ECI Technology | We Keep Your Chemistry Right

Leading Provider of Chemical Management Systems for Semiconductor, PV, and PCB Industries

Articles

Automated On-line Chemical Monitoring and Control for Hot Phos and Tungsten Etch in 3D NAND

Abstract:

In the process of fabricating 3D NAND devices, the complex deposition and etch have been proven to be challenging. Two etch processes: silicon nitride sacrificial removal and W etch-back in the 3D NAND word-line formation have been identified as the two critical steps that significantly impact the 3D NAND product yields. In this paper, we present the results of an automated on-line chemical management system that were specifically developed to enable real-time monitoring and control of both the sacrificial silicon nitride removal and W etch-back processes.

Keywords—Silicon Nitride Etch, Tungsten Etch, 3D NAND, Phosphoric-Acetic-Nitric Acid, PAN

Introduction

One of the key challenges of 3D NAND is scaling stack height for higher bit density. Unlike 2D planar NAND that is constrained by lithography, the bit density of 3D NAND is limited by the complex deposition and etch process steps while stacking the NAND structures in the vertical direction. The process of fabricating 3D NAND begins with multilayered silicon nitride and oxide deposition, followed by high aspect-ratio hole etch for the channel and word-line. The silicon nitride in the word-line is a sacrificial layer that is removed by immersion wet-etch, followed by dielectric (ONO) and tungsten metal gate, deposition and etch-back [1]. In this process flow, the silicon nitride sacrificial removal and W etch-back have been identified as the two critical steps that require accurate real-time metrology and process control.

Critical Wet Etch Processes

A. Sacrificial Silicon Nitride Etch using Hot Phosphoric Acid

The method of using hot phosphoric (Hot Phos) acid to etch silicon nitride is well understood and has been used in semiconductor manufacturing for many years. The control of temperatures and water content in H3PO4 was found critical in controlling the nitride and oxide etch rates. It was also found that seasoning the Hot Phos etching bath with silicate can further reduce the etching rate of SiO2 and improve the etch selectivity. Theoretically, a critically high etch selectivity can be achieved by seasoning the H3PO4 with high concentration of silica. Nevertheless, maintaining a stable etch process with such a high etch selectivity over time has been proven difficult to achieve without real-time monitoring and control, due to the dynamic bath loading behavior and etch by-products. A reliable real-time monitoring and control of Si is also important to prevent process induced defects due to Si precipitation.

At ECI, we have developed a suite of methods [2,3] designed to accurately analyze the components of the Hot Phos etch bath for stable and reliable monitoring and control of the etch process. These methods not only enable a reliable and stable etch process in the life time of the etching solution, but also the feed and bleed and cost savings that extend the lifetime of the etching baths. We have demonstrated that real-time results can be obtained using the methods implemented in our automated on-line system QualiSurf QSF-500 (see figure 1). To ensure the real-time results are accurate, we measured and compared the results with off-line Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES) (see Figure 2).

Figure 1. QualiSurf QFS-500 Series
Figure 2. Comparison of Measured Si vs ICP-AES

In the experiments of seasoning and etch, we demonstrated the capability of accurate monitoring and control of the Hot Phos silicon nitride etch process. The results are shown in Figure 2 and Figure 3.

Seasoning Process

Figure 3. Measured Si ppm in Hot Phos Seasoning Process
Figure 4. Monitoring Si during Feed and Bleed, Etch Process

B. PAN Tungsten (W) Etch

For a well-controlled selective etch of aluminum over Si or SiO2, PAN (Phosphoric-Acetic-Nitric acid) is commonly used. PAN is also considered for the W etch-back in the 3D NAND process. Similar to Aluminum etch, W oxidizes in the nitric acid forming a by-product W(NO3)x, which dissolves in the phosphoric acid. Acting as a wetting agent, the acetic acid in PAN facilitates the etch process by removing the H2 by-product.

Over the lifetime of the PAN solution, the concentration of H3PO4 increases due to the evaporation loss of the Nitric/Acetic/H20. To maintain a stable and consistent etch rate of W, the H3PO4 concentration must be controlled. Figure 5 illustrates the consequences of inconsistent etch of W where 3D NAND devices will short when W under-etches. At ECI, we have developed an on-line automated chemical management system that accurately monitors and controls the components of PAN. In a spiking experiment, different concentrations of H3PO4 were added into the bath. The system accurately measured the H3PO4 component concentrations as shown in Figure 6.

Figure 5. Tungsten (W) under and over-etch

 

Figure 6. PAN spiking experiment showing matched results of measured and expected

Conclusion

The demand for a higher bit density in 3D NAND will continue to push the limits of the fabrication process and stack height. The monitoring and control of the process becomes critically important as the number of stacking layers increases. In this paper, we presented the results of real-time on-line automated solutions in accurately monitoring and control Si3N4 and W etch.

References

[1] J.H. Jang, H.S.Kim, W.Cho and W.S.Lee, “Vertical cell array using TCAT(Terabit Cell Array Transistor) technology for ultra-high density NAND flash memory,” IEEE Symposium on VLSI Technology, page 192-193 2009.

[2] ECI Technology, Inc. Press Release, “Quali-Surf Qualifies in Japan and Taiwan Fabs”, Totowa, NJ, Feb 2, 2012.

[3] C. N. Bai, G. Liang, E. Shalyt, “Metrology for High Selective Silicon Nitride Etch”, Solid State Phenomena, Vol. 255, pp. 81-85, 2016.

Filed Under: Articles Tagged With: , , , ,

Voltammetric Detection of Low Copper Concentrations in Nickel Plating Baths

Abstract

Nickel electroplating is widely used in semiconductor manufacturing, primarily during the packaging stage. It is not used as a final coating, but instead as a barrier layer to prevent formation of copper–tin intermetallic compounds that affect the reliability of solder joints. The nickel is deposited from baths containing nickel salt (in relatively high concentrations), boric acid, and other ions.
The quality of the deposited nickel is highly dependent on the composition of the plating bath. Metallic contaminants are acceptable when their concentrations are below approximately 30 ppm. Copper, lead, zinc, and cadmium, even in relatively small quantities (higher than 30 ppm) produce a dull, black, or skip plate condition in the low-current-density areas. These metals may be removed from the plating solution by low-current-density dummyplating, but a sensitive and accurate analytical method must be used to determine when to treat the bath. Copper is considered a main contaminant due to its higher concentrations in the bath and its most detrimental effect on the nickel deposit. To prevent plating defects, the bath contaminants must be monitored.

Key Words

Nickel, electroplating, copper, contamination

I. Introduction

Traditionally, low concentrations of metals in plating solutions can be monitored using highly sensitive polarography methods or spectral methods such as AAS or ICP [1]. These methods can detect parts per trillion of copper and other metals. However, utilization of mercury electrodes makes this method less desirable due to safety and environmental concerns. In nickel-plating solutions, concentrations of copper are high enough that it would be more appropriate to measure copper with a solid electrode. Mercury electrodes are more sensitive than solid electrodes, but in the case of nickel-plating baths, such sensitivity is not required.
Prior publications describe methods for the determination of low copper concentrations using solid electrodes [2]. These methods work well when copper is present in low (ppm) concentrations and other compounds are present in low amounts or absent.In the nickel-plating bath, the main challenge is that concentrations of  nickel and other components (boric acid, chloride, or bromide ions) are very high, while the concentrations of copper are much lower [3,4]. For that reason, a new analytical method was carefully verified for possible interference with other bath components. In addition, nickel plating baths are utilized at high temperatures (40-50°C) to prevent precipitation of boric acid. This required additional precautions during analysis as boric acid can precipitate and distort the results. Results showed that if highly concentrated and hot nickel solutions are pre-treated (diluted and cooled) prior to analyses, the results of copper analysis can be accurate and reproducible.

II. Experimental Details

Chemicals and Materials- Solutions were prepared with boric acid (Fisher), nickel (II) chloride hexahydrate (Sigma-Aldrich), sodium chloride (Fisher), 50% aqueous nickel (II) sulfamate (Palm), Nikal BP wetting agent (Dow), and sodium bromide (Spectrum). For dilution tests pH 4.00 reference standard buffer from Ricca was used.

Instrumentation- Analyses were performed using an ECI QualiLab QL-10 bench top plating bath analyzer. A 4 mm Platinum Rotating Disk Electrode, an Ag/AgCL electrode with 0.1 M KCL junction solution, and a stainless steel rod counter electrode comprised the three-electrode system.

Procedures- Samples were prepared by dissolving properly measured quantities in de-ionized water. The analyzer performed modified pulse voltammetric stripping analysis (MPVS) on the samples. During the electrochemical scan, the platinum electrode surface was polarized with negative voltage to accumulate copper as per equation (1) and then subsequently polarized with positive voltage to dissolve accumulated metal.

Cu2 + 2e <=> Cu0

The reaction of copper deposition and dissolution is quite reversible, allowing collection of multiple electrochemical cycles. The dissolution peak area was selected as a main analytical signal. Electrochemical parameters such as scan rate and deposition potential were tested and optimized. The ranges are 10 to 50 mV/sec and -0.25 to -0.35 V (vs. Ag/AgCl reference electrode) respectively. The rotation speed of the platinum disk electrode was also validated in the range between 100 and 6000 rpm. Data was then processed using a proprietary algorithm where Cu peak areas were determined and compared.

II. Results and Discussions

A.      Parameter optimization

To establish a suitable electrochemical signal, initial parameter screening was performed. It was observed that the electrochemical outputs (voltammogram peaks) were sensitive to the changes in deposition voltage, scan rate, and rotation speed of the platinum disc electrode. The shape of the copper dissolution peak affects the subsequent data processing. The analysis parameters were optimized to provide the highest and most reproducible dissolution peak.

Fig. 1 shows optimization results of platinum electrode rotation rate.

Fig. 1. Rotation rate effect on voltammograms obtained from nickel bath with copper contaminant

As this graph shows, the most suitable peak shape is achieved at the highest possible rotation rate. This is because the copper concentration in solution is quite low and can be easily depleted in the layer near the electrode. Further increase in rotation rate was not beneficial due to high turbulence in the electrochemical cell that caused solution disturbance and splashing. 6000 rpm was found to be the optimum rotation rate and provides reproducible and suitable data for peak processing. In all subsequent electrochemical experiments, the rotation rate of the platinum electrodes was set at 6000 rpm.
The concentration of copper can potentially be as high as 30 ppm. However, as previously mentioned, the nickel plating baths contain high concentrations of boric acid and are kept hot to allow boric acid to remain in solution. When the bath is taken for analysis, the temperature can drop, causing precipitation of boric acid. This makes further analysis complicated. The solubility of boric acid at room temperature is about 47 g/l, but the presence of other ions in the solution will reduce its solubility [5]. We determined that for analytical purposes, boric acid concentration should be maintained below 30 g/l.
When rotation rate and dilution factor were optimized, copper calibration experiments were conducted (Fig. 2). These tests were performed in a copper concentration range from 0 to 10 ppm. The responses are linear through a wide range of rotation rates. However, at the highest rpm, the response is strongest, but not as linear as at lower rpm.
The increase in currents at higher rotation rates is expected due to an increase in the supply of reactants to the electrode surface. This should agree with the Levich equation (2), where mass transport limited currents are proportional to the square root of the rotation rate.

IL = (0.620)*n*F*A*D2/3*w1/2*v-1/6* C,

where ILis the Levich current (A), n is the number of moles of electrons transferred in a half reaction, F is the Faraday constant (C/mol), A is the electrode area (cm2), D is the diffusion coefficient (cm2/s), w is the angular rotation rate of the electrode (rad/s), v is the kinematic viscosity (cm2/s), and C is the concentration (mol/cm3).

Fig 2. Effect of rotation rate on plating change

Fig. 3 shows a linear relationship between the square root of the rotation rate and currents obtained during copper deposition experiments

Fig 3. Current as a function of rotation rate

Similar results were observed when scan rate was varied. Increased scan rate caused shorter deposition time and reduction of copper deposited on the electrode. This was undesirable as the decrease in the copper deposition reduces the sensitivity of the analysis. A linear relationship was observed for the entire range of scan rates tested.

B.Verification of method accuracy

The effects of all components in the bath were validated prior to the final testing of the analytical procedure.

The effects of all components in the bath were validated prior to the final testing of the analytical procedure.

Concentration changes in boric acid and nickel sulfamate have been shown to slightly distort the copper peak, affecting final analytical results. Table I summarizes the data collected in this study. It must be noted that pH changes could negatively affect the final result as well. This effect was observed when the plating solution was diluted by 50% with DI water. The result did not improve when higher dilutions were used.

Solution Signal, %
Target Nickel bath 100
Target Ni bath with 20% lower Boric Acid 74
Target Ni bath with 10% lower Ni Sulfamate 84

This interference could not be reduced by altering electrochemical parameters. Generally, the interference significantly increased when any of the main electrochemical settings were altered.
Because the bath must be diluted to avoid crystallization of boric acid, several different diluents were tested. We observed that dilution with pH 4 buffer (close to pH of the plating bath) aids in the elimination of the possible effects of boric acid and sulfamate. Essentially, both of these materials can affect pH values and buffer use or dilution simply cancels those effects.
During our testing, pH 4 buffer was used for bath dilution. The previously noted commercial buffer solution has enough buffering capacity to maintain the pH within ±0.1 units after dilution. Changes in the concentrations of multiple components in the bath had no effect (or only a negligible effect that was within analysis accuracy) on the analytical signal. A desirable dilution of the bath was achieved and boric acid did not precipitate.

Fig. 4 shows voltammograms obtained from the same nickel target bath with boric acid concentrations that are varied by dilution with DI water and with pH 4 buffer. Dilution with pH 4 buffer shows a clear advantage. The data are shown in Table II.

Solution Diluent Signal, %
Target Nickel bath pH Buffer 100
Target Ni bath with 20% lower Boric Acid pH Buffer 99.5
Target Ni bath with 10% lower Ni Sulfamate pH Buffer 99.8
Target Ni bath with 20% lower Boric Acid DI Water 74
Fig 4. Effect of bath dilution on voltammograms

Scans obtained from baths with varying boric acid concentrations looked very similar and subsequently led to the same integrated peak area. When all work on interferences was completed, we validated its precision by performing multiple analyses of three solutions with different Cu concentrations and target concentrations of other components. As shown in Table III, the results were highly repeatable, producing a Relative Standard Deviation (RSD) below 2%.

Table III Reproducibility of Analysis

Solution tested (5x each) Relative Standard Deviation, %
Cu 5 ppm 1.2
Cu 10 ppm 1.4
Cu 20 ppm 1.1

IV. Conclusion

A newly developed electrochemical method provides reproducible results of analysis for copper in nickel plating baths. This method has advantages in comparison to polarographic methods as it does not use a mercury working electrode. During this study, it was demonstrated that the matrix effects of other bath components can be eliminated by dilution with commercial buffer solution. This new analytical method can be easily automated to provide fast online (within 10 minutes) analysis results for copper in nickel electroplating baths.

Michael Pavlov, Mitchell Coffin, Danni Lin, and Eugene Shalyt
ECI Technology
60 Gordon Drive
Totowa, NJ 07512 USA
Ph: 973-773-8686; Fax: 973-773-8797
Email: mpavlov@ecitechnology.com

V. References

[1]    J. Heyrovsky, J. Kuta, “Principles of polarography”, Elsevier, (September 11, 2013)
[2]    B. Feier, I. Bajan, I. Fizesan, “Highly selective Detection of Copper (II) Using N, N – bis (acetylacetone)ethylenediimine as a receptor”, Int. J. Electrochem. Sci., (October, 2015), 121-139
[3]    M. Schlesinger, M. Paunovic, “Modern Electroplating”, Fifth edition, (2010)
[4]      D. Snyder, “Nickel Electroplating”, Products Finishing, Internet publication, (September 29, 2011)
[5]      R. Weast, Handbook of Chemistry and Physics, 63rd Edition, 1982-1983

Filed Under: Articles Tagged With: , , ,

Metrology for High Selective Si Nitride Etch

Silicon Nitride etch has been a building block of Semiconductor manufacturing for many years. The overall Si etch rate is dominated by the combination of process temperature and %H2O. Selectivity is controlled by Si level. Water content can be monitored through conductivity, refractive index, or the most preferred method, non-contact Near Infrared (NIR) spectroscopy. There is a variety of commercial analyzers designed for this purpose.

The main challenge is measurement of Si. We have previously described an automated method for analysis of Si in traditional Si3N4 etching solution. However, high selectivity processes require new solutions.

H3PO4 and H2O Measurement

H3PO4 and its counterpart H2O are measured by both NIR spectroscopy and conductivity methods. Table 1 summarizes the performance of the two methods.

Table 01

Comparison of H3PO4 results between on-line automated NIR method and off-line ICP-MS:

The on-line results are comparable to those of ICP-MS, but with much better time response (<5min) and automated sampling/feedback (lab analysis by ICP-MS can take several weeks with fab logistics.)

figure 02
figure 03

Conductivity

Conductivity represents mobility of the ions when under the driving force of an electrical field and is highly sensitive to temperature. Modern temperature control devices enable efficient temperature control so that the effect of temperature is greatly suppressed. The figure above shows a typical conductivity calibration curve with temperature correction, which has a good correlation with H3PO4 concentration.

Si Measurement

Silicon is measured by adding predetermined concentrations of Flouride ions to a predetermined amount of etchant solution, and measuring the potential of a Flouride Ion Specific Electrode (FISE) in this test solution. Under ideal conditions, the potential (E) of a FISE is given by the well known Nernst equation:

E = E0 – (2.303 RT/F) log [F–]

Si Measurement in Low Temperature Etch

Si is measured in a wet bench low temperature hot phosphoric etch process.

figure 04

Organosilicate Measurement

  • Reagent with Carboxylic acid added improves the sensitivity.
  • Sensitivity is further studied with various fractions of Acetic acid in the reagent
figure 05
figure 06
  • The accuracy of this method with Carboxylic acid in the reagent is evaluated by off-line ICP-MS method. The results from this improved Flouride method match those from ICP-MS.
  • Good stability of Organosilicate results by the same method with Carboxylic acid in the reagent.
  • All measured results have an accuracy of <2%.
figure 07
figure 08
figure 08

Conclusions

A variety of methods have been developed to measure the Silicon Nitride etch process bath in realtime. Results from these analyses can be used for tight process control to achieve high selectivity for Silicon Nitride removal.

References:

[1]S.J. Baffat, M.S. Lucey, M.R. Yalamanchilli “Hot Phosphoric Acid APC for Silicon Nitride Etch”, Semiconductor International, 8/1/2002
[2]Hong et al. “Compositions for Etching and Methods of Forming a Semiconductor Device Using the Same”, US Patent 9,136,120
[3]Cho et al. “Etching Composition and Method for Fabricating Semiconductor Device Using the Same”, US Patent 8,821,752
[4]Nowling et al. “Low Temperature Etching of Silicon Nitride Structures Using Phosphoric Acid Solutions”, US Patent 8,716,146
[5]Shalyt et al. “Analysis of Silicon Concentration in Phosphoric Acid Etchant Solutions”, US Patent 8,008,087
[6]E. Shalyt, G. Liang, P. Bratin, C. Lin “Real-Time Monitoring for Control of Cleaning and Etching Solutions” Proceeds of SPWCC Conference, USA, 2007
[7]Shalyt et al. “Analysis of Silicon Concentration in Phosphoric Acid Etchant Solutions” US Patent Application 20160018358

Filed Under: Articles Tagged With: , ,

Advanced Monitoring of TMAH Solution

Authored by Michael MacEwan, Jingjing Wang, Eugene Shalyt, Chuannan Bai, Guang Liang, Vishal Parekh
Presented at the 12th Annual Symposium on Ultra Clean Processing of Semiconductor Surfaces in Brussels, Belgium, 2014.

ECI Technology, Totowa, NJ 07512, USA

Outline

  • Project Goal
  • Importance of TMAH in Semiconductor Processing
  • Analysis Method Development
    • TMAH
    • Carbonate
    • Surfactant
  • Field Process Data
  • Conclusion

Project Goals

Develop Metrology to Facilitate Process Improvement and Knowledge for TMAH Solutions.

Chemical Structure of TMAH

Tetra Methyl Ammonium Hydroxide – a non-metallic source of the Hydroxyde ion.

Advantages of TMAH

  • No alkali metals {Li, Na, K} used as potential contaminant
  • Used in positive photoresist developers
  • Anisotropy (111):(100) ~ 1:10 to 1:35
  • Does not significantly etch SiO2 or Al (bond wire safe)

TMAH Applications – Lithography

It desolves acidic photoresist after exposure

Rate is concentration dependent and slows over time.

TMAH Applications – Etch

Etch rate is effected by TMAH: rate increases at lower concentration.

Methods of Analysis for TMAH

Application requirements are 0.002 wt% accuracy.

Analysis of TMAH – pH

  • Examples of concentration change on pH:
  • 0.01 wt% change in TMAH ~ 0.1 pH change
  • 0.001 wt% change in TMAH ~ 0.01 pH change

Current on-line techniques (ISFET, Optical) have performance to 0.1 pH units – not enough sensitivity

Analysis of TMAH – Titration

Titration is effected by the presence of carbonate
40ppm CO32- will cause 0.006 wt% error

Analysis of TMAH – NIR

TMAH concentration range from 0-3 wt%

Error is within ±0.015 wt% for all samples – still not enough accuracy to achieve the goal.

Analysis of TMAH – Conductivity

Excellent linearity within process operating range

±0.001 wt% reproducibility and accuracy obtained by temperature correction.

Analysis of Carbonate – Spectroscopy

  • Surfactant does not interfere with CO32- analysis
  • Test sample contains trace levels of CO32-
  • Range of 0-40ppm CO32- is linear

Carbonate absorption in TMAH

CO32- increase significantly in unprotected sample over 14 hours

Analysis of Surfactant – Surface Tension

Surfactant in TMAH solution decreases the ST
ST measurement is not sensitive in 172-175 ppm range.

Analysis of Surfactant – CVS

CVS of TMAH-Surfactant sample – Method is sensitive to low level of surfactant change.

CVS Response for Surfactant Concentration

Increasing surfactant concentration has a stronger effect.

Proposed configuration of Analyzer

TMAH Etch Field Process Data

Conclusion

  1. A comprehensive suite of automated analytical methods has been developed to meet the challenges of modern semiconductor TMAH applications for Sigma Etch and PR development:
  2. TMAH/DI
  3. Carbonate
  4. Surfactant

Methods have been successfully field-validated.

Thank you for your attention.

Filed Under: Articles Tagged With:

Advanced Process Control Of Nickel Electrodeposition for Packaging in Semiconductor Industry

Authored by Eugene Shalyt, Jingjing Wang, Vishal Parekh, Michael MacEwan
Presented at the ElectroChemical Society’s 226th Biannual meeting in Cancun, Mexico, 2014.

ECI Technology, Totowa, NJ 07512, USA

Outline

  • Use of Nickel Electrodeposition in Packaging
  • Components of electrolyte and their function
  • Metrology approaches and field validation with specific focus on
    • Surfactant
    • Leached Photoresist
    • Sulfur-Bearing Additive and/or Breakdown Product
    • Nickel
  • Conclusions

Use of Ni Electrodeposition

Nickel is not used as a final layer. It is used as a diffusion barrier between Cu and other final layers: Sn and it’s alloys, Pd, and Au.

Barrier Properties of Ni Electrodeposit

Components of Electrolyte

Root Cause of Pitting

Pitting Remediation

  • Reduce amount of H2 bubbles
    • maintaining pH > 4 to reduce partial current of H2 formulation
      however Ni(OH)2 formation can occur above pH 4.5
  • Reduce lifetime of bubbles
    • Intense agitation
    • Use of surfactant/wetter to lower surface tension

Effect of Surfactant on Voltammogram

Suppression vs. Wetter Property

There is no link. Chemical can have both properties, none, or only one.

Measurement of Surfactant by Surface Tension

  • Response is not linear
  • Matrix and temperature dependent
  • Surface tension results cannot be used for precise measurements

Measurement of Surfactant by Titration

R- SO3–Na+ + R4NCl–  –> RSO3NR4 + NaCl

Measurement of Leached Photoresist

Field Data on Accumulation of PR

Effect of PR on Voltammogram

Effect of S Compound

S Compound

Source:

  • Added on purpose (Saccharine derivative)
  • Sulfamate breakdown product

Monitoring of S Compound

  • HPLC
  • Polarography
  • Stress Measurement
  • Cyclic Voltammetry
  • Spectroscopy (new)

Monitoring of S Component

Field Data for S Component

Monitoring of Ni

  • Complexometric titration
    • Slow, high cost of ownership
  • Spectroscopy
    • Realtime, non-reagent, low cost of ownership

UV-Vis Spectra of Ni Solutions

Ni @ 66 g/l
matrix components are varied within process range

Majority of spectral range is highly selective to Ni and free of interferences

Monitoring of Other Components

Conclusions

  • Comprehensive suite of automated analytical methods have been developed to meet the challenges of modern semiconductor applications:
  • Nickel
  • Boric
  • Anode Activator (Cl, Br)
  • pH
  • Surfactant
  • Leached Photoresist
  • Sulfur-Bearing Additive and/or Breakdown Product

Filed Under: Articles Tagged With:

In situ monitoring of components to control dilute wet chemistries

Materials loss per cleaning cycle must be limited in advanced semiconductor process flows. Using dilute chemistries is necessary to achieve this goal. Fab engineers need to be able to analyze the concentrations of each component in processing solutions in real time to ensure that the chemistries are always in a narrowly defined process window. This article describes how multivariate methods can be applied to analyze chemistries of concentrations ranging from conventional to ultra-dilute levels.

Material loss during cleaning is particularly problematic for advanced technology generations where there are very shallow junctions at the drain and source areas, and any excess removal of silicon can be detrimental to the resulting transistors. The International Technology Roadmap for Semiconductors (ITRS) therefore limits the materials loss of each cleaning cycle at 0.5Å for both oxide and silicon for the 65nm technology node, and 0.3Å for the 45nm technology node [1].

To achieve minimal materials loss, leading R&D groups have resorted to using dilute chemistries for cleaning processes. However, if the solution becomes overly dilute, then cleaning becomes ineffective. If the solution is too concentrated, then materials loss will exceed that permitted by the process specification. Therefore, concentrations of all components in a cleaning solution must be precisely controlled at all times. Effective in situ methods are required to monitor the concentrations of cleaning solutions in order to make necessary timely adjustments by spiking or diluting key components.

Currently adapted monitoring methods such as resistivity/conductivity techniques, though widely used for single component solutions (HF, for example), cannot differentiate contributions from different components in multicomponent solutions, such as SC1, SC2, DSP, etc. Titration and other laboratory techniques, though highly accurate and fully capable of differentiating co-existing components in the same solution, are relatively slow and not ideal for real-time control of solution concentrations. Conventional optical methods, typically measuring absorbance at discrete wavelengths corresponding to the absorption peaks of each component, worked well for legacy concentration ranges such as 1:1:5 or 1:1:20 SC1. But as the solutions become more dilute, the absorption signals are very weak even at the peaks, and the corresponding signal-to-noise ratio becomes very small. In some cases, absorption peaks of key components overlap each other. Such conventional optical methods lack the sensitivity mandatory for analyzing dilute chemistries. An improved method is necessary to analyze highly dilute solutions, such as 1:1:100 SC1 and 1:200 HF for advanced technology nodes.

To reliably analyze ultra-dilute multicomponent wet processing solutions in real time, the method must be noninvasive to both chemicals and wafers, meeting the automatic operation requirements for wafer fabs. The corresponding analytical results can then be fed back to the wet processing equipment through closed-loop control, so that the concentrations are maintained within process windows.

Broadband monitoring

A broadband absorption monitoring hardware system can analyze optical absorption signals of semiconductor wet processing solutions in the near-infrared (NIR, 700-2500nm) wavelength range. However, unlike simpler optical systems that measure at only a very limited number of wavelengths corresponding to absorption peaks, this hardware measures the absorption spectra of the chemical solutions across a wide range of wavelengths at very small intervals. The corresponding spectrum, which includes several hundreds of data points, is then analyzed by a proprietary software routine using a multivariate method [2] to resolve the limited number of unknowns.

In SC1 for example, the only unknowns are the concentrations of NH4OH and H2O2 balanced by DI water. The abundance of data points ensures high accuracy of the analysis; the redundancy of information compensates for noises associated with signal collection. Figure 1 shows examples of as-collected spectra by the hardware before being fed into the microprocessor for further processing.

The solutions were buffered HF, and each widely separated and differentiable spectrum corresponds to a distinctly different NH4F concentration. The two wavelength bands show examples of the discrete ranges typically used by conventional optical methods.

In many real fab cases, particularly for dilute solutions, minute changes in concentrations cannot be easily differentiated from raw data, rendering the conventional optical methods ineffective. In such cases, further processing of the spectrum by a multivariate method is mandatory to achieve the desired accuracy to determine the component concentrations. A schematic of such a setup is shown in Fig. 2a.

Light emitted from a high-stability broad-wavelength-range light source is diverted by a multiplexer into multiple channels through optical fibers to individual measuring modules. The setup is capable of measuring eight or more samples/points of different compositions.

Within each measuring module, piping (in a re-circulation loop or at point-of-use) of the wet processing equipment is encased by a flow-through cell (Fig. 2b).

Figure 02b. The flow-through cell hardware used to take the measurements

Incoming light is partially absorbed by the solution inside the piping before being collected by the optical fiber at the other side of the tubing. Light signals from multiple locations are then detected in sequence by a diode array detector before being forwarded to the computer for signal processing. The flow-through cell blocks stray light from the cleanroom and holds optical fibers onto precision locations and orientations, with respect to piping, to minimize measurement noises.

The flow-through cell is an integral part of the fluid path and does not alter the diameter from upstream or downstream piping. This eliminates concerns about optical distortion due to compression of fluid or release of dissolved air. No optics are in physical contact with the solutions in the piping and the reliability of the setup is expected to be high. Each analysis takes 1.5 sec, including software processing and collecting multiple full spectra at 0.4 millisec/spectrum.

Experimental results

One significant improvement of the current work over the authors’ earlier results [3, 4] is temperature correction. Small temperature variations at measurement points are unavoidable because of temperature changes in the solution and in the ambient, among other factors. Optical absorption signals are known to be extremely sensitive to minute temperature changes. For example, measurements of a dilute HF solution prepared at 620ppm varied perceptibly due to slight changes in temperature despite real-time correction. To resolve this problem, new spectral temperature correction software was developed that fundamentally minimizes measurement variability. The new spectral approach ensures that the measurement is insensitive to temperature variations over several degrees Celsius.

Multivariate measurements of NH4OH in ultra-dilute 1:1:500 SC1 solutions correlated better than 0.995 with the known concentrations of as-prepared solutions. Statistics of measuring H2O2 in SC1 showed similar reliability.

Figure 3 shows concentration changes of the two key components of an originally 1:1:50 SC1 mixture over a 16-hr period in a recirculation loop without active concentration control.

Figure 03. SC1 concentration changes over a 16-hr period in a recirculation loop without spiking/dilution, showing the ability to clearly resolve the components

NH4OH decreases much more slowly than H2O2, thus the latter should be replenished more frequently into the tank. Such different trends cannot be resolved by resistivity/conductivity analytical techniques since contributions from ammonium hydroxide and hydrogen peroxide cannot be differentiated.

The multivariate technique measured several different concentrations of dilute HF between 1000 and 3000ppm (corresponding to approximately 1:570 and 1:190 dilutions, respectively) over an extended period of time. The temperature of the measuring environment was not actively controlled, varying between 20° and 25°C. Results show that all datasets were within 2.5% of the known as-prepared concentrations, and the repeatability was better than 0.7% in all cases. In general, when measuring HF of a 50,000ppm concentration, a standard deviation of 200ppm (0.4% σ) is achievable.

We monitored the concentration of a dilute HF solution as it changed from 1500ppm to 8350ppm by spiking and then back to 1500ppm by dilution. Because of the short 1.5 sec analytical time of each measurement, several data points during mixing were captured in the graph (Fig. 4).

The effectiveness of the current method to measure solutions in transition can help fabs precisely control tank or point-of-use concentrations associated with process recipe changes.

Other cleaning solutions, including BEOL organic solvents, aqueous solutions, HF/H2SO4 blends, as well as many proprietary mixes were measured using the same setup, and similar accuracies as the previously mentioned examples were achieved.

Conclusion

A proven multivariate method measures ultra-dilute single- and multi-component wet processing solutions in real time. Simpler optical methods to measure concentrations often have problems associated with unavoidable small temperature variations. The multivariate method and system can be used to help control the concentrations of critical processing solutions in fabs, vital for necessary wafer cleaning without excessive material removal.

Authored by

Chenting Lin, Peter Bratin, Guang Liang, Michael Schneider, Eugene Shalyt
ECI Technology, Totowa, NJ

Reference

1. “Surface Preparation Technology Requirements: Near-term Years,” Table 68a, Front End Process, ITRS, 2005 Edition.

2. I.T. Jolliffe, Principal Components Analysis, 2nd ed., Springer, 2002.

3. E. Shalyt, et al., “Real-time Monitoring of Dilute Multicomponent Wet Processing Chemistries,” pp. 163-172, 25th SPWCC Proceedings, 2005.

4. Y. Shekel, et al., “Real-time Chemical Monitoring by NIR Spectroscopy,” pp. 245-250, Proceedings of the 208th ECS Meeting, Los Angles, CA, 2005.

Filed Under: Articles Tagged With: