During the manufacture of semiconductors, hundreds of processes are used that do not take place in a constant sequence. Between these processing steps, wafers are carried and stored in dedicated plastic enclosures, known as Front Opening Unified Pods (FOUPs).
Image Credit: TOFWERK
Specific flaws in wafers have been associated with an increased amount of time between processes (“queue times”) and with the interaction of wafers with compounds outgassing from the surfaces within the FOUPs.1
Sensitive and accurate measurement of the outgassing compounds could direct process changes to reduce defects associated with queue time and improve the cleaning process of separate FOUPs before loading with new batches of wafers.
Most significantly, these measurements could inform the development of sophisticated FOUPs using new surface treatment procedures and novel polymeric materials.2 This study presents the application of a TOFWERK Vocus CI-TOF mass spectrometer for constant tracking of FOUP outgassing after a process that replicated standard cleaning procedures.
Experimental Procedure
FOUP outgassing (approximately 50 L) was tracked with the help of a Vocus CI-TOF mass spectrometer equipped with an Aim Reactor using iodide reagent ions (refer to Figure 1). The Vocus CI-TOF instantly samples the atmosphere and promptly reports the concentrations of trace organic and inorganic compounds in the atmosphere.
Experiments were performed by spraying a solution containing nitric acid (HNO3), acetic acid (CH3COOH), formic acid (CH2O2), hydrobromic acid (HBr), and hydrochloric acid (HCl) into the FOUP and subsequently flushing the FOUP with nitrogen to replicate the cleaning process.
The corresponding mass deposited into the FOUP from the solution spanned from 0.15 to 1 µg. Hydrofluoric acid (HF) was added to the FOUP through a permeation tube with an emission rate of 125 ng/minute.
Using a constant flow of N2 (2 L/minute), the inner volume of the FOUP was flushed to make sure that the interior of the FOUP was suitably blended and to replicate the cleaning of the FOUP container. This led to a FOUP ventilation rate of less than 60 minutes.
Three steps were involved in the measurement protocol: (1) quantifying the FOUP background for 5 minutes to determine the clean FOUP background, (2) positioning the HF permeation tube within the FOUP for two minutes and then instantly inject the acid solution and (3) continuously quantifying the mixing and subsequent decay of the introduced compounds until the concentrations return to background values.
Figure 1. Schematic diagram of the experimental procedure.
Image Credit: TOFWERK
Results
After the acid solution is injected, the mixing within the FOUP took around 3 to 4 minutes (including evaporation of the administered solution) before the flushing triggered the decay of analyte concentrations within the FOUP. Figure 2a demonstrates an example of nitric acid decay and also the reproducibility of the decay of acetic acid between repeated experiments (refer to Figure 2b).
Figure 2. (a) Normalized concentration C(t) of nitric acid (red) and its double exponential fit (blue). (b) Normalized concentration of acetic acid after deposition of 1 µg (first experiment) and 0.15 µg (second experiment), showing the reproducibility of the system.
Image Credit: TOFWERK
All compounds displayed a double exponential decay, with a few sticky compounds prevailing at trace concentrations of 10 to 100 pptv, even 100 minutes following injection.
The compound-dependent time constants were retrieved using the double exponential fit (Equation 1). They represent the flushing timescales of every compound from the FOUP. τ1 in equation 1 denotes the e-folding time for the rapid decay (gas volumetric exchange in the FOUP), while the second time constant (τ2) denotes the slower outgassing from FOUP surfaces.
| C(t) = C1e-t/τ1 + C2e-t/τ2 + Cb |
Equation 1 |
τ2 is considerably longer and relies on the interactions of the acid with the inner surfaces of the FOUP. Figure 2a displays an example of the double exponential fit for HNO3, which has considerable interaction with the FOUP walls and as a result persists relatively longer than other tested acids.
Figure 3 demonstrates the reaction of HNO3, HCl, HF, and HBr to nitrogen flushing over the first 45 minutes as soon as stable concentrations were reached in the FOUP. Table 1 sums up the time constants (τi) from the double exponential fits shown in Figure 3.
Figure 3. Exponential decay of different inorganic acids during the first 45 minutes of FOUP flushing. The response of the acids to flushing is related to their vapor pressure and surface interactions with the inner FOUP surfaces.
Image Credit: TOFWERK
Table 1. Decay time constants (τ) for each acid shown in Figure 3. The values of τ2 are calculated according to the fit when the concentration starts to stabilize. Source: TOFWERK
The majority of the acids have a similar response in the first few minutes in the FOUP when controlled by volumetric flushing. But on longer timescales, certain acids prevail at trace concentrations of 10 to 30 pptv for several hours, as shown in Figure 4.
Figure 4. Concentration decay of common inorganic acids in FAB environment. The markers show the quantification limit of each compound. Arrows on the right axis show the 1-minute LOD of the Vocus CI-TOF. Diamonds show the point in the 11-hour long flushing experiment where the measured signal falls below the LOD of the instrument. For HCl and HNO3, measurable signal persists even after 11 hours.
Image Credit: TOFWERK
Acetic, formic, hydrobromic, and hydrofluoric acids all reached near-background concentrations (a decrease of 90%) in the first one hour, signifying no major attenuation or memory on the inner surfaces of the FOUP (refer to Table 2).
To put this in simpler terms, cleaning these substances within an otherwise vacant FOUP is probably simple, and thus easy to predict the optimal ending point of a FOUP cleaning procedure.
But the relatively slower decay of hydrochloric and nitric acids indicates that cleaning procedures that are not improved for the slow outgassing of such acids or that cannot spot them at adequately low concentrations may suffer from later outgassing of acids from the inner surfaces of the FOUP, which could present an Airborne Molecular Contamination (AMC) issue within the FOUP atmosphere, thereby decreasing yield of wafers.
Table 2 sums up the performance of Vocus CI-TOF for detecting both organic and inorganic acids that are pertinent in semiconductor fabrication plant (FAB) settings. The instrument has a quick time response (a few seconds for most compounds, T90) that enables a single instrument to screen several different measurement points or be installed on a mobile platform to quantify at different points in the FAB.
Table 2. Vocus CI-TOF detection limits and response time. Source: TOFWERK
The excellent detection limits and easy independent operation present a paradigm shift in FAB operators’ potential to measure airborne and surface-bound AMC at ever-lower concentrations as line widths are driven to ever-smaller dimensions.
Thanks to the combination of the Vocus CI-TOF and the FAB, a better understanding of FOUP cleanliness can be achieved at pptv concentrations and in real time. By introducing a novel online acid detection technology, enhancements in the FAB operator’s potential to regulate the FOUP — and more generally the FAB surroundings — will decrease wafer defects caused by mask degradation or surface contaminations.
Although TOFWERK has emphasized the application in FOUP cleaning, there are several other applications in the FAB in which the Vocus CI-TOF would be quite suitable, for instance, the detection of trace acids in FAB AMC monitoring, estimation and control of scrubber efficiency, quality control of gases fed into deposition and etch reactors as well as in lithography equipment. All these can be met with unparalleled precision and speed by the Vocus CI-TOF.
References
- Jeong et al. Control of Wafer Slot-Dependent Outgassing Defects during Semiconductor Manufacture Processes. 2019. doi.org/10.1109/ASMC.2019.8791794.
- Gonzalez-Aguirre et al. Control of HF Volatile Contamination in FOUP Environment by Advanced Polymers and Clean Gas Purge. 2015. doi.org/10.4028/www.scientific.net/SSP.219.247.