b. In other embodiments, as is shown in FIG. 1 b, the dynamic signal 131 a, 131 b may be obtained upon establishing a current flow through the electrolyte 112, wherein the voltage drop across the reactor bowl 111 may provide information of the present tool status. In other embodiments, additionally or alternatively, the current through the electrode assembly 113, the electrolyte 112 and the substrate 101 may be sampled and may be analyzed by the failure detection unit 130 so as to at least detect whether or not an invalid tool status has occurred.
FIG. 1 c schematically shows an exemplary wave form of the signals 131 a, 131 b when representing a voltage drop measured by the detector 132, as shown in FIG. 1 b. Hereby, it is assumed that the control unit 120 operates the controllable power supply 121 in a pulsed constant current mode, wherein a group of current pulses of identical height and duration is established followed by an intermediate period with no current supplied. Thereafter, the group of current pulses may be repeated, followed by a further deposition-free period. This sequence may be continued until the current-time integral corresponds to the target value for the process recipe under consideration. In FIG. 1 c, the dashed line may represent the progression of the current over time supplied to the electrode assembly 113, while the solid lines may represent the corresponding voltage signals obtained by the voltage detector 132 as the dynamic signal 131 a, 131 b. For convenience, only a single voltage signal is shown in FIG. 1 c. It should further be appreciated that the duration of a single voltage or current pulse in FIG. 1 c may be on the order of magnitude of milliseconds, while the amount of current per current pulse may range to several tenths of ampere. Based on the signal 131 a, 131 b, i.e., in the example of FIG. 1 c, the voltage signal (solid line), the failure detection unit 130 may correspondingly operate on the data representing the signal 131 a, 131 b so as to enable a comparison with appropriately defined reference data. The reference data may specify a characteristic status of the tool 110, for instance a valid tool status, which may represent an operating mode of the tool 110 with defect-free hardware components, such as the electrode assembly 113, for a given process recipe. Appropriate reference data may be obtained by gathering one or more of the signals 131 a, 131 b for a well-defined tool status of the tool 110.
In one illustrative embodiment, reference data are gathered for at least two different tool statuses, which may thus allow a quantitative estimation of the tool status at least with respect to one or more specific tool parameters. For instance, the signals 131 a, 131 b may, for otherwise identical operating conditions, be obtained for an electrode assembly 113 in a valid status and in an invalid status, wherein, in illustrative embodiments, also a plurality of intermediate statuses may be investigated to obtain corresponding reference data. In other cases, various signals reflecting different tool status with respect to a different electrolyte composition may be gathered and may correspondingly be processed to obtain respective reference data. Similarly, reference data may be obtained for two or more different operating positions of the substrate 101 to thereby obtain reference data regarding any mispositioning of the substrate 101. The dynamic signals 131 a, 131 b, irrespective of whether reference data or actual process data are considered, may be processed in any appropriate manner to thereby enable a rapid and reliable comparison of the reference data with the signals 131 a, 131 b obtained during an actual deposition process. In illustrative embodiments, the raw data representing, for instance, the voltage signals of FIG. 1 c, may be processed to obtain one or more statistically significant values, which may be referred to as statistical key numbers, which represent a quantitative measure of at least one quality of the signals 131 a, 131 b. For instance, an integration over time of the raw data, possibly in combination with the current values corresponding thereto, may provide an “overview” of the global behavior of the electroplating tool 110. For example, the total energy supplied by the controllable power supply 121, which is actually introduced into the reactor bowl 111, may be calculated on the basis of the time integral of the voltage signal during the entire deposition process or during a specified part thereof, which may then be compared with corresponding reference values of the power for an invalid or a valid reference status to provide a first criterion for the failure status of the tool 110.
In addition, or alternatively, appropriate filtering and/or clipping techniques may be used to significantly reduce the amount of data or to increase the efficiency of any processes for extracting information from the signals 131 a, 131 b. Efficient filtering techniques, such as high pass filtering, low pass filtering and band pass filtering, and the like, may be very efficient in removing unwanted signal components, which may otherwise compromise or “obscure” the statistical relevance of extracted values or value ranges. For example, a low pass filter may remove high frequency components in the signal, thereby providing a smooth voltage signal so that even individual voltage pulses may be compared with corresponding reference pulses, by for instance calculating a corresponding mean value. In other embodiments, data reduction may be advantageous as, for instance, the analyzation of reference data may have revealed that only one or a low number of voltage pulses per each group may suffice to represent the time progression of the dynamic signal 131 a, 131 b. It should be appreciated that corresponding clipping and filtering criteria may be established on the basis of previously gathered data, in particular on the basis of previously gathered reference data for which well-established tool conditions are known.
Based on the signals 131 a, 131 b, the failure detection unit 130 may detect at least an invalid tool status by, for instance, comparing one or more statistical key numbers with corresponding reference values, and may indicate the corresponding failure state of the tool 110 prior to the processing of a next one of the plurality of substrates 101. For this purpose, in some illustrative embodiments, the failure detection unit 130 may operatively be coupled to the control unit 120, wherein the control unit 120 is configured to receive a corresponding failure status indication and to initiate a corresponding tool activity. For example, when an invalid tool status is detected by the unit 130, the control unit 120 may instruct the tool 110 to discontinue operation upon completion of the deposition process presently running in the tool 110. In other embodiments, the failure detection unit 130 may, as previously discussed, be configured to estimate the tool status in a more quantitative manner so that the corresponding status indication may enable an enhanced process control by the control unit 120. For example, the quantitative measure of the presently prevailing tool status determined by the failure detection unit 130 may be used as a control variable or as an offset or machine constant for a model predictive control strategy implemented in the control unit 120. For instance, based on post-process data and/or pre-process data, the control algorithm implemented in the control unit 120 may calculate appropriate manipulated variables for the plurality of substrates to be processed in the tool 110, wherein the status indication provided by the unit 130 may be used as offset values that may provide corrections for each single run, wherein, in particular, the release of the process tool 110 for the next substrate is determined by the status indication supplied by the unit 130.
In other embodiments, a plurality of predetermined machine activities may be associated with respective status indications established by the unit 130. That is, a plurality of status codes may have been established, for instance on the basis of reference data and a plurality of well-defined tool conditions, wherein each failure code is associated with a dedicated tool activity, wherein corresponding instruction tables and the like may be implemented in the control unit 120 so as to appropriately respond to the failure status detected by the unit 130. In some embodiments, the predefined machine activities may incorporate specific maintenance actions. In this case, a certain degree of self-diagnosis may be established in the system 100, thereby significantly enhancing tool utilization as the tool 110 may be brought back into production more rapidly compared to conventional electroplating systems. Moreover, in one illustrative embodiment, the detection unit 130, possibly in combination with the control unit 120, may provide a certain predictability of the tool behavior with respect to one or more tool parameters. For example, a certain criterion may be established to predict an estimated time period for the replacement or maintenance of one or more specified hardware components. For instance, the failure detection unit 130 may recognize a degradation of a specific component, such as the electrode assembly 113, which may be compensated for automatically by the operational mode, for instance a constant current mode, substantially without affecting the quality of the solder bumps 108. Nevertheless, by providing an estimated time for a replacement of the electrode assembly 113, the availability of the tool 110 may be estimated more reliably and therefore process flow management in a semiconductor facility may significantly be enhanced. For this purpose, the failure detection unit 130 may indicate to an operator or to a supervising control system a corresponding predicted time to maintenance or time to failure of a specific hardware component.
FIG. 1 d schematically shows the failure detection unit 130 in more detail in accordance with further illustrative embodiments. The unit 130 may comprise a status signal input 133, which is configured to receive the dynamic status signals 131 a, 131 b. For example, the input 133 may comprise hardware components to receive the signals 131 a, 131 b in analogous or digital form, depending on the configuration of the detector 132. In other embodiments, the input 133 may have incorporated therein an analog to digital converter to provide the dynamic status signals 131 a, 131 b in digital form for further processing in the unit 130. Moreover, a data pre-processor 134 may be provided in some embodiments, wherein the data pre-processor 134 may be configured to operate on the data received from the input 133 and to provide the data in a format that allows a rapid and reliable extraction of information. For instance, the data pre-processor 134 may comprise any filtering and clipping mechanisms for smoothing the raw data and reducing data complexity by, for instance, discarding certain raw data outside of well-defined time slots and/or outside of predefined value ranges.
The unit 130 may further comprise a key number extractor 135 that may be coupled to the data pre-processor 134 and the input 133 to operate on the raw data as well as on the pre-processed data of the signals 131 a, 131 b. The key number extractor 135 is configured to reduce the data obtained in a statistically significant manner to provide “meaningful” numbers or number ranges that allow a rapid comparison with corresponding reference data. For instance, the key number extractor 135 may comprise means and components for integrating and/or differentiating and/or summing and/or transforming and/or multiplying and/or any combinations of these processes for operating on the data received. It should be appreciated that any other data manipulation algorithms may be implemented in the key number extractor 135 as long as these algorithms are sufficiently fast to provide the key numbers within a time period that is comparable to the operation time of a single substrate in the tool 110.
The unit 130 may further comprise a comparator 136 that is configured to compare the dynamic signals 131 a, 131 b with appropriately formatted reference data, which may be provided by an external source or which may be stored in a memory device (not shown) within the unit 130. It should be appreciated that the comparator 136 may be designed, depending on the available computational power, to appropriately compare data as provided by the input 133 and/or by the data pre-processor 134 with corresponding reference data. In one particular embodiment, the comparator is coupled to the key number extractor 135 and compares reference data, provided in the form of respective reference key numbers, with the key numbers provided by the extractor 135, thereby allowing a rapid and reliable estimation of the currently prevailing failure state of the tool 110. The comparator 136 may be implemented as a rule based fault classification engine, which may recognize failure states and classify them in accordance with a pre-established hierarchical system. For example, the various failure states recognized by the comparator 136 may represent elements of different hierarchy levels, which may reflect any tool-specific activities in response to the recognized failure status. For example, a highest hierarchy level or fault class may indicate failure states that require an immediate discontinuation of the operation of the tool 110 so that the tool 110 will not be released for the next substrate to be processed. Lower-lying hierarchy levels may indicate less “dramatic” failure states, which may in some embodiments be used for enhanced control efficiency in combination with the APC strategies that may be implemented in the control unit 120, as is previously described.
Moreover, the unit 130 may comprise a time-to-failure predictor 137, which may be configured to provide a prediction for a time to maintenance or time to failure of one or more specified hardware components. For example, the predictor 137 may estimate on the basis of the hierarchy structure, the future behavior of the tool 110. For this purpose, the predictor 137 may monitor the time development of the tool status or a portion thereof, i.e., specific key numbers may represent specified aspects of the total tool status, wherein the change and the change rate of moving from one hierarchy level to another may be used in estimating a quantitative measure for predicting the failure of a specified component. It should be appreciated that the approach with a hierarchy structure for predicting the time to failure is of illustrative nature only and other appropriate algorithms may be used. For instance, one or more of the statistical key numbers may be analyzed with respect to their time development without using any hierarchy structure, wherein the predictor 137 may instead operate on the basis of corresponding reference data that may have been obtained on the basis of specifically designed test runs or which may have been obtained on the basis of empirical data from a large number of substrates previously processed in the tool 110. Thus, the present status of the tool 110 may be described more precisely compared to conventional electroplating systems without failure status detection, wherein also enhanced process control and/or tool reliability and availability may be achieved by the predictor 137.
As a result, the present invention provides a method and a system that provides significantly increased process reliability by using a dynamic failure signal during an electroplating process so as to indicate a tool status and in particular estimate a failure status prior to processing a subsequent substrate. In particular embodiments, the dynamic failure signal may be represented by one or more voltage signals and/or one or more current signals obtained from “sensitive” areas within the electroplating reactor, such as the electrode assembly, which therefore implicitly contain information on the tool condition with respect to the currently processed substrate, wherein at least a portion of this information may be extracted and may be used at least for deciding whether or not the tool is to be released for the processing of the next substrate. In particular embodiments, in addition to identifying an invalid tool status, the information extracted from the dynamic failure signal may also be used in enhancing the control efficiency for the electroplating tool in that corresponding tool activities or maintenance activities may be associated with a plurality of tool states, which may be recognized on the basis of the dynamic failure signal. Consequently, an inline process control is established that may operate on a substrate basis, thereby significantly reducing yield loss during the formation of solder bumps.
The particular embodiments disclosed above are illustrative only, as the invention may be modified and practiced in different but equivalent manners apparent to those skilled in the art having the benefit of the teachings herein. For example, the process steps set forth above may be performed in a different order. Furthermore, no limitations are intended to the details of construction or design herein shown, other than as described in the claims below. It is therefore evident that the particular embodiments disclosed above may be altered or modified and all such variations are considered within the scope and spirit of the invention. Accordingly, the protection sought herein is as set forth in the claims below.

Claims

1. A method, comprising:

forming a plurality of bumps on a first substrate by an electrochemical deposition process in an electroplating tool;

obtaining a dynamic status signal during the processing of said first substrate, said dynamic status signal representing a dynamic behavior of at least one tool parameter during the electrochemical deposition process;

estimating a current tool status on the basis of said dynamic status signal; and

releasing said electroplating tool for forming a plurality of bumps on at least one subsequently processed substrate on the basis of said estimated tool status.

2. The method of claim 1, wherein estimating said current tool status comprises providing reference data for at least one reference status of said electroplating tool and comparing said dynamic status signal of said first substrate with said reference data.

3. The method of claim 2, wherein providing said reference data comprises defining one or more fault conditions for said at least one process parameter.

4. The method of claim 3, wherein each fault condition is associated with a specific fault procedure for the further handling of the electroplating tool.

5. The method of claim 4, wherein each fault procedure comprises establishing an updated tool status for the processing of said subsequent substrate by at least one of adapting at least one manipulated variable of a process recipe to be applied to said plurality of substrates and initiating a specific maintenance action.

6. The method of claim 1, further comprising predicting an expected time to failure of at least one hardware component of said electroplating tool.

7. The method of claim 1, wherein estimating said current tool status comprises extracting a plurality of key values from said dynamic status signal and comparing said key values with reference key values.

8. The method of claim 7, wherein extracting said plurality of key values comprises processing raw data of said dynamic status signal by at least one of compressing said raw data and filtering said raw data.

9. The method of claim 8, wherein at least some of said key values are directly calculated from said raw data.

10. The method of claim 1, wherein said dynamic status signal represents a voltage signal obtained from an anode of said electroplating tool.

11. The method of claim 1, wherein said dynamic status signal represents a current flowing through an anode of said electroplating tool.

12. A method, comprising:

processing a first substrate in an electroplating tool that operates on a single substrate basis;

monitoring a failure status of said electroplating tool on the basis of a dynamic status signal obtained during the processing of said first substrate; and

comparing said failure status with at least one reference status prior to processing an additional substrate.

13. The method of claim 12, further comprising at least one of releasing said electroplating tool for processing said additional substrate and initiating a maintenance action on the basis of said comparison.

14. The method of claim 12, further comprising obtaining reference data and defining a plurality of reference statuses that represent one or more fault conditions for at least one tool parameter.

15. The method of claim 14, wherein each fault condition is associated with a specific fault procedure for the further handling of the electroplating tool.

16. The method of claim 15, wherein each fault procedure comprises establishing an updated tool status for the processing of said next substrate by at least one of adapting at least one manipulated variable of a process recipe to be applied to said plurality of substrates and initiating a specific maintenance action.

17. The method of claim 12, further comprising predicting an expected time to failure of at least one hardware component of said electroplating tool on the basis of said failure status.

18. The method of claim 12, wherein monitoring said failure status comprises extracting a plurality of key values from said dynamic status signal that represent said failure status.

19. The method of claim 18, wherein extracting said plurality of key values comprises processing raw data of said dynamic status signal by at least one of compressing said raw data and filtering said raw data.

20. The method of claim 19, wherein at least some of said key values are directly calculated from said raw data.

21. The method of claim 12, wherein said dynamic status signal represents at least one of a voltage signal and a current signal obtained from an anode of said electroplating tool.

22. The method of claim 12, wherein processing said substrate comprises forming a plurality of bumps above respective contact areas, said bumps being configured for a direct contact to contact regions of a carrier substrate.

23. A system, comprising:

an electroplating tool having an anode assembly; and

a failure detection unit connected to said electroplating tool, said failure detection unit being configured to indicate a failure status of said electroplating tool for each process run on the basis of a dynamic status signal.

24. The system of claim 23, wherein said dynamic status signal represents at least one of a voltage signal and a current signal obtained from said anode assembly of said electroplating tool.

25. The system of claim 24, wherein said anode assembly comprises a plurality of anode segments and said dynamic status signal is obtained from each of said anode segments.

26. The system of claim 23, further comprising a control unit operatively coupled to said electroplating tool and said failure detection unit, said control unit being configured to control operation of said electroplating tool on the basis of said failure status supplied by said failure detection unit.

27. The system of claim 26, wherein said failure detection unit is further configured to indicate a type of maintenance action on the basis of said failure status so as to re-establish a valid tool status prior to starting a process run.