Semiconductors have been used for sensors since the middle of the 20th century and have been involved in some of the most advanced physical experiments and discoveries. They have helped to uncover the mechanism of radiation-induced damage as well as identifying many nuclear physics principles.
Miniaturized smart sensors based on semiconductors make use of novel materials and new processes to replace the cumbersome and costly conventional sensors. These are used for several applications, including navigation systems, optical devices, and motion sensors, as well as health monitors.
Introduction to Semiconductor Sensors
A sensor detects an event or a change in a factor and responds with an output. The measured phenomenon may be chemical, electrical, mechanical, radiant, magnetic or thermal.
The sensor picks up a signal and passes it to a modifier, the first part of the sensor’s measurement system. This transmits the modified signal to the output transducer. The output transducer emits the final signal in the form of the output. The output is typically an electrical signal.
Most sensors operate using semiconductors and are therefore called semiconductor sensors. The materials most commonly used for semiconductors include silicon and other members of groups III to V. This is because they are readily available, can be easily used, have the required physical characteristics and are cost-effective.
Among these materials, the use of silicon has progressed enormously, allowing the manufacture of miniature silicon circuits less than a micron in diameter. Moreover, batch production techniques allow numerous sensors to be rolled out at the same time at a low cost.
The use of modern semiconductors is based on stacked layers that have different two-dimensional patterns; being built-up to produce a 3-dimensional device. These are manufactured either by surface micromachining or by bulk micromachining. Essential steps in sensor fabrication include deposition, lithography, and etching.
What is the Basic Design of a Semiconductor Sensor?
Semiconductors are in general used for the detection of various kinds of charged particles and photons. Their detection capacity is based on the occurrence of ionization and their applications are extensive in scope.
The crystalline nature characterizes a semiconductor, as lattice symmetry is fundamental to its behavior. The atoms show the presence of electron energy bands with an energy gap between the valence and conduction bands.
Impurities can be introduced (‘doping’) with n-type donor atoms or p-type acceptor atoms located near it in the periodic table, which have excess mobile electrons and holes, respectively. The dopants provide shallow doping levels which ionize at about 300 K. The basis of most semiconductor sensors is the p-n diode and the MOS capacitor.
When two transistors are used at different but unchanging collector current densities, the base-emitter voltages produced are different. The difference is based on the absolute transistor temperature and is converted to an electrical current or voltage. This is finally converted to Celsius or Fahrenheit.
Semiconductor-based temperature sensors, or integrated circuit (IC) temperature sensors, operate with reverse bias, have a small capacitance and a low leakage current. They are formed on thin wafers of silicon. They are compact, produce linear outputs, and have a small range of temperature. They also have a low cost and are accurate following calibration. However, they cannot be interchanged easily.
Many types of semiconductor temperature sensors are available, and are classified into five categories:
- Voltage output - linear output, low output impedance.
- Current output - constant current regulator displaying high impedance.
- Digital output - both sensor and analog-to-digital converter integrated on one chip, typically specific to microprocessor chips rather than measuring devices.
- Resistance output - exploits temperature vs. bulk resistance properties of semiconductors to provide more stable sensors, but may heat up which can be hazardous.
- Simple diodes - the least expensive, but require two-point calibration and stable current input for good results
These sensors are more fragile, both electrically and mechanically, compared to many other types of temperature sensors, and are therefore best used in embedded applications.
Semiconductor temperature sensors are quite sensitive, but since they are usually packaged in standard housing for embedding into an application, the precision measurement of temperature may be affected. Also, interchangeability accuracy is below par. Calibration to two or three-point calibration improves the accuracy, especially with a limited temperature range, and when the change in temperature is small.
How are Semiconductor Sensors Ised?
Microstrips of silicon are used extensively in particle physics experiments, because of the high resolution and efficiency they offer. They are durable and inexpensive, with small magnetic effects. The detector material which is present between a pair of electrodes is affected by the radiation, setting free a variable number of charge carriers. This helps measure the ionizing radiation incident on the strip.
Advanced detectors will help analyze the type of radiation, by their alpha, beta and gamma nature, as well as by activation analysis.
Silicon diodes are also used as optical sensors in telescopes and satellites, in X-ray detectors, or in photodiodes to detect scintillation.
Semiconductor pressure sensors are found in the following four varieties:
- Absolute pressure sensor
- Differential pressure sensor
- Gauge pressure sensor
- Negative pressure sensor
Semiconductor gas sensors convert changes in physicochemical properties into electric current signals, which are picked up by the sensing device.
These have good sensitivity, respond rapidly, are stable over the long term and can be integrated into devices for gas sensing. However, one disadvantage is their non-specificity to gases which share, for instance, the same reducing property or combustible nature.