Shopping Cart

Dew point analyzers, Oxygen analyzers, Oxygen sensors, Pirani Gauges & Thin Film Deposition

Dew point analyzers, Oxygen analyzers, Oxygen sensors, Pirani Gauges & Thin Film Deposition

Dew point analyzer (-80 ~ 20 ℃)

The dew point analyzer ( ) is suitable for various online analysis occasions with strict control requirements for moisture content. It is an economical and universal solution that can meet the control requirements of various industrial processes.

Product information

The online dew point meter adopts imported dew point sensor, which has fast response speed, high sensitivity and easy operation. It can measure the trace moisture content in various gases, and is suitable for various online analysis occasions that have strict control requirements for moisture content. The economical and universal solution can meet the control requirements of various industrial processes.


◆ Imported capacitive dew point sensor with fast response and high sensitivity

◆ Optional precision filter

◆ 128 × 64 dot matrix LCD display, full Chinese operation menu

◆ Simultaneous display of dew point and volume ratio (PPM / V), intuitive reading

◆ Measurement data is automatically stored, recording interval can be set freely, and alarm points can be set arbitrarily

◆ Standard RS232 or RS485 communication port (optional), can realize two-way communication with computer

◆ Has a very long calibration cycle, with a typical value of two years.

Technical Parameters

Display: 128 × 64 liquid crystal display Measuring range: Basic type: -60 ~ 20 ℃ Standard type: -80 ~ 20 ℃ Extended type: -100 ~ 20 ℃ Measuring accuracy: ≤ ± 2 ℃ (0 ~ -60 ℃) ≤ ± 2 ℃ (below -60 ℃) Resolution: 0.1 ℃ Response time: -40 → 10 ℃ 5s (10s) 10 → -40 ℃ 15s (240s) Repeatability: Cv <± 1% Working temperature: -40 ~ 60 ℃ Ambient humidity: 0 ~ 90% RH Sample gas flow: 2 ~ 4L / min Sample gas pressure: 0.05MPa≤Inlet pressure≤0.5MPa Working power: AC220V ± 10%, 50 / 60Hz, Power consumption ≤10VA Analog output signal: 0 / 4 ~ 20mA Digital output signal: RS232 Alarm contact capacity: AC125V 0.3A AC110V 0.5A DC24 3A Measuring medium: H2, He, inert gas, mixed gas, hydrocarbon and other sensors Principle: Thin film capacitive sensor life:> 2 Annual (normal use conditions) Instrument weight: about 2.5kg Installation size: 133mm × 133mm × 250mm (W × H × D) Application fields: petrochemical, natural gas, industrial gas, semiconductor, drying industry, food industry, power, machinery Online measurement of gas and moisture in manufacturing, air separation, and pharmaceutical industries.

Oxygen analyzer(

Oxygen analyzer is an industrial online process analysis instrument, which is not only widely used in the detection of oxygen concentration in mixed gas in heating furnaces, chemical reaction vessels, underground wells, industrial nitrogen production, etc., but also widely used in dissolved oxygen and sewage in boiler Detection of dissolved oxygen outside the treatment plant.

There are many types of oxygen analyzers, different detection principles, and strong pertinence. Therefore, appropriate instruments should be selected according to different applications and different process conditions.

Oxygen Analyzer Type

Oxygen analyzer for zirconia


◆ Integrated structure of sensor and transmitter, aluminum alloy case is lighter and smaller than similar instruments

◆ Large screen LCD dot matrix display, man-machine dialogue

◆ Chinese menu style function selection

◆ Automatic measurement data storage, with paperless recorder function

◆ Arbitrary setting of measured value upper and lower limit alarm output

◆ Automatic range switching

◆ Communication RS232

product description:

The zirconia oxygen analyzer is an oxygen sensor composed of a zirconia solid electrolyte. The signal change is an intelligent online analysis instrument composed of a new microprocessor as the core.

The main purpose:

◆ Air separation, automatic analysis of oxygen content in chemical process

◆ Production of semiconductors and magnetic materials

◆ Float glass, cement building materials industry

◆ Automatic analysis of oxygen content in various industrial furnaces and heat treatment processes

◆ Scientific research on electronic components, biopharmaceuticals, etc.

Trace oxygen analyzer

Instrument Features

Integrated design reduces the influence of external interference on measured values;

The probe uses special materials to make it more resistant to corrosion;

Instrument housing with high protection level;

Built-in display and button design, can ensure the service life of the instrument part even in harsh environments;

Standard DN65 flange mounting makes installation simple and convenient;

High-precision automatic temperature compensation system to eliminate the influence of ambient temperature;

Simple operation, long service life and easy maintenance;

Application areas:

It is widely used in the measurement of flue gas on-line continuous monitoring system CEMS. It can also be used for humidity measurement and control of wood, building materials, papermaking, chemical, pharmaceutical, fiber, textile, tobacco, vegetables, and food processing. This humidity meter can also be used Humidity measurement in high temperature environment such as ceramic drying kiln, electrode drying furnace.

On-line oxygen analyzer

Instrument Features

The original imported wide area ratio sensor is a measuring unit with high measurement accuracy, fast response speed, no power consumption and no life consumption, and good interchangeability.

Calibration at any point can meet the measurement accuracy of the whole range

One-way or two-way RS232 or RS485 communication, can realize one-way and two-way communication with the computer

Measuring range 1PPm-25.00% O2, automatic range switching

No reference gas required, independent of working environment oxygen concentration

High-precision automatic temperature compensation system to eliminate the influence of ambient temperature

Wide range AC power supply for wider application

Application occasion

On-line oxygen analysis of shielding gases such as wave soldering, reflow electronics, soldering furnaces for semiconductor packaging industry and inert gas sintering furnaces.

Precautions for use

Oxygen analysis instruments are generally divided into three categories according to different principles: fuel cell oxygen analyzers, zirconia sensors, and magnetic oxygen analyzers. Oxygen analysis instruments have many precautions in use, otherwise problems such as inaccurate analysis results are very easy to occur.

  1. Oxygen analyzer. Before initial use, leak detection should be performed on connection points, solder joints, valves, etc. to ensure that oxygen in the air will not penetrate into the pipeline and the inside of the instrument, resulting in high measurement values.
  2. Before using the instrument again, purify the pipe system, blow out the leaked air, and ensure that there is no air leak when connecting the sampling pipeline.
  3. The change of the oxygen content in the sample gas will be affected by the material and surface roughness of the pipeline. Therefore, copper pipes or polished stainless steel pipes are generally used for connecting pipelines instead of plastic pipes and rubber pipes.
  4. In microanalysis, it is necessary to avoid all kinds of dead ends such as pipe fittings, valves, gauges and so on to contaminate the sample gas, so the gas path system must be simplified as much as possible, and the dead ends of the connecting parts must be small. Power to prevent the escape of dissolved oxygen and cause pollution. It is best to use water seals, oil seals and wax seals to ensure accurate data.

Oxygen sensor(

  1. What are the types and working principles of oxygen probes

There are two main types of oxygen probes on the market today, namely electrochemical fuel cell and zirconia sensor technology. These two technologies work very differently.

Electrochemical fuel cell technology

The working principle of an electrochemical fuel cell is to use oxygen as the positive electrode and metal lead as the negative electrode, and the electrolyte is potassium hydroxide. The electrolyte and metallic lead are contained in a small, solvent-resistant cylindrical container, and the opening is sealed with a polytetrafluoroethylene film. The use of a completely sealed fuel cell oxygen sensor is one of the current international methods for measuring oxygen. Oxygen diffuses into the battery through a polytetrafluoroethylene membrane. When oxygen gets electrons on the positive electrode and metal lead loses electrons on the negative electrode to generate a current, the current is proportional to the oxygen concentration.

The negative reaction formula is: 2Pb -4e-== 2Pb2 +

The positive electrode is: O2 + 2H2O + 4e-== 4OH-

Total reaction: O2 + 2H2O + 2Pb == 2Pb2 + 4OH

  1. Analysis of advantages and disadvantages of two oxygen probes

Advantages of electrochemical fuel cells:

  1. The zero point is accurate, not easy to drift, only a single point correction is needed; it can be seen from the chemical reaction formula that when there is no oxygen, there is no current, so the zero point of the electrochemical fuel cell is accurate, and only a single point correction is required for correction.
  2. Strong resistance to organic solvents. Electrochemical fuel cell detection of oxygen is a redox chemical process at normal temperature, so it is less affected by a small amount of common organic solvents.
  3. The replacement cost is low, and only the fuel cell needs to be replaced. When the fuel cell fails and the oxygen analyzer fails, the user only needs to replace the chemical fuel cell in the oxygen probe, and the replacement cost is less than one-third of the entire probe price.


  1. Contacting the oxygen probe with high-concentration oxygen for a long time will affect its life. The battery needs to be stored in a high-purity inert gas, which is not a problem in the glove box application.
  2. Cannot be used to detect high temperature gas.

Zirconia sensor


  1. The reaction time is fast, and the oxygen concentration can be measured at high temperature, high pressure and low pressure, such as boiler exhaust gas;
  2. It can be in contact with air when it is not tested, and can be stored in air for a long time, because the reference gas of the zirconia sensor is air.


  1. Not suitable for gas containing organic solvents; zirconia sensors need to work at high temperatures. When the detection gas contains an organic solvent, the organic solvent will chemically react on the electrode, and the oxygen will be consumed by reacting with the organic matter, so that the measured value of oxygen is lower than the actual value.
  2. The zero point is easy to deviate. When measuring PPM-grade gas, it is often necessary to use standard gas for calibration. Because air is used as the reference gas, when 1PPM gas is detected, the concentration difference is over 200,000 times, and the accurate detection is difficult. Calibrate with a standard gas before each test.
  3. Only the entire oxygen probe can be replaced, and the replacement cost is high.
  4. Which oxygen probe is more suitable for glove box

Based on the above analysis, electrochemical fuel cell technology is more suitable for testing trace oxygen in glove boxes. Itex glove box selects the gas detection technology suitable for the atmosphere of the glove box——electrochemical fuel cell technology, which has the advantages of high accuracy and no influence of organic solvents, and avoids the problems of zirconia oxygen analyzer affected by organic solvents.

Pirani vacuum gauge(

Pirani vacuum gauges, and translational Lanny vacuum gauges. This kind of vacuum gauge is a kind of thermal conductivity vacuum gauge.


Using the Wheatstone bridge’s compensation principle, one of the filaments acts as one of the resistors in the circuit. When the gas molecular density of the filament changes, its thermal conductivity will be different, so the temperature displayed by the filament will also be different, which indirectly affects the resistance value. In the fixed voltage, fixed current and fixed temperature, the last method is used for sensitivity. It is worth noting that the thermal conductivity of different gases is different, so calibration is required when measuring.

The vacuum gauge in this line judges the pressure of the gas by measuring the degree of heat conduction between a heating element and a receiving heating element.

When the Kn value (Knudsen number) is between 0.01 <Kn <10, the amount of heat conduction is proportional to the pressure.

At low pressures, heat can be dissipated in three ways, namely the heat conduction of the surrounding gas, heat radiation, and the main body of the pillar heating element.

In order to improve the accuracy of the measurement pressure, the heat loss of the latter two types is avoided as much as possible.

Structure and composition

The vacuum gauge is generally divided into two parts:

Induction head: mostly metal or glass shell with filament or other temperature sensing element to sense vacuum pressure.

Control head: Provide necessary circuits for the sensor head and do signal amplification and signal digitization.


Pirani vacuum gauge

Thermoelectric vacuum gauge

Thermistor vacuum gauge

Constant current type

The temperature of the filament will change with the change of vacuum pressure. Since the filament is a material with a high temperature coefficient of resistance, the change in the temperature of the filament will cause the resistance value of the filament to change. The resistance change can be measured by the resistance bridge sensing method. The change in vacuum pressure was demonstrated.

The rightmost of the Wheatstone bridges in the figure below has a sensor head filament and a reference sensor head filament. The reference sensor head has a structure very similar to that of the sensor head, but its interior has been evacuated to a relatively high vacuum state and sealed, and is as close as possible to the pressure sensing position of the sensor head, which has the function of ambient temperature compensation. The other side of the bridge has a variable resistor R1 and a fixed resistor R2 in this order. The bridge is connected in parallel with a power source, and a balanced current meter G is indirectly connected between the two induction heads and two resistors.

When the vacuum pressure in the sensor head is changed, the resistance value of the filament resistance of the sensor head will also change along with it, which will cause the reading of the balanced ammeter in the electric bridge to be different from zero. At this time, the variable resistor R1 in the bridge can be adjusted to make the reading of the balanced ammeter reset to zero. When the current provided by the power supply is constant, the change in the resistance of the filament can be detected by the changes in the voltage of the two sections of the bridge, and the vacuum pressure value can be read from the voltmeter across the bridge through the conversion and proofreading work of the party.

Constant voltage

When a certain voltage is maintained on the filament in the induction head, the temperature of the filament will change with the vacuum pressure. Take the vacuum pressure as an example, at this time, the heat energy taken away by heat conduction will increase, and the temperature of the filament will decrease. Since the filament is a high resistance temperature coefficient material, the temperature of the filament will decrease, which will cause the resistance value of the filament to increase. When the voltage does not change, the current through the filament will decrease, so the decrease in current also proves that the vacuum pressure increases. Conversely, when the vacuum pressure becomes low, it can be proved that the vacuum pressure becomes low from the increase of the current.


Effect of different gases on Pirani vacuum gauge

Because heat transfer is not only related to pressure, but also to the molar amount and molecular structure of the gas. Different gases can lead to different results when measuring pressure.

In general, the larger the atoms or molecules of an existing gas, the lower the capacity or efficiency of heat conduction.

Application range

chemical industry

Electronics and microelectronics

Vapor deposition

Electron beam welding

Vacuum metallurgy

Performance characteristics


Large measuring range

Good reproducibility

Low production cost and reasonable price

Faster response time


Different gas types affect measurement accuracy (usually calibrated with nitrogen or air)

Measure pollution sensitive

Thin film deposition technology

Before understanding the lift-off pattern transfer, you first need to understand the thin film deposition technology suitable for the lift-off method. There are many methods for depositing thin films, including physical and chemical vapor phase methods, molecular beam epitaxy methods, spin coating or spraying methods, and electroplating methods. But commonly used are physical and chemical vapor deposition.

The so-called physical vapour deposition (PVD) includes thermal evaporation deposition and plasma sputtering deposition. Thermal evaporation deposition can be divided into resistance evaporation deposition and electron beam evaporation deposition according to different evaporation methods. Plasma sputtering deposition can also be divided into DC sputtering, RF sputtering, and magnetron RF sputtering according to different plasma generation methods. Chemical vapour deposition (CVD) includes low pressure (LPCVD), atmospheric pressure (APCVD), plasma enhanced (PECVD) and metal organic compound (MOCVD). In recent years, a new type of thin film deposition technology, namely atomic layer deposition (ALD), has received widespread attention.

Comparison of main features of common thin film deposition techniques

To investigate whether a thin film deposition method is suitable for the pattern transfer of the stripping method, two points need to be paid attention to: first, the substrate temperature rise during the deposition process, and second, the directivity of the deposition. Since the stripping method relies on the photoresist pattern as a mask, each photoresist has a stripping transition temperature (Tg). When the temperature is higher than Tg, the photoresist starts to soften and flow, causing the photoresist pattern to collapse. Or deformation. As can be seen from Table 2, the CVD deposition method generally requires a higher substrate temperature, and is therefore not suitable for pattern transfer by the lift-off method [1].

The directivity of the deposition directly determines whether the deposited film can be successfully peeled off. Figure 2 (a) is the result of thin film deposition and peeling under good deposition direction. When the photoresist film layer is removed, only the area not covered by the mask leaves the metal film pattern. Table 2 (b) shows the results of thin film deposition and peeling in the case of poor directivity. Due to the poor directivity of the deposition, a thin film is also deposited on the sidewall of the photoresist pattern, connecting the top of the photoresist with the metal film on the substrate as a whole. When the photoresist layer is removed, one possibility is that the entire metal film is also removed along with the photoresist layer, which may happen when the deposited metal film is not firmly bonded to the substrate. Another possibility is that as shown in Table 2 (b), although the metal film on the top of the photoresist is removed, the metal film on the side wall is integrated with the metal film on the substrate and remains on the substrate. Based on the above two points, only thermal evaporation thin film deposition can meet the requirements of substrate temperature rise and good directivity at the same time, so it is also most suitable for pattern transfer by solvent stripping.

Thin film deposition(  process

Thin film deposition is an essential link in the manufacturing process of integrated circuits. Traditional thin film deposition processes mainly include PVD, CVD and other vapor deposition processes:

PVD (Physical Vapor Deposition): Under vacuum conditions, a physical method is used to vaporize the surface of a material source (solid or liquid) into gaseous atoms, molecules or parts of it, and then ionize it through a low-pressure gas (or plasma) process. Technology for the surface deposition of thin films with some special functions. The main methods of PVD include vacuum evaporation, sputtering coating, etc., not only can deposit metal films, alloy films, but also compounds, ceramics, semiconductors, polymer films, etc. The materials involved include all solids (C, Ta, W difficult) , Halides and thermally stable compounds.

CVD (Chemical Vapor Deposition): A method that uses one or more vapor-phase compounds or elements containing thin-film elements to perform a chemical reaction on the substrate surface to form a thin film. The CVD method can produce thin film materials including metals other than alkali and alkaline earth (difficult to Ag, Au), carbides, nitrides, borides, oxides, sulfides, selenides, tellurides, metal compounds, alloys, and the like.


As integrated circuits become more integrated and smaller in size, high-k dielectrics are gradually replacing traditional silicon oxide gates. At the same time, the aspect ratio is getting larger and larger. Capabilities have put forward higher requirements, so ALD has been increasingly adopted as a new deposition process that can meet the above requirements:


ALD (Atomic Layer Deposition): It can be understood as a phase-change CVD process. A method of forming a deposited film by alternately passing pulses of a gas phase precursor into a reactor and chemisorbing and reacting on the substrate. Different from traditional CVD, in the process of ALD, the reaction precursors are alternately deposited, and the chemical reaction of the new layer of atomic film is directly related to the previous layer. In this way, only one layer of atoms is deposited per reaction. . ALD deposited materials include metals, oxides, carbon (nitrogen, sulfur, silicon) compounds, various semiconductor materials and superconducting materials.

ALD deposition material


Compared with traditional PVD and CVD deposition processes, ALD makes full use of surface saturation reactions. It is inherently capable of thickness control and high stability, and is less sensitive to changes in temperature and reactant flux. Therefore, the ALD-deposited film has both high purity and high density, is flat and highly conformable, and can achieve good step coverage even for structures with aspect ratios up to 100: 1. The main disadvantage of ALD was that the deposition speed was slower, about 1 Å / min. However, as the thickness of the deposited thin film layer is getting thinner and thinner, the impact of this disadvantage is no longer a problem. ALD is becoming more and more widely used in gate oxide layers, diffusion barrier layers, and electrode thin film layers in memory structures.

Comparison of PVD, CVD and ALD process characteristics


For the ALD process, the selection of the precursor usually needs to meet the following requirements: (1) sufficient vapor pressure at the deposition temperature to ensure that it can fully cover the surface of the substrate material; (2) good thermal and chemical stability, Self-decomposition does not occur at the deposition temperature; (3) High reactivity, which guarantees rapid chemical adsorption on the surface of the substrate, or rapid and effective reaction with the surface groups of the material, so that the surface film has high purity; (4) The reaction by-products are not corrosive to the substrate and surface film; (5) The materials are widely sourced and have low toxicity.


Commonly used ALD precursors include non-metallic precursors and metal precursors. Non-metallic precursors such as halides (SiCl4, AlCl3, etc.), nitrides (NH3, (CH3) NNH2, BuNH2, etc.), metal precursors such as alkyl precursors (Ga (CH3) 3, Mg (C2H5) 2), β-diketone precursors (La (thd) 3, Ca (thd) 2), alkoxide precursors (Ta (OC2H5) 5, Zr [(OC) (CH3) 3] 4), alkylamines and silamines Base precursors (Ti [N (C2H5CH3) 2] 4, Pr [N (SiMe3) 2] 3), etc.


The manufacturing process of 3D NAND is very complicated. It mainly includes high aspect ratio trenches, no doping on source or drain, and perfectly parallel walls. Tens of stairsteps, Uniform layer across wafer, Single-Lithostairstep, Hard mask etching , Processing inside of hole, Deposition on hole sides, Polysilicon channels, Charge trap storage, Through-hole etching of various materials ( Etch through varying materials), composite multilayer film deposition (Deposition of tens of layers), etc.

3D NAND basic process flow


Because 3D NAND complex structures require high aspect ratios, related processes include stack deposition, high aspect ratio channel hole etching, word line metallization, step etching, high aspect ratio slit etching, and step contact line forming. Among them, stacked deposition and in-line metallization place extremely high requirements on the deposition process. In this respect, the ALD process has advantages over the traditional CVD and PVD processes.

3D NAND structure and key deposition and etching processes


The ALD process can effectively reduce stress. The manufacturing process of 3D NAND stacked memory cells starts with alternating thin film deposition. It is important to accurately control the uniformity of the thickness of each layer. Wafer warpage and local film stress directly affect the lithographic stacking accuracy, film thickness and repetition. The characteristics affect the effective volume of the memory cell and the consistency of the lithography / etching performance. Therefore, thin film stress control and good uniformity are critical to wafer yield. At the same time, in the 3D NAND using the replacement gate process, the wire connection of the memory cells in the same layer is realized by tungsten filling. The traditional chemical CVD tungsten film has high tensile stress, which will cause wafer warpage, and the fluorine element brought by the process Will spread to adjacent layers, causing defects and affecting yield. The ALD process using low-fluorine tungsten (LFW) can produce a smoother surface topography and more closely fit each filling layer, thereby reducing the stress generated by the deposition process. Compared with traditional CVD tungsten deposition technology, ALD low-fluorine tungsten technology can reduce the stress (GPa → hMPa) by more than an order of magnitude, the fluorine content of 99%, and the resistivity of more than 30%.

Higher uniformity of ALD deposited films with minimal stress


In addition to effectively controlling stress, ALD has better step coverage and can meet the requirements of high aspect ratio in 3D NAND manufacturing. With the increase of the number of 3D NAND layers, the aspect ratio of the channel is also increasing. During manufacturing, the vertical and lateral high-K dielectric (Al2O3) and titanium barrier layer (Ti / TiN) are required in the 100: 1 aspect ratio channel. And other substances. Because the ALD deposition process can effectively control the thickness and uniformity of the thin film, it can achieve uniform coverage of high aspect ratio channels, while PVD and CVD cannot achieve uniform coverage for steps with high aspect ratio. A step with an aspect ratio that is too high may cause the top opening to become blocked.

Methods such as PVD often fail to achieve uniform coverage at high aspect ratios


Not only is 3D NAND the same, the most important and difficult challenge for planar DRAM is the high aspect ratio of storage capacitors. The storage capacitor’s aspect ratio will increase multiples as the component process shrinks, which will make the planar DRAM process shrink more and more difficult. Therefore, the ALD process also has important applications in the flat DRAM and NAND.

Storage industry drives precursor market to expand, 3D structure stimulates demand growth


The precursor is the main raw material of the semiconductor thin film deposition process. IC precursors can be summarized as: a class of substances that are applied to semiconductor manufacturing processes, carry target elements, are in a gaseous or volatile liquid state, have chemical thermal stability, and have corresponding reactivity or physical properties. In semiconductor manufacturing processes including thin film, photolithography, interconnection, and doping technologies, precursors are mainly used in vapor deposition (including PVD, CVD, and ALD) to form various thin film layers that meet the requirements of semiconductor manufacturing. Semiconductor precursors can be divided into: TEOS (ethyl orthosilicate), boron phosphorus (B, P) dopants, metal precursors, high-k precursors, low-k precursors, and so on.

Precursor materials for semiconductors

With the development of downstream chip, memory and other chip industries, the overall market size of precursors will maintain rapid growth. According to Japan’s Fuji Economic Data, the sales volume of global precursors was about 85.5 billion yen (about 750 million US dollars) in 2014, an increase of 28.81% year-on-year; it is expected to reach 135.8 billion yen (about 1.20 billion US dollars) in 2019, and the CAGR will reach 10 %.

According to a report by Techcet, a semiconductor materials market research company recommended by Chemical and Gases Manufacturers Group of SEMI, the complex 3D structure has greatly stimulated the demand for high-k deposition processes, driving the rapid growth of high-k and metal precursor markets, 2015 The high-k and metal precursor market size for ALD / CVD was USD 185 million in 2019, and it will grow to USD 325 million in 2019, with a CAGR of approximately 15%.

Leave a Reply

Your email address will not be published.

Free Worldwide shipping

On all orders above $50

Easy 30 days returns

30 days money back guarantee

International Warranty

Offered in the country of usage

100% Secure Checkout

PayPal / MasterCard / Visa