U.S. patent application number 15/185658 was filed with the patent office on 2016-10-13 for deterministic feedback blender.
The applicant listed for this patent is Air Liquide Electronics U.S. LP. Invention is credited to Kevin T. O'DOUGHERTY.
Application Number | 20160296902 15/185658 |
Document ID | / |
Family ID | 57111242 |
Filed Date | 2016-10-13 |
United States Patent
Application |
20160296902 |
Kind Code |
A1 |
O'DOUGHERTY; Kevin T. |
October 13, 2016 |
DETERMINISTIC FEEDBACK BLENDER
Abstract
Methods and systems for high precision, continuous blending of
mixtures, and particularly mixtures having at least two distinct
chemical components, are disclosed. More particularly, the
disclosed methods and systems provide high precision, continuous
blending of buffered oxide etch mixtures containing water, ammonium
fluoride, and hydrofluoric acid.
Inventors: |
O'DOUGHERTY; Kevin T.;
(Arden Hills, MN) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Air Liquide Electronics U.S. LP |
Dallas |
TX |
US |
|
|
Family ID: |
57111242 |
Appl. No.: |
15/185658 |
Filed: |
June 17, 2016 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G05D 11/138
20130101 |
International
Class: |
B01F 15/04 20060101
B01F015/04; G05D 11/13 20060101 G05D011/13; B01F 15/00 20060101
B01F015/00 |
Claims
1. An inline blending method of components that alter each other's
assays, the method comprising: a. mixing Component A, Component B,
and a solvent in an inline blender to form a mixture; b. analyzing
the mixture downstream of the inline blender to determine a
concentration of Component A and a concentration of Component B; c.
maintaining a target concentration in the mixture of Component A
within 0.008% w/w and maintaining a target concentration in the
mixture of Component B within 0.22% w/w by adjusting a flow rate of
Component A, Component B, and/or the solvent based on the
concentration of Component A and the concentration of Component
B.
2. The inline blending method of claim 1, wherein the solvent is
ultra purified water.
3. The inline blending method of claim 2, wherein Component A
comprises species that affect the concentration of Component B.
4. The inline blending method of claim 2, wherein Component B
comprises species that affect the concentration of Component A.
5. The inline blending method of claim 4, wherein Component A is HF
and Component B is NH.sub.4F.
6. The inline blending method of claim 2, wherein Component A is
NH.sub.4OH and Component B is H.sub.2O.sub.2.
7. The inline blending method of claim 2, wherein Component A is
HCl and Component B is H.sub.2O.sub.2.
8. The inline blending method of claim 2, wherein Component A is
H.sub.2SO.sub.4 and Component B is H.sub.2O.sub.2.
9. A method of mixing chemical fluids in an inline blender to
produce a mixture having a concentration of the chemical fluids
within 0.22% w/w of a target concentration, the method comprising:
a. introducing a first chemical fluid into an inline blender via a
first flow control device; b. introducing a second chemical fluid
into the inline blender via a second flow control device; c.
introducing a third chemical fluid into the inline blender via a
third flow control device; d. mixing the first chemical fluid, the
second chemical fluid, and the third chemical fluid in a mixing
zone of the inline blender to form a mixture; e. monitoring the
mixture downstream from the mixing zone for a first chemical fluid
concentration and a second chemical fluid concentration; f.
adjusting the first flow control device based on the first chemical
fluid concentration; and g. adjusting the second flow control
device based on the second chemical fluid concentration.
10. The method of claim 9, wherein the first flow control device,
the second flow control device, and the third flow control device
is an orifice, a flow control valve, a stepper throttle, or
combinations thereof.
11. The method of claim 9, wherein the mixing zone is a tube or a
pipe having an unrestricted flow path.
12. The method of claim 9, wherein the mixing zone comprises an
element to induce mixing.
13. The method of claim 12, wherein the element to induce mixing is
a static mixer, a stirrer, a vortex element, or combinations
thereof.
14. The method of claim 9, wherein the first chemical fluid
concentration and the second chemical fluid concentration are
monitored by combinations of a conductivity meter, a pH meter, a
refractometer, a turbidity monitor, a Raman spectrometer, an
infrared spectrometer, a UV/VIS spectrometer, a densitometer, an
ultra-sonic meter, and a particle counter.
15. The method of claim 9, wherein the first chemical fluid and the
second chemical fluid are not water.
16. The method of claim 9, wherein the first chemical fluid is
H.sub.2SO.sub.4, HF, NH.sub.4OH, or HCl.
17. The method of claim 9, wherein the second chemical fluid is
H.sub.2O.sub.2 or NH.sub.4F.
18. The method of claim 1, wherein the third chemical fluid is
water.
19. The method of claim 9, wherein the first chemical fluid
comprises ions that affect the second chemical fluid
concentration.
20. The method of claim 9, wherein the second chemical fluid
comprises ions that affect the first chemical fluid concentration.
Description
TECHNICAL FIELD
[0001] Methods and systems for high precision, continuous blending
of mixtures, and particularly mixtures having at least two distinct
chemical compounds, are disclosed. More particularly, the disclosed
methods and systems provide high precision, inline blending of,
among others, buffered oxide etch mixtures containing water,
ammonium fluoride, and hydrofluoric acid.
BACKGROUND
[0002] In various industries, chemical delivery systems are used to
supply chemicals to processing tools. Illustrative industries
include the semiconductor, pharmaceutical, biomedical, food
processing, household product, personal care product, or petroleum
industries.
[0003] The chemicals being delivered by a given chemical delivery
system depend, of course, on the particular processes being
performed. Accordingly, the particular chemicals supplied to a
semiconductor processing tool depend upon the processes being
performed inside the tool. Illustrative semiconductor processes
include etching, cleaning, chemical mechanical polishing (CMP), and
wet deposition (e.g., spin-on, copper electroless, electroplating,
etc.).
[0004] Commonly, two or more fluids are combined to form the
desired mixture for the particular process. The mixtures may be
prepared in batches, on- or off-site, and then shipped to the
processing location. Alternatively, the mixtures may be prepared at
the point-of-use using a suitable blender system.
[0005] U.S. Pat. No. 6,050,283 to Air Liquide America Corp.
discloses a system and method for mixing and/or diluting ultrapure
fluids, such as liquid acids, for semiconductor processing.
[0006] U.S. Pat. No. 6,799,883 to Air Liquide America L.P.
discloses a method and apparatus for continuously blending a
chemical solution for use in semiconductor processing.
[0007] U.S. Pat. No. 6,923,568 to Celerity, Inc. discloses a method
and apparatus for blending and supplying process materials,
particularly ultra-high purity chemicals.
[0008] U.S. Pat. No. 7,344,297 to Air Liquide Electronics U.S., LP,
discloses a method and apparatus for asynchronous blending and
supply of chemical solutions.
[0009] A need remains for methods and systems for high precision,
continuous blending of solutions.
NOTATION AND NOMENCLATURE
[0010] Certain abbreviations, symbols, and terms are used
throughout the following description and claims, and include:
[0011] As used herein, the indefinite article "a" or "an" means one
or more.
[0012] As used herein, the terms "approximately" or "about"
mean.+-.2% of the value stated.
[0013] As used herein, the term "inline" or "continuous" means that
the blending process substantially simultaneously feeds the
chemicals to while removing the product mixture from a mixing zone
without interruption. The inline blending process is distinct from
a "batch" process, in which defined quantities of chemicals are
mixed, typically in a mixing tank, to produce a batch, or specific
quantity, of the product mixture.
[0014] As used herein, the term "slurry" means a chemically active
or buffered solution containing suspended solids. Slurries are
typically used to remove and/or planarize deposited materials.
[0015] As used herein, the term "species" means atoms, molecules,
molecular fragments, ions, etc. resulting from the methods
disclosed. In other words, the disclosed blending methods of any of
the components, fluids, acids, bases, oxidizers, reducers, or
chemicals disclosed herein may produce species (i.e., atoms,
molecules, molecular fragments, ions, etc.) of the item subject to
blending.
[0016] As used herein, the phrase "maintaining a target
concentration in the mixture of Component X within # % w/w" means
not permitting the concentration of the mixture to exceed the # %
w/w difference from the target concentration.
SUMMARY
[0017] Inline blending methods are disclosed. The components that
are blended may alter each other's assays, complicating the
standard inline blending processes which typically mix components
based on volume or mass. Component A, Component B, and a solvent
are mixed in an inline blender to form a mixture. The mixture is
analyzed downstream of the inline blender to determine the
concentration of Component A and the concentration of Component B.
A target concentration in the mixture of Component A is maintained
within 0.008% w/w and a target concentration in the mixture of
Component B is maintained within 0.22% w/w by adjusting the flow
rate of Component A, Component B, and/or the solvent based on the
concentration of Component A and the concentration of Component B.
The disclosed inline blending methods may include one or more of
the following aspects: [0018] maintaining the target concentration
notwithstanding the source of Component A and/or Component B and/or
the solvent; [0019] the mixture being a solution; [0020] the
mixture being a slurry; [0021] the slurry comprising a basic
solution; [0022] the slurry comprising an acidic solution; [0023]
the slurry comprising abrasive particles; [0024] the abrasive
particles being silica (SiO.sub.2), alumina (Al.sub.2O.sub.3),
calcium carbonate (CaCO.sub.3), ceria (CeO.sub.2), zirconia
(ZrO.sub.2), or titania (TiO2); [0025] the solvent being ultra pure
water; [0026] the inline blender being a static mixer; [0027] the
inline blender being a stirrer; [0028] the inline blender being a
vortex element; [0029] the inline blender being a combination of a
static mixer and a vortex element; [0030] downstream of the inline
blender being a tube or pipe having an unrestricted flow path;
[0031] downstream of the inline blender not being a mixing or
holding tank; [0032] wherein Component A, Component B, and the
solvent are mixed together in one inline blender to form the
mixture; [0033] wherein Component A, Component B, the solvent, and
the abrasive particles are mixed together in one inline blender to
form the mixture; [0034] maintaining the target concentration of
Component A within 0.001% w/w; [0035] maintaining the target
concentration of Component B within 0.001% w/w; [0036] Component A
comprising species that affect the concentration of Component B;
[0037] Component B comprising species that affect the concentration
of Component A; [0038] the flow rate of Component A, Component B,
and/or the solvent being adjusted using a PID algorithm; [0039]
Component A being HF and Component B being NH.sub.4F; [0040]
Component A being a slurry and Component B being H.sub.2O.sub.2
[0041] Component A being NH.sub.4OH and Component B being
H.sub.2O.sub.2; [0042] Component A being HCl and Component B being
H.sub.2O.sub.2; [0043] Component A being H.sub.2SO.sub.4 and
Component B being H.sub.2O.sub.2; [0044] Component A being
NH.sub.4OH and Component B being ammonium acetate; [0045] Component
A being NaHCO.sub.3 and Component B being NaOH; [0046] analyzing
the concentration of Component A using a different analyzer than
that used to measure the concentration of Component B; [0047] the
concentration of Component A and the concentration of Component B
being monitored by a combination of a conductivity meter, a pH
meter, a refractometer, a turbidity monitor, a Raman spectrometer,
an infrared spectrometer, a UV/VIS spectrometer, a densitometer, an
ultra-sonic meter, and a particle counter; [0048] the concentration
of HF being determined using a pH meter; [0049] the concentration
of the slurry being determined by a densitometer; [0050] the
concentration of H.sub.2O.sub.2 being determined by a
refractometer; [0051] the concentration of HCl being determined
using a pH meter; [0052] the concentration of HF being determined
using a conductivity meter; [0053] the concentration of HCl being
determined using a conductivity meter; [0054] the concentration of
H.sub.2SO.sub.4 being determined using a conductivity meter; [0055]
the concentration of NH.sub.4F being determined using a
conductivity meter; [0056] the concentration of NH.sub.4OH being
determined using a conductivity meter; [0057] the concentration of
H.sub.2O.sub.2 being determined using a refractometer; [0058]
adjusting the flow rate of Component A using a PID algorithm;
[0059] adjusting the flow rate of Component B using a PID
algorithm; [0060] the flow rate of HF being adjusted using a direct
acting PID algorithm; and [0061] the flow rate of NH.sub.4F being
adjusting using a reverse acting PID algorithm.
[0062] Methods of mixing chemical fluids in an inline blender to
produce a mixture having a concentration of the chemical fluids
within 0.22% w/w of a target concentration are also disclosed. A
first chemical fluid is introduced into an inline blender via a
first flow control device. A second chemical fluid is introduced
into the inline blender via a second flow control device. A third
chemical fluid is introduced into the inline blender via a third
flow control device. The first chemical fluid, the second chemical
fluid, and the third chemical fluid are mixed in a mixing zone of
the inline blender to form a mixture. The mixture is monitored
downstream from the mixing zone for a first chemical fluid
concentration and a second chemical fluid concentration. The first
flow control device is adjusted based on the first chemical fluid
concentration. The second flow control device is adjusted based on
the second chemical fluid concentration. The disclosed methods may
include one or more of the following aspects: [0063] maintaining
the target concentration notwithstanding the source of the first
chemical fluid and/or the second chemical fluid and/or the third
chemical fluid; [0064] the first flow control device, the second
flow control device, and the third flow control device being an
orifice, a flow control valve, a stepper throttle, or combinations
thereof; [0065] the mixing zone being a tube or a pipe having an
unrestricted flow path; [0066] the mixing zone comprising an
element to induce mixing; [0067] the element to induce mixing being
a static mixer, a stirrer, a vortex element, or combinations
thereof; [0068] downstream of the inline blender being a tube or
pipe having an unrestricted flow path; [0069] downstream of the
inline blender not being a mixing or holding tank; [0070] the
mixture being a solution; [0071] the mixture being a slurry; [0072]
the slurry comprising a basic solution; [0073] the slurry
comprising an acidic solution; [0074] the slurry comprising
abrasive particles; [0075] the abrasive particles being silica
(SiO.sub.2), alumina (Al.sub.2O.sub.3), calcium carbonate
(CaCO.sub.3), ceria (CeO.sub.2), zirconia (ZrO.sub.2), or titania
(TiO2); [0076] wherein the first chemical fluid, the second
chemical fluid, and the third chemical fluid are mixed together in
one mixing zone of the inline blender to form the mixture; [0077]
maintaining the target concentration of the first chemical fluid
within 0.008% w/w; [0078] maintaining the target concentration of
the second chemical fluid within 0.22% w/w; [0079] the first
chemical fluid and the second chemical fluid not being water;
[0080] the first chemical fluid being H.sub.2SO.sub.4, HF,
NH.sub.4OH, or HCl; [0081] the second chemical fluid being
H.sub.2O.sub.2 or NH.sub.4F; [0082] the third chemical fluid being
water; [0083] the first chemical fluid being HF and the second
chemical fluid being NH.sub.4F; [0084] the first chemical fluid
being NH.sub.4OH and the second chemical fluid being
H.sub.2O.sub.2; [0085] the first chemical fluid being HCl and the
second chemical fluid being H.sub.2O.sub.2; [0086] the first
chemical fluid being H.sub.2SO.sub.4 and the second chemical fluid
being H.sub.2O.sub.2; [0087] the first chemical fluid being
NH.sub.4OH and the second chemical fluid being ammonium acetate;
[0088] the first chemical fluid being NaHCO.sub.3 and the second
chemical fluid being NaOH; [0089] the first chemical fluid
comprising ions that affect the second chemical fluid
concentration; [0090] the second chemical fluid comprising ions
that affect the first chemical fluid concentration; [0091] the ions
being selected from H.sup.+, NH.sub.4.sup.+, SO.sub.4.sup.2-,
F.sup.-, OH.sup.-, or Cl.sup.-; [0092] the ions being selected from
H.sup.+, NH.sub.4.sup.+, H.sup.-, OH.sup.-, OOH.sup.-,
O.sub.2.sup.-, or F.sup.-; [0093] analyzing the concentration of
the first chemical fluid using a different analyzer than that used
to measure the concentration of second chemical fluid; [0094] the
first chemical fluid concentration and the second chemical fluid
concentration being monitored by a combination of a conductivity
meter, a pH meter, a refractometer, a turbidity monitor, a Raman
spectrometer, an infrared spectrometer, a UV/VIS spectrometer, a
densitometer, an ultra-sonic meter and a particle counter; [0095]
the first chemical fluid concentration being monitored by a
temperature-adjusted conductivity meter; [0096] the second chemical
fluid concentration being monitored by a temperature-adjusted pH
meter; [0097] the concentration of HF being determined using a pH
meter; [0098] the concentration of slurry being determined by a
densitometer; [0099] the concentration of H.sub.2O.sub.2 being
determined by a refractometer; [0100] the concentration of HCl
being determined using a pH meter; [0101] the concentration of HF
being determined using a conductivity meter; [0102] the
concentration of HCl being determined using a conductivity meter;
[0103] the concentration of H.sub.2SO.sub.4 being determined using
a conductivity meter; [0104] the concentration of NH.sub.4F being
determined using a conductivity meter; [0105] the concentration of
NH.sub.4OH being determined using a conductivity meter; [0106] the
flow rate of H.sub.2O.sub.2 being adjusted using a refractometer;
[0107] adjusting the flow rate of the first chemical fluid using a
PID algorithm; [0108] adjusting the flow rate of second chemical
fluid using a PID algorithm; [0109] the flow rate of HF being
adjusted using a direct acting PID algorithm; and [0110] the flow
rate of NH.sub.4F being adjusting using a reverse acting PID
algorithm.
[0111] Methods of mixing acids and bases in an inline blender to
produce mixtures having consistent concentrations are also
disclosed. A solvent is introduced into an inline blender via a
first flow control device. An acid is introduced into the inline
blender via a second flow control device. A base is introduced into
the inline blender via a third flow control device. The acid, base,
and solvent are mixed in a mixing zone of the inline blender to
form a mixture. The mixture is monitored downstream from the mixing
zone for an acid concentration and a base concentration. The second
flow control device is adjusted based on the acid concentration.
The third flow control device is adjusted based on the base
concentration. The disclosed methods include one or more of the
following aspects: [0112] maintaining the acid concentration and
the base concentration notwithstanding the source of the acid
and/or the base; [0113] the first flow control device, the second
flow control device, and the third flow control device being an
orifice, a flow control valve, a stepper throttle, or combinations
thereof; [0114] wherein the acid, base, and solvent are mixed
together in one mixing zone of the inline blender to form the
mixture; [0115] the mixing zone being a tube or a pipe having an
unrestricted flow path; [0116] the mixing zone comprising an
element to induce mixing; [0117] the element to induce mixing being
a static mixer, a stirrer, a vortex element, or combinations
thereof; [0118] downstream of the mixing zone being a tube or pipe
having an unrestricted flow path; [0119] downstream of the mixing
zone not being a mixing or holding tank; [0120] the mixture being a
solution; [0121] monitoring the acid concentration using a
different analyzer than that used to measure the base
concentration; [0122] the acid concentration and base concentration
being monitored by a combination of a conductivity meter, a pH
meter, a refractometer, a turbidity monitor, a Raman spectrometer,
an infrared spectrometer, a UV/VIS spectrometer, a densitometer, an
ultra-sonic meter and a particle counter; [0123] maintaining the
target concentration of the acid within 0.01% w/w; [0124]
maintaining the target concentration of the base within 0.01% w/w;
[0125] maintaining the target concentration of the acid within
0.001% w/w; [0126] maintaining the target concentration of the base
within 0.001% w/w; [0127] the solvent being water; [0128] the acid
and the base not being water; [0129] the acid comprising ions that
affect the base concentration; [0130] the acid being HF; [0131] the
ions being selected from the group consisting of H.sup.+ and
F.sup.-; [0132] the base comprising ions that affect the acid
concentration; [0133] the base being NH.sub.4F; [0134] the ions
being selected from the group consisting of H.sup.+,
NH.sub.4.sup.+, H.sup.-, or F.sup.-; [0135] the base concentration
being monitored by a temperature-compensated conductivity meter;
and [0136] the acid concentration being monitored by a
temperature-compensated pH meter.
[0137] Methods of mixing an oxidizer and a reducer in an inline
blender to produce a mixture having a consistent concentration are
also disclosed. A solvent is introduced into an inline blender via
a first flow control device. An oxidizer is introduced into the
inline blender via a second flow control device. A reducer is
introduced into the inline blender via a third flow control device.
The oxidizer, reducer, and solvent are mixed in a mixing zone of
the inline blender to form a mixture. The mixture is monitored
downstream from the mixing zone for an oxidizer concentration and a
reducer concentration. The second flow control device is adjusted
based on the oxidizer concentration. The third flow control device
is adjusted based on the reducer concentration. The disclosed
methods may include one or more of the following aspects: [0138]
maintaining the oxidizer concentration and the reduce concentration
notwithstanding the source of oxidizer and/or the reducer; [0139]
the first flow control device, the second flow control device, and
the third flow control device being an orifice, a flow control
valve, a stepper throttle, or combinations thereof; [0140] wherein
the oxidizer, reducer, and solvent are mixed together in one mixing
zone of the inline blender to form the mixture; [0141] the mixing
zone being a tube or a pipe having an unrestricted flow path;
[0142] the mixing zone comprising an element to induce mixing;
[0143] the element to induce mixing being a static mixer, a
stirrer, a vortex element, or combinations thereof; [0144]
downstream of the mixing zone being a tube or pipe having an
unrestricted flow path; [0145] downstream of the mixing zone not
being a mixing or holding tank; [0146] the mixture being a
solution; [0147] monitoring the oxidizer concentration using a
different analyzer than that used to measure the reducer
concentration; [0148] the oxidizer concentration and reducer
concentration being monitored by a combination of a conductivity
meter, a pH meter, a refractometer, a turbidity monitor, a Raman
spectrometer, an infrared spectrometer, a UV/VIS spectrometer, a
densitometer, an ultra-sonic meter and a particle counter; [0149]
maintaining the target concentration of the oxidizer within 0.01%
w/w; [0150] maintaining the target concentration of the reducer
within 0.01% w/w; [0151] maintaining the target concentration of
the oxidizer within 0.001% w/w; [0152] maintaining the target
concentration of the reducer within 0.001% w/w; [0153] the solvent
being water; [0154] the oxidizer and the reducer not being water;
[0155] the oxidizer comprising ions that affect the reducer
concentration; [0156] the oxidizer being H.sub.2O.sub.2 or
NH.sub.4OH; [0157] the ions being H.sup.+, NH.sub.4.sup.+, or
OH.sup.-; [0158] the reducer comprising ions that affect the
oxidizer concentration; [0159] the reducer being H.sub.2O.sub.2,
H.sub.2SO.sub.4, or HCl; [0160] the ions being H.sup.+, H.sup.-,
OH.sup.-, OOH.sup.-, O.sup.2-, HSO4.sup.-, SO4.sup.2-, or Cl.sup.-;
[0161] monitoring the reducer concentration by a
temperature-compensated conductivity meter; and [0162] monitoring
the oxidizer concentration by a temperature-compensated pH
meter.
[0163] An inline chemical blender for mixing chemicals in real-time
is also disclosed. Three chemical fluid flow control devices
communicate three chemical fluids from their sources to a mixing
zone of the inline chemical blender. Three sensors are located
downstream of the mixing zone, one of which monitors the
temperature of the mixture and two of which measure the
concentrations of two of the chemical fluids in the mixture. The
disclosed inline chemical blender may include one or more of the
following aspects: [0164] each flow control device being an
orifice, flow control valve, stepper throttle, or combinations
thereof; [0165] the mixing zone being a tube or pipe; [0166] the
mixing zone being a tank or reservoir; [0167] downstream of the
mixing zone being a tube or pipe having an unrestricted flow path;
[0168] downstream of the mixing zone not being a mixing or holding
tank; [0169] measuring the concentration of a first chemical fluid
using a different sensor than that used to measure the
concentration of a second chemical fluid; [0170] maintaining a
concentration of the first chemical fluid in the mixture and a
concentration of the second chemical fluid in the mixture
notwithstanding the source of the first chemical fluid and/or the
second chemical fluid; [0171] the sensors being combinations of a
temperature sensor, a conductivity meter, a pH meter, a
refractometer, a turbidity monitor, a Raman spectrometer, an
infrared spectrometer, a UV/VIS spectrometer, a densitometer, an
ultra-sonic meter and a particle counter; [0172] the sensors
communicating with the flow control devices via a PID algorithm;
[0173] the PID algorithm adjusting the flow rate of a chemical
fluid proportionally to the pH results of the chemical fluid;
[0174] the PID algorithm increasing the flow rate of a chemical
fluid when the pH results of the chemical fluid increase; [0175]
the PID algorithm adjusting the flow rate of a chemical fluid
inversely to the conductivity results of the chemical fluid; [0176]
the PID algorithm decreasing the flow rate of a chemical fluid when
the conductivity results of the chemical fluid increase; and [0177]
further comprising a drain downstream of the mixing zone.
BRIEF DESCRIPTION OF THE DRAWINGS
[0178] For a further understanding of the nature and objects of the
present invention, reference should be made to the following
detailed description, taken in conjunction with the accompanying
drawings, in which like elements are given the same or analogous
reference numbers and wherein:
[0179] FIG. 1 is a block schematic diagram of a prior art inline
blender;
[0180] FIG. 2 is a graph showing concentration (wt. %) versus
conductivity (mS/cm) of two blend points produced by the prior art
inline blender of FIG. 1;
[0181] FIG. 3 is a block schematic diagram of an inline blender
with a proposed offline chemical analysis feedforward step;
[0182] FIG. 4 is a block schematic diagram of the disclosed inline
blender;
[0183] FIG. 5 is a graph showing pH versus HF concentration for
different molarity NH.sub.4F solutions; and
[0184] FIG. 6 is the graph demonstrating how to select the pH and
conductivity set points necessary to produce the blend having the
desired concentration, notwithstanding the assay of the source
materials.
DESCRIPTION OF PREFERRED EMBODIMENTS
[0185] Inline blending multiple chemical constituents in a way that
is accurate and repeatable is difficult when the chemical
constituents interact with and affect each other. Even if the assay
of each chemical constituent consistently remains within a
specification window, the effect of compounding error may make it
difficult to control the specification for the final blended
product.
[0186] The prior art volume-based and weight-based blending
processes do not address this issue, since they depend solely on
the known specification of the incoming material.
[0187] Metrology-based feedback inline blenders use techniques such
as conductivity, pH, refractive index, density, turbidity, and
other measurements, in order to control the flow rate of the
chemicals and maintain the final blended material
specification.
[0188] Previous continuous inline blenders, of the type
schematically represented in FIG. 1, have a series of "Blend
Points" with one constituent species added at each blend point and
controlled locally with a feedback metrology device, typically
conductivity, but also Refractive Index, pH, Ultra Sonic,
densitometers, etc. The first complication to final blend accuracy
in these series of continuous inline blenders arises at the second
blend point. The concentration of the first blend point is immune
to the single species source assay, such as 49% HF acid, because
the first series blend point with feedback provides for the correct
first blend point assay even if the chemical source is 48% HF acid.
The feedback from the first blend point allows more flow of the 48%
HF source to compensate. However, the second blendpoint is not
immune to the single species source assay, even when the second
chemical source is a single species, such as NH.sub.4F having no
other added species, such as HF or NH.sub.3. if the single species
NH.sub.4F source assay varies from the expected 40% NH.sub.4F w/w
(e.g., 39% or 41% NH.sub.4F w/w), the variance ripples upstream and
alters the intended assay of the first blend point by diluting or
concentrating it. In other words, the addition of the second
chemical having an assay that differs from expected results in an
improper concentration of the first chemical in the final
solution.
[0189] Concentration control is further impacted when the two
chemical compounds to be blended alter each other's concentrations.
In other words, prior art systems use of subsequent sensors to
measure the concentration of each sequentially added component
fails to address the impact a subsequent component may have on the
concentration of an earlier component. As a result, some solutions
cannot be blended without ambiguous output based upon the prior art
single component/single sensor embodiment. For example, any traces
of NH.sub.3 or HF in NH.sub.4F may negatively affect the total HF
concentration in a buffered oxide etch solution.
[0190] A term used to describe the robustness of a blender for
source chemical concentrations is "Source Immunity". This describes
the level to which the final inline blend concentration target can
be maintained with variations in the concentration of the input
source chemicals. Single Species Source Immunity exists only at the
first blend point. In other words, the final output of the blending
with feedback from a conductivity signal for simple HF dilution
blending is immune to the input HF concentration (i.e., 48% wt/wt,
49% wt/wt, 46% wt/wt, etc.).
[0191] For each subsequent series blend point (i.e., second, third,
etc.), the precise source assay input must be known and constant,
or concentration or dilution of the first constituent may take
place and result in an out of specification assay condition for the
first component at the second and/or third blend points. This type
of blend error ripples upstream through the series of continuous
inline blenders. The second component may be out of specification
at the third and fourth blend points.
[0192] This continuous inline blender limitation is worsened when
certain species are present in more than one of the chemical
constituents being blended. For example, NH.sub.4F may include
H.sub.2O, OH.sup.-, NH.sub.4F, NH.sub.3, NH.sub.4.sup.+, HF,
H.sup.+, and F.sup.-. When HF and NH.sub.4F are blended, any HF or
F-- present therein may alter the concentration of HF in the final
blend.
[0193] In the semiconductor industry, due to the high purity
required for the chemicals and blends containing these chemicals,
gas phase generation of the chemical or blend is typical and
preferable compared to simple solid phase dissolution of various
salts etc. As a result, the final assay may also include free
species in the gas phase due to difficulty controlling the final
gas phase end point.
[0194] When multiple chemicals are blended, for example deionized
water (DIW)+Chemical A+Chemical B, the prior art teaches that
Chemical A is added at a first chemical mixing zone, controlled by
a first metrology feedback, and Chemical B is added sequentially
into a second chemical mixing zone, controlled by a second
metrology feedback. See, e.g., U.S. Pat. No. 6,050,283 to Air
Liquide America Corp. Additional chemical additions are possible,
each one added in sequence with a dedicated metrology feedback
control. When the incoming source material specification changes,
the blender described above is not able to detect or correct for
that condition for all upstream chemicals, and the blended material
may be out of specification.
[0195] Accurate mixing of chemicals is particularly important in at
least the semiconductor industry due to the continuous shrinking of
the product size necessitated by Moore's law. Failure to maintain
the specified concentrations may introduce product variations or
even failure. Product degradation may occur when the product is
shipped from off-site or if a batch sits for too long.
[0196] Buffered chemical solutions using salt buffers are
particularly afflicted by the limitations of the prior art
continuous inline blenders. For example, many buffered oxide etch
formulations will use an ammonium fluoride (i.e., NH.sub.4F) salt
buffer to create a buffered hydrofluoric acid solution for etch
control. The original NH.sub.4F solution is manufactured by
dissolving NH.sub.3 gas into a hydrofluoric acid solution and
attempting to stop the manufacturing process at the neutral point
(i.e., the point at which there is no excess HF or NH.sub.3 in the
solution). If the reaction is stopped too early, the final ammonium
fluoride becomes HF rich and if the reaction is stopped late, the
resulting ammonium fluoride would be NH.sub.3 rich. Alternatively,
the entire ammonium fluoride salt buffer may be generated from the
gas phase by first dissolving anhydrous HF into water to form HF
acid and then dissolving NH.sub.3 gas into the HF acid to formulate
the ammonium fluoride. In this case, if the reaction is stopped too
early, the final ammonium fluoride becomes HF rich and if the
reaction is stopped late, the resulting ammonium fluoride would be
NH.sub.3 rich. As a result, the final blend assay is affected by
the unknown and non-constant free species in the ammonium fluoride
source material.
[0197] Preparation of a buffered oxide etch solution is described
as follows using a typical prior art blender shown in FIG. 1 in
block schematic form. In FIG. 1, Ultrapure Water, or UPW, 100 is
blended with 49% w/w HF 200 at blendpoint 250 to produce dilute HF.
The UPW 100 may be delivered to blendpoint 250 using a flow meter
(not shown). The HF 200 may be delivered to blendpoint 250 using a
proportioning valve (not shown). The proportioning valve (not
shown) may deliver more or less HF 200 to the blendpoint 250 based
on temperature corrected conductivity feedback 251 to achieve the
desired dilute HF assay. The dilute HF produced at blendpoint 250
is at the desired concentration independent of the actual HF 200
source assay (i.e., source immunity).
[0198] The dilute HF produced at blendpoint 250 is then blended
with ammonium fluoride 300, which may contain free NH.sub.3 or HF,
at blendpoint 350 to produce the buffered oxide etch solution 400.
Although depicted as a "block" in FIG. 1, one of ordinary skill in
the art will recognize that the buffered oxide etch solution 400 is
delivered either directly to the processing equipment or to a
drain. The ammonium fluoride 300 may be delivered to blendpoint 350
using a proportioning valve (not shown).
[0199] At blendpoint 350, the first general error may arise as a
function of the actual ammonium fluoride assay--40.0%+/-1%
NH.sub.4F by wt. If this assay is not known and held constant
across all sources of ammonium fluoride 300, the HF assay in the
buffered oxide etch solution 400 may be forced out of
specification. In addition, the final HF concentration may be
affected by the unknown and non-constant free species, NH.sub.3 or
HF, existing in the gas phase production generation of ammonium
fluoride source material 300. The final free HF assay in the
buffered oxide etch solution 400 typically requires a very high
tolerance due to the sensitive process in which the etch solution
is used.
[0200] As a result, the buffered oxide etch solution 400 only
produces acceptable HF assays for one exact and specific assay of
the ammonium fluoride source material 300, for example 40.0%
NH.sub.4F and 0.05% Free HF by wt. Every time a new lot of ammonium
fluoride 300 is brought online or if the ammonium fluoride 300
undergoes any changes in concentration while being used, the
blender set points must be changed to produce the required buffered
oxide etch solution assays. The ammonium fluoride assay may be
performed using titration, functional silicon wafer etch tests, or
a combination of both, all of which are time consuming and time
sensitive.
[0201] The UPW 100, HF 200, and NH.sub.4F 300 are retrieved from
industry standard storage units. The storage units are in fluid
communication with blendpoint 250 and 350. Blendpoint 250 and 350
are in fluid communication with a drain (not shown) and processing
equipment (not shown). More particularly, supply lines supply the
UPW 100, HF 200, and NH.sub.4F 300 to the blendpoints 250 and 350.
Similarly, supply lines supply the resulting buffered oxide etch
solution to a drain (not shown) or processing equipment (not
shown). The supply lines are industry standard supply lines. The
supply lines may include flow control devices, such as orifices,
flow control valves, stepper throttles, or combinations thereof;
valves, such as check valves or electronic valves; and/or pressure
regulators.
[0202] Blendpoint 250 or 350 is a mixing zone, also known as an
inline blender. The mixing zone may be a tube or pipe having an
unrestricted flow path. Alternatively, the mixing zone may be a T
junction. In either alternative, the tube or pipe may include an
element to induce mixing, such as a static mixer, a stirrer, a
vortex element, or combinations thereof. One of ordinary skill in
the art will recognize that the suitable mixing zone will be
determined based upon both the mixing conditions required and
pressure necessary at the point of use. More particularly, a
viscous solution or slurry may require more forceful mixing than a
flowing solution and, as a result, an element to induce mixing.
However, an element to induce mixing may result in loss of pressure
across the system. One of ordinary skill in the art will recognize
the mixing zone necessary based on equipment setup and the product
to be mixed.
[0203] The supply line downstream of blendpoint 250 or 350 includes
a sensor (not shown) that measures a characteristic of the blend.
Exemplary sensors include conductivity meters, pH meters,
refractometers, turbidity monitors, Raman spectrometers, infrared
spectrometers, a densitometer, an ultra-sonic meter, spectrometers,
and particle counters. The sensors provide real-time feedback
(i.e., 251, 351) to the relevant flow control device and adjusts
the flow rate based on the analysis results. For example, if the
conductivity is a little high, a conductivity meter may instruct
the flow control device for HF 100 to decrease the flow rate. If
the conductivity is too high so that the resulting buffered oxide
etch solution 400 is unusable, the sensor may instruct a valve in
the line to divert the solution to the drain. The sensors and flow
control devices may communicate via a controller. The communication
may occur via electrical wiring or wireless communication links.
The controller may include a processor that is programmable to
implement any one or more suitable types of process control, such
as proportional-integral-derivative (PID) feedback control.
Exemplary controllers include the PLC Simatic S7-300 system from
Siemens Corp. Also commonly used are Allen Bradley CompactLogix PLC
Control and the Allen Bradley RSLogix programming suite.
[0204] Surprisingly, Applicants have found that neither the
refractometer nor the conductivity meter were capable of
determining the concentration of free species in the resulting
buffered oxide etch solution. However, pH analysis properly
reflects free species in the solution.
[0205] FIG. 2 is a graph that illustrates the resulting two blend
point operating curve that demonstrates the inadequacy of the prior
art inline blender of FIG. 1. The graph shows the conductivity (in
mS/cm) versus concentration (in wt. %) of two blend points: 250 and
350. Point 250 provides the concentration and conductivity of the
solution at blendpoint 250 in FIG. 1, more particularly, 0.08% w/w
HF and 3.300 mS/cm. Point 350 provides the concentration and
conductivity of the solution from blendpoint 350 in FIG. 1, more
particularly, 14.8% w/w NH.sub.4F and 160 mS/cm. The operating
curve shown in FIG. 2 is only valid for one unique ammonium
fluoride source 300 assay. To obtain the buffered HF 400 having
14.8% wt. NH.sub.4F and 0.08% wt. HF from a HF 200 source and an
ammonium fluoride source 300 of 40% NH.sub.4F by wt. and 0.000%
free species, the blendpoint 250 is set to 3.300 mS/cm and the HF
is added from source HF 200 until a conductivity of 3.300 mS/cm is
reached at blendpoint 250. NH.sub.4F is then added from source
NH.sub.4F 300 at blendpoint 350 to a conductivity of 160 mS/cm. If
the ammonium fluoride source 300 changes, the 3.300 mS/cm setting
at blendpoint 250 will no longer be valid to obtain the final
required buffered HF 400 solution. A unique blendpoint 250 must be
found for every ammonium fluoride source 300 assay as explained
above.
[0206] Various techniques were tested to try to minimize or remove
this ammonium fluoride assay dependence, including reversing the
blend order to blend ammonium fluoride first followed by HF, all
unsuccessfully, because there is no method of compensating for the
varying assays of the NH.sub.4F source material using the blender
arrangement of FIG. 1.
[0207] Based on the limitations of the prior art continuous inline
blender of FIG. 1, a "Feedforward" step was proposed that includes
the assay of the ammonium fluoride 300 and the amount of free
species NH.sub.3 or HF to alter the setpoint for HF source at the
blendpoint 250 analysis. A block schematic of this proposal is
shown in FIG. 3. An automated means (not shown) provides the
concentration of free species and ammonium fluoride in the ammonium
fluoride source 300 to blendpoint 250 as indicated by the
feedforward step 352. The automated means must remain accurate for
the NH.sub.4F assay as well as free species NH.sub.3 or HF across a
full domain of source 300 input possibilities, pH for example, vs.
just one single value--the final disclosed blender setpoint. Based
on mass balance information and calculations, the new setpoint 352
was sent to the blendpoint 250. Various automated metrology devices
were explored to provide this automated solution, including
conductivity, refractive index and pH. Due to mass balance offsets,
convoluted intermediate steps with errors, and difficulty obtaining
the true assay of the ammonium fluoride 300, not to mention the
long time interval, consumables and equipment costs necessary to
run such an automated analytical assay process, this proposal was
rejected. Blendpoint 350 of FIGS. 2 and 3 must remain fixed because
the final buffered oxide etch solution 400 assay must be fixed.
[0208] FIG. 4 is a block schematic of the inline blender that
solves the problems that have not been resolved in the prior art.
As will be described in more detail below, blendpoint 250 of FIGS.
1 and 3 has been eliminated and blendpoint 350 now provides
temperature compensated pH feedback 251 to the proportioning valve
(not shown) for HF 200 and temperature compensated conductivity
feedback 351 to the proportioning valve (not shown) for the
ammonium fluoride source 300.
[0209] Chemical equilibrium arises within a buffering salt and
acid. As discussed above for the buffered oxide etch solution,
using NH.sub.4F as the salt buffer results in the following
chemical solution species: [0210] H.sub.2O (I) [0211] HF (aq)
[0212] H.sup.+ [0213] OH.sup.- [0214] NH.sub.4.sup.+ [0215]
NH.sub.3 [0216] F.sup.- In FIG. 4, the NH.sub.4F salt is fully
dissolved in the aqueous solution at NH.sub.4F source 300. The pH
feedback 251 and conductivity feedback 351 must take place after
complete mixing of UPW 100, HF 200, and NH.sub.4F 300 (i.e.,
completion of any chemical reaction, salt formation from NH.sub.3
rich material, etc.) and establishment of a full chemical
equilibrium for all the ion species.
[0217] At chemical equilibrium, an Initial, Change, Equilibrium
("ICE") matrix for the components to be reacted may be established
to determine the sensor settings needed. Alternatively, for many
cases, the shortcut Henderson-Haselbalch equation may be used
(i.e., pH=pK.sub.a+log.sub.10 ([A-]/[HA]), wherein K.sub.a is the
acid dissociation constant, [HA] is the molar concentration of the
un-dissociated weak acid, and [A-] is the molar concentration of
the conjugate base of HA). One of ordinary skill in the art will
recognize that the ICE matrix or Henderson-Haselbach equations are
temperature dependent.
[0218] The resulting chemical operating curve is shown in FIG. 5,
which is a graph showing pH of the blend versus HF molar
concentration for different molarity NH.sub.4F solutions. The FIG.
5 graph shows a family of curves representing the molarity of the
salt buffer NH.sub.4F. 10 curves are shown for NH.sub.4F molarity
ranging from 0.5 M to 5.0 M. The molarity of the NH.sub.4F buffer
solution provides the desired conductivity reading. For example,
the solid line represents 0.5 M NH.sub.4F; the dotted line
represents 1.0 M NH.sub.4F; the triangled line represents 1.5 M
NH.sub.4F; the squared line represents 2.0 M NH.sub.4F; the
dot-dash line represents 2.5 M NH.sub.4F; the dashed line
represents 3.0 M NH.sub.4F; the dot-dot-dash line represents 3.5 M
NH.sub.4F; the circled line represents 4.0 M NH.sub.4F; the line
represents 4.5 M NH.sub.4F; and the diamond line represents 5.0 M
NH.sub.4F. Data of concentration versus conductivity for various
chemicals are published at 25.degree. C. For each molarity of
NH.sub.4F above, a specific temperature compensated conductivity is
manifested. For highest accuracy in the disclosed blender, the
actual temperature compensated conductivity at each salt buffer
concentration is determined empirically.
[0219] Each NH.sub.4F molarity curve intersects the pH and HF
concentration axes at different points. As shown by the solid graph
grid line at pH=4.5, a constant pH may be maintained with
increasing HF concentration due to the buffering effect obtained by
the increasing NH.sub.4F M. The desired concentration of the oxide
etch buffer solution may be selected to determine the needed
molarity/conductivity of the NH.sub.4F and pH of the resulting
oxide etch buffer solution. For example, as shown in FIG. 6, if the
desired blend contains 14.8% NH.sub.4F by wt. (4.00 M or 160.00
mS/cm) and 0.08000% HF by wt (0.04 M), the resulting pH is 5.1. The
resulting setpoints for the blender sensors using this example are:
[0220] Temperature Compensated Conductivity 351=160.00 mS/cm [0221]
Temperature Compensated pH 251=5.1
[0222] Instead of a cascaded series of inline blend points used by
the traditional continuous blender, just one blend point 350 may be
used as the confluence of all the constituents, and feedback is
established to obtain the setpoints from the operating curve shown
in FIG. 5. As a result, the blender occupies a smaller footprint
than the prior art multi-blend point inline blender and further may
require less maintenance.
[0223] Experimentation in a full scale blender has demonstrated
that using conventional proportional-integral-derivative (PID)
feedback control, final blend output assay control is
automatically, (via the disclosed PID feedback control),
obtained/maintained, regardless of HF or NH.sub.4F source assays.
The conventional PID feedback control system is an electronic
closed-loop control between the flow meters, control valves (or
pump control), and analyzers. The analyzer results are used to
adjust the flow meters and/or control valves in order to maintain
the final desired blend assay.
[0224] The NH.sub.4F Molarity/Concentration is automatically
controlled with a "Reverse Acting", (SP-PV), traditional PID
algorithm--as conductivity rises, NH.sub.4F flow drops via the PID
algorithm.
[0225] The HF Molarity/Concentration is automatically controlled
with a "Direct Acting", (PV-SP), traditional PID algorithm--as pH
rises, HF flow increases via PID algorithm.
[0226] These two automatic PID feedback algorithms ensure the final
blend output is at the temperature compensated conductivity and
temperature compensated pH and therefore at the unique and required
blend output assay.
[0227] When a new NH.sub.4F source is placed online, regardless of
the NH.sub.4F assay or the amount of free NH.sub.3 or HF, final
blend assay is controlled by automatically via the PID feedback as
explained above to maintain the desired conductivity and pH and
thus the final stable and consistent NH.sub.4F and HF assay. The
final blend does not change as a function of the origin of the free
HF species--either in the HF source or the NH.sub.4F source as long
as the six species alone are present, mixing is complete, chemical
equilibrium is established, and process variables are temperature
compensated as discussed previously. Nor does the bulk
concentration of salt in the source NH.sub.4F effect the final
blend assay, but instead simply by the flow of NH.sub.4F driven
from the Reverse Acting PID algorithm.
[0228] The resulting solutions/slurries having the desired
concentration may be prepared on-demand without worrying about any
changes to the concentrations of the starting materials. More
particularly, the disclosed methods produce solutions and slurries
on demand that are able to maintain concentrations within 0.22% w/w
of the desired concentration. Depending on the chemical and
measuring technique, the concentration may be maintained within
0.008% w/w of the desired goal. As shown in the examples that
follow, this is an improvement over the results obtained using the
prior art blender of FIG. 1.
[0229] While tests to date have been performed using the buffered
oxide etch solution (i.e., NH.sub.4F/HF/H.sub.2O), Applicants
believe that the disclosed blending process may also be used for
other solutions requiring high precision that are common in the
electronics industry including, but not limited to, solutions such
as NH.sub.4OH/H.sub.2O.sub.2/H.sub.2O, HCl/H.sub.2O.sub.2/H.sub.2O,
or H.sub.2SO.sub.4/H.sub.2O.sub.2/H.sub.2O or slurries containing
abrasive particles, such as silica (SiO.sub.2), alumina
(Al.sub.2O.sub.3), calcium carbonate (CaCO.sub.3), ceria
(CeO.sub.2), zirconia (ZrO.sub.2), or titania (TiO2) in a basic
solution such as NH.sub.4OH/ammonium acetate/H.sub.2O or
NaHCO.sub.3/NaOH/H.sub.2O or an acidic solution such as
NH.sub.4F/HF/H.sub.2O.
EXAMPLES
[0230] The following non-limiting examples are provided to further
illustrate embodiments of the invention. However, the examples are
not intended to be all inclusive and are not intended to limit the
scope of the inventions described herein.
Disclosed Blender Example
[0231] The inline blender of FIG. 4 was built in the research
laboratory to determine whether the final concentration was
maintained after changing NH.sub.4F source 300. Due to the research
settings and waste management requirements, the concentrations in
these examples are less than typically encountered in the
industry.
[0232] The trial runs were performed with an Ultra Pure Water (UPW)
flow rate of 1.75 liters per minute (Lpm) flowing from UPW source
100. A pH probe and conductivity probe for feedback elements 251
and 351 was added after blendpoint 350. A PTFE six stage static
mixer purchased from Edlon Inc. was used for blendpoint 350.
[0233] Each Feedback Metrology Subsystem was tuned and optimized
for this UPW flow rate for maximum stability/response. This was
accomplished using the resident onboard Allen Bradley CompactLogix
PLC Control and the Allen Bradley RSLogix programming suite using
the PIDE (Proportional-Integral-Derivative Enhanced) Function
Block.
[0234] The AutoTune feature of the PIDE Function Block was used to
tune the PID parameters for optimum operation at the setpoints of
Conductivity=50.00 mS/cm, pH=5.50. These AutoTune deduced PID
parameters were rounded off and used for the Deterministic Blend
Runs.
[0235] The following parameters were set:
TABLE-US-00001 Parameter Description Parameter Value Units UPW
Blender Flow Rate 1750 mL/min NH.sub.4F Conductivity 0.0183
.degree. C..sup.-1 Temperature Compensation Alpha pH
Instrumentation Endress and Hauser No units Temperature Auto Temp
Compensation Compensation NH.sub.4F PIDE Feedback Parameters PID
Equation Type Independent Control Action Reverse Acting SP-PV No
units Loop Update Time 100 Milliseconds Deadband 0.0 K.sub.p 0.4
unitless K.sub.i 3.6 minutes.sup.-1 K.sub.d 0.007 minutes HF PIDE
Feedback Parameters PID Equation Type Independent Control Action
Direct Acting PV-SP No units Loop Update Time 100 milliseconds
Deadband 0.0 K.sub.p 20 unitless K.sub.i 80.0 minutes.sup.-1
K.sub.d 0.4 minutes
[0236] To prepare for the experimental runs of the disclosed
blender, three source containers of chemical were prepared: [0237]
1. Source 200 was 49% HF [0238] 2. Source 300 supply #1 contained
40% NH.sub.4F w/w and neutral free species, i.e. no free NH.sub.3
or HF [0239] 3. Source 300 supply #2 contained 40% NH.sub.4F w/w
and 0.9% free HF w/w
[0240] The second HF rich source 300 was designed to challenge the
system in comparison to the neutral free species source 300 supply
#1 and for the system to throttle back the amount of HF delivered
to the blend via source 200 and maintain the same final blend as
more HF is delivered via source 300 supply #2. As a result, the
concentrations of the first source 300 supply #1 will be considered
the desired concentration.
[0241] At the end of this full scale blender test cycle, samples of
the blender output were collected and retained and titrated for HF
and NH.sub.4F assay with a Mettler Toledo Excellence T70 Titrator.
External measurements of the conductivity and the pH were also
measured in a beaker using: [0242] Mettler Toledo InLab 717
Conductivity Probe [0243] Calibrated with a 12.9 mS/cm conductivity
standard. [0244] Mettler Toledo DG111-SC pH Probe [0245] Calibrated
using 5.00 and 8.00 pH standards.
[0246] The results of the two output blends from the disclosed
blender FIG. 4 are shown in the table below: Both disclosed
Deterministic Blend samples blended with source 200 HF 49% w/w.
TABLE-US-00002 External Measured External NH.sub.4F Sample
Conductivity Measured HF Assay Assay Description (mS/cm) pH (% w/w)
(% w/w) Source 300 50.7494 5.278 0.1655 2.7030 Supply #1 40%
NH.sub.4F w/w 0.00% free HF w/w 0.00% free NH.sub.3 Source 300
50.8206 5.247 0.1675 2.6881 Supply #2 40% NH.sub.4F w/w 0.90% free
HF w/w 0.00% free NH.sub.3F Assay change +0.0020 -0.0149 (% w/w)
Accuracy = % Assay 1.2% 0.56% composition change = (Assay change/
initial result) .times. 100
[0247] As can be seen, the system maintained the concentration of
HF within +0.0020% w/w and NH.sub.4F within -0.0149% w/w in the
final product notwithstanding the change of NH.sub.4F source 300.
These results are well within the claimed concentration limits.
Using the targeted temperature compensated pH and conductivity, a
final product assay with small variations was achieved. In
addition, these results were obtained on small volume samples, and
do not benefit from the long term averaging that continuous dynamic
blending provides. As a result, when implemented into a single
blend point configuration and using two control loops for the
different feeds, a much tighter control on the variation of the
targeted pH and conductivity is expected. The tighter control on
these critical control parameters will lead to a much tighter
control on the final product assay notwithstanding variations in
the feed product assay.
Prior Art Example
[0248] In order to contrast the blend output result from the prior
art blender of FIG. 1, a series blender with two blend points with
conductivity feedback as in FIG. 2 was simulated. Blendpoint 250
was reproduced in a large beaker with dilute HF to a conductivity
of 50 mS/cm as measured by the same external conductivity probe as
in the example above, a Mettler Toledo InLab 717 Conductivity
Probe. The sample was split into two small beakers. In one beaker,
source 300 supply #1 was added to a conductivity of 50 mS/cm and in
the other beaker, source 300 supply #2 was added to a conductivity
of 50 mS/cm; this is the series operation of the prior art blender
with conductivity feedback. The two beakers were then titrated with
the same Mettler Toledo Excellence T70 Titrator as in the example
above to obtain the HF assay. The conductivity and pH were measured
with the Mettler Toledo InLab 717 Conductivity Probe and the
Mettler Toledo DG111-SC pH Probe in the same way as was done in the
example above.
[0249] The results of the two output blends from the prior art
blender simulation of FIG. 1 are shown in the table below:
TABLE-US-00003 External Measured External NH.sub.4F Sample
Conductivity Measured HF Assay Assay Description (mS/cm) pH (% w/w)
(% w/w) Source 300 49.436 4.908 0.1751 Not Supply #1 measured 40%
NH.sub.4F w/w 0.00% free HF w/w 0.00% free NH.sub.3 Source 300
49.256 4.795 0.2661 Not Supply #2 measured 40% NH.sub.4F w/w 0.90%
free HF w/w 0.00% free NH.sub.3F Assay change +0.0910 N/A (% w/w)
Accuracy = % Assay +51.97% N/A composition change
[0250] As can be seen, the HF assay alone is more than 10 times
larger than the claimed target concentration window (i.e.,
.+-.0.008% w/w). As a result, the two exemplary blend output tables
confirm the essence of the disclosed invention; that the
Deterministic Blender can compensate for changes to NH.sub.4F
source 300, where the prior art is blind and unable to adjust to
the changing source 300 levels of free HF. Neither of these
examples take advantage of the inherent benefits of dynamic
blending, which generally leads to a tighter control on the
targeted final blend control characteristics i.e., pH and
conductivity. The disclosed processes simply provide a means of
having the correct target for the conductivity and pH without
concern for the feed assay.
[0251] It will be understood that many additional changes in the
details, materials, steps, and arrangement of parts, which have
been herein described and illustrated in order to explain the
nature of the invention, may be made by those skilled in the art
within the principle and scope of the invention as expressed in the
appended claims. Thus, the present invention is not intended to be
limited to the specific embodiments in the examples given above
and/or the attached drawings.
* * * * *