pH Sensing

Water exists as a mixture of H⁺ and OH^– ions, continually associating and dissociating in a dance. Water molecules in solution may exist as as H₂O, as H₃O⁺, or as dissociated H⁺ and OH^–. pH is the measure of the concentration of hydronium ions (H₃O⁺) present in a solution, expressed on a logarithmic scale:

pH = -log₁₀(a_H+)

Water has a pH of ~7.  Solutions with values below this are acidic, or higher in hydronium ion activity. Solutions with values above this are basic, or lower in hydronium ion activity. Most biological processes and systems are in the pH 5 – 9 range. Blood, for example, has a pH of 7.365. Even a very small decrease in pH to 7.35 leads to acidosis, a condition which results in irreversible cell damage.

Several methods exist for pH sensing. The simplest and most inexpensive is pH paper, in which paper is infused with various indicator compounds that change color in response to pH. Though not very accurate, they are useful for estimating pH range in non-critical applications like home aquariums, swimming pools, and in titrations to determine the extent of a chemical reaction. They are useful for pH 0 through 12.

Electrodes are the most common pH sensor, using a milli-voltmeter with very high input impedance to measure the potential difference between a hydronium-sensitive electrode and a reference electrode. Electrodes work best for pH in the range 2.5 – 10.5, as the Nernst law governing their function breaks down outside this range. Some designs are capable of working at more extreme pH by using special procedures and construction. They are often salinity dependent, and are not compatible with all chemical environments.

Optrodes use the interaction of light with pH sensitive molecules to determine pH. This may be done using colorimetric absorbance, fluorescence lifetime, or fluorescence intensity, all of which use an optical detector. Though they are limited in range to about 4 pH units, they can offer increased chemical compatibility and faster response time than optrodes, and may be less affected by ionic strength (depending on the optrode type).

Our pH sensors are optrodes based on colorimetric changes of an indicator in a flexible, highly responsive sol-gel material that can be applied directly to a probe or used as a discrete patch for non-contact measurements. They are designed for use over the biological range (pH 5 – 9), and have been optimized for response time and thermal performance. Unlike other optical pH systems that measure fluorescence, our systems are ratiometric and resistant to drift. Additionally, they can be used with oxygen optrodes to provide pH and O₂ sensing in a single, small package.

Our pH sensors are based on optical detection of spectral changes in a pH-sensitive dye immobilized in a modified sol-gel matrix. Hydrogen ions from solution diffuse into the sol-gel and interact with the dye, causing a change in color between yellow (acidic) and blue-green (basic).

Light from a broadband visible/near-infrared light source travels down a transmissive or reflective probe to the pH sensor material, measuring the change in absorption at 620 nm as well as at a reference wavelength. As pH increases, so does the absorbance at 620 nm. By looking at the absorbance and comparing to absorbance of a reference solution with a pH value of 1, pH of the sample can be calculated using a ratiometric algorithm. The method used is designed to minimize the effects of fiber movement, light source drift, moderate changes in dye concentration, and environmental factors, providing stable and reliable pH readings. Calibration against a series of standard buffer solutions ensures accuracy.

Our sensors are based on an organically modified sol-gel (Ormosil) which has been engineered to minimize impact of ionic strength on pH measurements. As compared to a polymer matrix, it offers faster response time, more flexibility in indicator materials, greater chemical compatibility, and enhanced thermal and optical performance. Our sol-gel formulation offers a fast response, as low as 10 seconds, and maintains calibration for over a month with only a single baseline correction. The flexible structure of the sol-gel allows it to easily be incorporated into a variety of product packages.

Our early pH optrodes operated on the same principle described above, but measured absorbance of the sol-gel in transmission using a transmissive dip probe. The disadvantage of this was that if the sample solution was colored, turbid, or had sediment, the optical signal was affected. Ambient light and bubbles trapped within the probe tip also impacted measurement accuracy, and the displaced mirror surface at the probe tip was susceptible to degradation in harsh environments.

To eliminate these effects, we developed a reflective sensor design whereby the pH sensor material could be applied directly to the transparent wall of a vessel for measurement with an external probe, with the solution in contact on the backside of the sensor patch. A layer of gold mesh allows hydrogen ions to diffuse freely from solution into the sol-gel material, while providing enough reflectivity to eliminate the need for a discrete mirror at the probe tip. The inert nature of gold ensures broad chemical compatibility and good reflectivity at the wavelengths of measurement.

Our ratiometric method of pH sensing requires two wavelengths: 620 nm and a reference wavelength to correct for drift. The reference wavelength must be one at which the absorbance of the sensor does not change with pH. It can be the isosbestic point (~510 nm for our sensors), or any wavelength ≥750nm.

A tungsten halogen light source from the HL-2000 or LS-1 series covers all of these wavelengths best, though the LLS-Warm White broadband white light LED could also be used.

One of the greatest advantages of using an optrode for pH measurement is that the sol-gel matrix allows a much wider range of sampling options and form factors, from transmissive probes to pH-sensitive cuvettes and self-adhesive pH patches. We can also apply our pH sensors as a coating on Petri dishes, microtiter plates, flow cells, or other media where the sample volume may be limited.

All of these form factors have a response time of roughly 30 seconds, though shorter response times may be possible, depending on salinity, temperature, and liquid flow at the sensor interface. The salinity of the solution has the greatest impact on response time. While the accuracy of our pH sensors is not affected by salinity, the time needed to equilibrate and read a stable pH value is determined by the ionic strength of the solution. Higher salinity samples respond more quickly than lower salinity samples.

When working with any of these form factors, it is important to remember that absorbance is being measured, so any movement or change in the optics will affect the absorbance reading and therefore the pH value calculated. It is recommended to leave all optics and sensors in place once calibrated and throughout measurement, including cuvettes.

Transmissive pH Dip Probes

These probes are good for process lines that need continuous monitoring of pH, or for handheld spot-checks in the lab or in the field. They are composed of a high-stability bifurcated fiber (RE-BIFBORO-2) in a T300 or TP300 dip probe sleeve and cap, with a transmissive pH sensor patch applied to the lens. The disposable patches are peel-and-stick, allowing them to be replaced on-site as needed.

These probes are available as kits in stainless steel or in PEEK. Stainless steel probes can corrode in some solutions, notably seawater, making PEEK a good general-purpose choice. The components can also be purchased separately. A 5 mm or 10 mm pathlength tip is recommended to reduce the risk of bubbles being trapped in the probe tip and affecting measurements.

Light travels down the bifurcated fiber, is transmitted through the pH patch, through the sample solution in the probe cap, and is reflected by a mirror in the cap back through the sample and pH patch back into the fiber for measurement by the spectrometer.

Transmissive probes are good for process lines that need continuous monitoring of pH, or for handheld spot-checking in the lab or in the field. They can be Swagelok’d or pressure-fitted into a standard ¼” fitting, which is ideal for process lines and various reaction vessels. Additionally, the patches can be purchased in custom shapes or sizes and applied directly to process vessels for in-situ monitoring of pH.

Typical handheld applications include general laboratory use and R&D as a direct replacement for pH electrodes, as well as field R&D, water sample testing, pools, and aquariums. They have also been integrated into process applications like monitoring of power plant cooling water, food and beverage processing, pharmaceutical processing, and seawater.

Note that transmissive probes are only suitable for samples that are clear and non-turbid. Since the light travels through the sample as well as the pH sensor patch, any color or sediment in the sample will affect the absorbance measurement and invalidate the pH reading.

Smart pH Cuvettes

As an alternative to using a transmissive probe, we have coated the pH sensing material onto the inner walls of standard cuvettes. These are available as plastic (PMMA) cuvettes for general purpose use, or quartz cuvettes for high temperature applications. The PMMA cuvettes themselves can be considered semi-disposable, as they can be used multiple times, but they are not intended for long-term permanent use.

Smart pH cuvettes must be used in a cuvette holder, with fibers to route light from the light source and to the spectrometer. The CUV-UV and direct attach CUV-FL-DA cuvette holders are the most popular choices, though any cuvette holder could be used. A pair of 100 µm or 200 µm core fibers should result in good signal levels (consult the fiber section below for more information on fiber-spectrometer combinations).

To simplify configuration, we offer convenient smart pH cuvette kits. For desktop spectrometers like the USB4000, the SC-PH-DESKTOP-KIT2 contains a CUV-UV cuvette holder, two 200 µm fibers (QP200-025-VIS-NIR), and an 8-pack of coated PMMA cuvettes (SC-PH-VIS1M-SAM). For the Jaz spectrometer, the SC-PH-JAZ-KIT2 is more compact, with a direct-attach CUV-FL-DA cuvette holder, a single 200 µm fiber (QP200-025-VIS-NIR), an SD-MEMORY card for storing data onboard, and an 8-pack of coated PMMA cuvettes (SC-PH-VIS1M-SAM).

Smart pH Cuvettes are ideal for small volume samples and for quick desktop readings, as they are inexpensive and disposable, and have a small footprint. They are often used as a permanent desktop setup for small samples in testing or R&D labs to check cell culture buffers, lake/river/ocean water samples, pharmaceutical solutions like contact storage, eye care, etc., and to test low-conductivity samples such as boiler water. In the field, they are an ideal portable setup for small samples to test environmental water samples, commercial pools, and aquariums.

As with the transmissive dip probe, smart pH cuvettes are only recommended for samples that are clear and non-turbid. Since the light travels through the sample as well as the pH sensor patch, any color or sediment in the sample will affect the absorbance measurement and invalidate the pH reading.

Smart pH cuvettes can be reused simply by washing, with up to 50 uses or more for a single cuvette. The cuvette should be disposed of when absorbance at pH 11 drops to less than 0.1 (relative to a pH 1 reference). High temperatures, high pH, and high pressure, as well photobleaching from the light source will destabilize and reduce the lifetime of the cuvette. All of our pH sensors are very stable at low pH values.

Reflective pH Patches

Samples that are colored or turbid pose an issue for transmissive pH sensors. We have solved that problem by developing a reflective pH sensor that can be applied to the interior of a transmissive container and read with a reflection probe. It consists of our existing transmissive pH sensor material, overlaid with a very fine gold mesh and a top piece of adhesive to create a peel-and-stick patch.

The electoformed gold mesh is liquid-permeable, allowing diffusion of hydrogen ions into the sensor material while acting as a mirror to reflect light back to the detector. Since the light being measured is reflected at the gold mesh and never transmits the sample itself, the sensor is immune to the effects of ambient light, sample color, and turbidity.

The patches themselves come in a package of 5 (PH-BCG-REFLECT), all self adhesive and cut in ½”x½” squares. They are also available in custom sizes and shapes down to 5×5 mm for use on the transmissive wall of a vessel or process line to provide non-intrusive pH sensing.

A specially designed reflection probe is used with reflective pH patches to maximize signal to the detector, allowing measurements to be made through container or vessel walls. The RE-BIFBORO-2 probe uses a randomized bifurcated borosilicate fiber bundle to provide uniform illumination and collection of light, and maximizes signal from the patch.

Reflective pH patches are a clear choice for any application calling for non-intrusive pH measurement. They work very well in bioreactors and flasks for monitoring beer, wine and E. coli fermentation processes. This novel design makes it easy to add a pH sensor to a spin-flask, beaker, fish tank, process line, or any other transmissive vessel. They are almost entirely immune to ambient light and sample color, and are ideal for mud and turbid sample analysis, as well as seawater monitoring.

It is important to choose the right combination of light source, sampling optic, fibers, and spectrometer to get a signal that does not saturate the spectrometer at any pH level in the calibration range of pH 1 – 11, but makes use of the spectrometer’s full dynamic range. The table below gives some guidance on the expected light level for different fibers when working with an HL2000 light source and different USB4000 spectrometer configurations (without an L2 lens).

Any spectrometer capable of measuring absorbance at 620 nm and 750 – 800 nm can be used for pH measurement. This would include the USB series, STS, Jaz, and many others, including preconfigured UV-VIS and VIS-NIR systems. The exact configuration varies depending on the pH sensor form factor being used, but a 50 µm slit will work well in most cases if no L2 detector collection lens is used.

When using a USB4000 spectrometer, the recommended configurations by sampling optic are:

• Transmissive probe: 25 µm slit, no L2, 200 – 850 nm or 350 – 1000 nm
• Smart pH cuvettes: 25 or 50 µm slit, no L2, 200 – 850 nm or 350 – 1000 nm
• Reflective pH patches: 50 or 100 µm slit, no L2, 200 – 850 nm or 350 – 1000 nm

For field measurements, the handheld Jaz offers an easy and portable solution. Its SD card runs a script that allows use of the factory calibration or a complete calibration. It also shows live pH values and gives you the ability to save data directly to the card. If using the Jaz with a direct-attach cuvette holder, use the smaller of the slits recommended above to avoid saturating the detector.

The pH module available for use with SpectraSuite provides simplified calibration, convenient pH readings, customizable data logging and comprehensive exportation of data and calibration information.

It can be downloaded from the software page on the website; a download customized to the Jaz spectrometer is also available. The module allows the user to choose between independent calibration (6 buffers), factory Calibration (3 buffers), and loading a pre-existing calibration file.

The module includes the following additional features:

• Export of calibration file
• Logging of pH data at user-specified intervals
• Export of logged data to an Excel-compatible file, which also has all reference spectra for troubleshooting purposes
• Static temperature compensation, which adjusts the pK value based on the user’s isothermal conditions

The Jaz program is similar, and offers the same calibration and logging options. Note that it is very important to use a specific file structure on the SD card for the program to run correctly.

There are two options for calibrating a pH sensor: a full calibration using 6 buffers, and a calibration reset using the factory calibration value and 3 buffers. LabChem.net is our recommended source for buffer solutions.

The most important thing to remember when calibrating a pH sensor is that the calibration is heavily dependent on the optics. Any change to the light coupling anywhere along the optical path during or after calibration will affect absorbance measurements and therefore the calibration. This includes detaching fibers, moving sensor material relative to the optics, or changing light sources.

An independent, or full calibration uses a complete set of 6 buffers to establish the slope and intercept. This requires buffers for pH 1, 5, 6, 7, 8, and 11. It generates fresh coefficients and yields the most accurate performance from the sensor, though takes more time to complete.

A calibration reset can reduce the time needed for calibration by using only 3 buffers (pH 1, 8 , and 11) in combination with a  factory calibration coefficient (the pK value), measured at the sensor batch level. It effectively resets the slope and maximum absorbance coefficients in the calibration algorithm. Accuracy is best near the reset point (pH 8), and goes down as you move away from it. This type of calibration is adequate for quick, basic measurements, but is not recommended for high accuracy.

Once calibrated, a sensor can re-use a previously stored calibration file. This is extremely convenient and allows for immediate start-up, as no buffers are required. Keep in mind, however, that the optics must have remained entirely unchanged from the last use for it to be valid.

Does every smart pH cuvette need to be individually calibrated before use? Not necessarily. Cuvettes are very reproducible, so once a cuvette from a given batch has been calibrated, others should have roughly ~5% accuracy vs 1-2%. This should be characterized by the user under the exact conditions of use first.

The range of our pH sensors is listed as pH 5 – 9 , which is the region of greatest sensitivity and linearity using the current algorithm. The sensor material still responds outside this range, more notably in the acidic region, and linearity is achieved down to pH ~4.5. A lookup table or a polynomial fit calibration can be used to achieve good results when operating the pH sensor in the acidic range (below pH 5).

Calibration consists of exposing the pH sensor to standard buffer solutions and measuring absorbance. A linear relationship exists between the pH and the ratio of the sample’s absorbance to the change in absorbance relative to absorbance at pH 11. Once the pH sensor is calibrated and its slope and pK (y-intercept) is determined, the calibration can be used to measure pH of unknown solutions.

Since fiber movement, light source instability, and environmental factors can cause the absorbance spectrum to shift vertically, a baseline wavelength is used for correction, making the actual calculation look more like this:

The reference wavelength can be any wavelength for which absorbance does not change appreciably with pH. The isosbestic point at ~510 nm can be used, but since the exact wavelength must be used in order to avoid change in absorbance with pH, it is better to use any wavelength longer than 750 nm.

• Don’t move the optics: Since this is an absorbance measurement, any movement or change in the optics will change the absorbance value at 620 nm, and therefore change the pH value being calculated. For the smart pH cuvettes, customers often remove the cuvette(s) between reference buffers or samples. It is better to leave the cuvette in the holder, and then add or remove liquid with a pipette. For the reflective pH patches, the probe needs to be butted-up against the vessel wall, and a good stand is needed to keep the probe and patch still relative to one another.
• Watch out for bubbles: Ensure there are no bubbles over the sensor interface; the sensor will not give correct values if it is reading a gaseous bubble rather than the actual liquid.
• Carefully install the hardware: Make sure that the patch is attached firmly to the vessel, that it has been “wetted”, that all fiber connections are tight, and have been checked to ensure proper transmission of light.
• Don’t saturate the detector at pH 1: A check should be done to ensure the detector is not saturated when pH 1 is exposed to the sensor. It represents the highest intensity seen by the detector, and is at greatest risk for saturation.
• Allow equilibration time: Moving between calibration buffers without sufficient equilibration time results in a poor calibration. Many users jump quickly between buffers, and then wonder why their numbers are off.
• Allow sufficient response time when referencing: It is easy to move too quickly between taking references in the software, when the sensor is still moving towards its final value. This manifests as an upwards drift or inaccurate readings after calibration is completed. For example, the last reference buffer in the software is pH 8, but if you hit “Acquire and Finish” before it truly levels-out at pH 8, the software assigns that current value to “pH 8.00” and the reading will continue drifting up to 8.02, 8.04, 8.06, and further.
• Ensure you have the proper system configuration: The right combination of light source, spectrometer, sampling optic and fiber must be used that doesn’t give too much or too little light. If the system is saturating, even at the lowest integration times, readings will not be possible. Conversely, if the system has very little light, the software will crash. This is because the software tries to adjust the integration time to 80% saturation, so for very low light this increases the integration time into the seconds range, which coupled with the 10x averaging causes the system to lock up and crash.

These sensors can be safely cleaned using a 50:50 mixture of isopropyl alcohol (IPA) and water. Brief exposure to ethanol (EtOH) is fine, provided it is less than 1 hour. Hydrogen peroxide (H₂O₂) is not recommended.

Ethylene oxide (EtO): All of our pH sensor formats can be sterilized using ethylene oxide (EtO). This includes reflective and transmissive probes, cuvettes, and patches. Lifespan of the probe is unaffected.

Gamma radiation: Both transmissive probes and cuvettes may be sterilized using gamma radiation. Lifespan of the probe is unaffected. This method is incompatible with the gold mesh used in the reflective pH patches.

Autoclaving: Both our T300 and TP300 transmissive probes may be autoclaved, provided no patch is present (the patch material is not high-temperature resistant). Lifespan of the pH probe is limited to about 5 cycles if openly exposed to steam, but is unaffected if an autoclave bag is used. Any custom pH sensors on BK-7 or quartz substrates can also be autoclaved.

A single-point recalibration is required after sterilization for the reflective pH patches, but not for cuvettes/probes so long as no optics have been moved.

• A razorblade is useful in separating the lining.
• Wear gloves so that oils from fingers do not get absorbed into the gold mesh pores and restrict diffusion.
• Push firmly against the patch across the entire surface to ensure good hold.
• Use a pipette to “wet” the mesh. Suck up liquid and jettison it onto the patch several times. Wetting can be perceived visually from the back side. This only needs to be done initially to promote diffusion.
• Position the probe and vessel in a very sturdy arrangement. An optical stage is very useful for this.

The precision of the sensor is limited by the precision of the spectrometer being used, as well as the total absorbance of the sensor. For an indicator with an absorbance of 1, and a spectrometer with a precision of 0.001 AU, a pH precision of ~0.001 pH units is possible with careful calibration.

Many solvents are compatible with our pH sensors, including water, alcohols, ammonia, sodium hypochlorite, and some organic solvents. Acetone and peroxides will degrade the sensor material, and should be avoided.

Group 2 salts and custom buffers pose an issue for our pH sensors. Atypical behavior is seen when measuring solutions containing only or predominantly group 2 salts, such as magnesium, calcium, and strontium. Some mixtures with group 1 and group 2 salts such as seawater, however, give expected and valid responses. Certain buffers have also shown uncharacteristic responses in the past, including tris- and acetic acid-based buffers. We recommend group 1 phosphate buffers for calibration. We use such buffers here at the factory, sourced from LabChem.

High alkalinity samples (pH 9+) can lead to drift during long-term exposure (over 150 hours). This is the upper limit of our recommended range, even with a polynomial fit calibration.

Photobleaching is always a potential concern when light interacts with matter, particularly during long-term continuous monitoring. Fluorescence-based pH sensors are very susceptible to photobleaching over time, which is a major drawback. Our colorimetric approach exhibits a much smaller drift due to photobleaching, which can be mitigated by reducing the cumulative exposure to the light source used.

The triarylmethane dyes used in our pH sensors absorb strongly in the 550 – 600 nm range, so long term exposure to high intensities at these wavelengths has a type of “bleaching” effect similar to leaving a shirt out in the sun for a few weeks. Using a lower intensity light source or an intermittent duty cycle can minimize the effect for measurements over the course of hundreds of hours. Use of a light source at only the specific wavelengths of interest would also reduce the total flux contributing to photobleaching.

USP Class VI Certification represents approval from United States Pharmacopeia, a non-governmental standards setting organization, for our oxygen and pH sensor patches to be used in a range of bio-sensitive applications. Our self-adhesive acrylic patches were tested through extensive leaching studies with aqueous, alcohol-based, oil-based, and organic liquids. Toxicity studies were also performed in mice, rats, and rabbits through injections, and the pH sand oxygen sensor chemistry was found to be non-harmful to all test subjects.

Class VI is its most stringent testing protocol for classification of plastics used in medicines and other health care technologies. With USP Class VI certification, the biocompatibility, toxicity and extractables of Ocean Optics’ sensor coatings are assured to be compatible with biological and pharmaceutical processes and implantable devices.

The approval certifies our pH and oxygen sensor patches for use in applications such as catheters, pharmaceutical processes, and monitoring of long-term drug storage. Testing and certification were provided by the North American Science Associates (NAMSA) in Northwood, Ohio.

All of the pH sensors currently available use an acrylic substrate (PMMA), and as such the temperature limit is about 60 – 70°C. High temperature substrates are available, though the indicator dye chemistry will only survive to 130 – 140°C. We offer the sensor chemistry on BK-7 glass or quartz for higher temperature applications, though increased temperature does shorten the life of the sensor and increases drift during long-term measurements.

All of the current pH sensors use bromocresol green (BCG). Bromocresol green works across the full pH range, but pH 5 – 9 is the range is where the calibration is linear. A polynomial fit can be used to extend the calibration range, though this must be done manually.

There are two other sensor dyes that work well in our sol-gel matrix. These include bromophenol blue for the lower pH range (pH < 5), and bromocresol purple for the higher pH range (pH > 9). If interested in one of these alternate indicators, samples can be sent for testing. Calibration and pH calculation for alternate indicator dyes must be performed outside the software, as it is not yet supported.

The pH of pure water decreases with increasing temperatures, though it is a very small effect. A 4 -5°C change in temperature is needed to see a change in pH of just 0.01. In fact, the pH of pure water at 50 °C decreases only from 7 to 6.55. The Van’t Hoff correlation equations below can be used to compensate for temperature effects if necessary, however.