我刚刚开始制作DIY空气净化器的时候，很多人发邮件问我能不能制作一个车载净化器。我当时第一反应是“这主意太棒了！”——高速公路上充斥着可怕的污染，我们当然要净化空气！
但作为一个nerd(书呆子)，我想先做些测验。那些实验结果却是我始料未及的——我们不需要车载空气净化器！理由如下：
Uber汽车测试
我带着我的Dylos DC1700激光粒子计数器走上了上海的街头（这样的测试仪可靠么？）。空气质量是这样的：
左边的数字表示0.5微米及以上的粒子数目，安全值大约是2,500。所以可见当天空气质量较差。而右边的数字表示2.5微米及以上的较大粒子数。
结果
之后我打了Uber，继续测量空气。上车后的空气质量如下图：
好多了！0.5微米的粒子数只是户外空气所含粒子数的25%。几分钟后：
妈妈你看，不用净化器！现在降至户外空气的11%。
车里的空气为什么干净得多呢？
原来汽车空调系统里已有内置滤网了，并且他们过滤颗粒污染物效果很好。下图是汽车内置的滤网，在美国亚马逊才10块美元。
由于自己本身没接触过什么空气净化器的实验，而且网上也没有说明这些滤网的级别，根本不清楚滤网是否能捕获小粒子，所以这些数据迟早能派上用场！
普通车实验
这个简单的测试表明汽车的内置滤网能大大降低微粒数目，甚至是那些小于2.5微米的！但是坐一次Uber不太严谨，这辆汽车本身是否有特殊性呢？为了确保精准，我与Ohad（一位Smart Air粉丝）在以色列用他的车做测试。他开的是一辆普通的2010年马自达3，没有安装任何特殊的滤网或净化器。
测试方法
Ohad开车同时用Dylos粒子计数器测量。一进车就开了空调，紧闭车窗，使用 “车内循环空气”模式。就这样，五分钟之内以分钟为单位持续测量。测试时间不长，所以这次车厢测量还是挺保守的。
等一下，以色列空气污染真的有那么严重吗？当然没有北京严重，但以色列的很多地方最近污染严重超标。在Ohad的实验里，0.5微米粒子变化数从2,464（将近WHO的上限）到10,000以上（上限的四倍）。所以空气中的颗粒够多！
结果
我求出了所有8次实验的平均，结果如下：
当空调开至最大功率时，5分钟内的实验平均值结果显示了0.5 微米粒子减少97.9%，2.5微米粒子减少99.3%。最近两个广东人也做了类似的实验，结果跟我的测试很相似。
原来我们的汽车本身已经是非常好的空气净化器了！另配车载空气净化器厂家就要哭了。
不开空调测试
不过等一下，精明的读者也许会问：我们怎么知道这要归功于车载空气系统，还是车内的人在吸入粒子呢？亦或只是单纯由于粒子逐渐沉降呢？
为了看看是不是车内空气系统对减少粒子数在起作用，Ohad关闭了空气系统，再进行了一次一样的实验。
在空气系统不开的实验中(红色线)，颗粒数目只出现了一点下降。我猜这主要是因为颗粒沉降。不过汽车空气系统开着的时候(蓝线)，下降得更快。
如果比较8次开空调和5次不开空调的测试，开空调时（左）消除的颗粒百分比不开空调（右）高：
结论
不需要特殊车载空气净化器防止微粒空气污染，因为汽车空气系统已经最大程度减少了微粒污染。我们只需这么做：
1. 把车窗关上。
2. 为了最快地降低PM2.5，使用 “车内循环空气” 模式。
3. 风量开中档或高档。
4. 每隔几个月就检查并更换滤网。
5. 好好想想是不是真的需要一个汽车空气净化器。
例外：气体污染
但还有个例外：汽车滤网能捕获颗粒，但它们不会捕获气体，比方说NO2。因此，也可以考虑用活性炭滤网。（活性炭真的能减少气体污染吗？）
任何人都不需要用汽车空气净化器吗?
如果我在中国有汽车的话，自己不会买一个汽车空气净化器，不过我也不认为所有人都不该拥有。比如说，对污染特别敏感的人或者有小孩子的父母可以考虑用空气净化器加强保护对抗空气污染。
公开数据
和往常一样，原始数据和实验方法细节都是公开的，方便和我一样的实验控们看看。也有空调高档、 低档和不开空调的比较实验。
Hello, I just randomly stumbled into this old article while doing research on this subject and noticed some flaws I’d like to point out! Not that I expect anyone to answer at this point, but just to clarify things for people like me stumbling into this after all this time!
The argument seems to be “old cars have the same quality air filtration systems as Tesla HVAC’s ‘biodefence mode'” if I understood this correctly. This argument is then justified incorrectly by comparing Tesla to an old car recirculating air already inside the car through a non-HEPA filter over and over again. It should be obvious the air will get cleaner and cleaner like this (probably linearly at about the same pace that the air is being filtered). The reality of the matter is: Tesla has a state of the art HEPA filtration system that doesn’t recirculate air, but takes in new air and then purifies it. The car is also over-pressurized to avoid leaks from outside from getting in without being filtered. The same methods are used in pharmaceutical cleanrooms during medicine manufacturing!
The problem with recirculating air is that while it will also have low levels of outside pollutants like HEPA filtered air, it will start to build up CO2 levels and moisture and the air quality will become bad (not from outside pollutants though).
You can really see the difference in quality of these systems when compared fairly in Tesla’s video comparing their “biodefence mode” (=HEPA+other methods) with BMW’s stock filtration system:
https://www.youtube.com/watch?v=Dx9wrIfqhjM
Can you do the same test in the car Tesla model S who had a medical quality HEPA Filter , please ?
Cabin air filter seems to clean particulates effectively, and as you said, it will not work this good against gases (like NO2, VOCs). Did you perform some tests with gases too? Since some cars might only have an activated carbon filter (like my fiesta), then using a car air purifier would make sense in places like Delhi.
It depends on the filter efficiency rating of the vehicle manufacturer. Having used the Dylos and other improved particles with algorithms that convert to mass concentration, I found many vehicles in India with less than desirable ac cabin filter results ( most were nowhere as good as the Mazda or the Uber in your tests). Not surprisingly the foreign car manufacturers’ results were as good as the Mazda. Some European vehicles have HEPA filter too.
I think car air purifiers are an excellent concept. Thanks for sharing this article!
Have you repeated the tests with the AC on but not in recirculation mode?
Good question! We didn’t do that. My guess is it’d be less effective (but still quite effective). For long trips, we might want to use the “bring in outdoor air” mode because CO2 would probably go pretty high in the car.